<<< SHAVE::DECUS:[DECUSERVE]HARDWARE_HELP.NOTE;1 >>> -< HARDWARE_HELP >- ================================================================================ Note 1.0 Welcome to the HARDWARE_HELP conference No replies EISNER::MEDLIN "Terry Medlin" 3 lines 24-MAR-1987 21:05 -------------------------------------------------------------------------------- The purpose of this conference is to discuss hardware problems, glitches, ECOs, diagnostics, how-to questions, etc.  ================================================================================ Note 2.0 This topic reserved for future use by the moderator No replies EISNER::MEDLIN "Terry Medlin" 2 lines 24-MAR-1987 21:05 -------------------------------------------------------------------------------- The most current information concerning this conference will be posted as responses to this topic.  ================================================================================ Note 3.0 This topic reserved for future use by the moderator No replies EISNER::MEDLIN "Terry Medlin" 2 lines 24-MAR-1987 21:05 -------------------------------------------------------------------------------- The most current information concerning this conference will be posted as responses to this topic.  ================================================================================ Note 4.0 This topic reserved for future use by the moderator 1 reply EISNER::MEDLIN "Terry Medlin" 2 lines 24-MAR-1987 21:05 -------------------------------------------------------------------------------- The most current information concerning this conference will be posted as responses to this topic. ================================================================================ Note 4.1 This topic reserved for future use by the moderator 1 of 1 EISNER::PROVOST 154 lines 8-JUN-1988 13:05 -< Table of Contents >- -------------------------------------------------------------------------------- Below is an index of TOPICs in this conference. Those marked with a * are generic, and should be looked at periodically. Others are more specific to particular subjects. -------------------------------------------------------------------------------- HARDWARE_HELP Created: 24-MAR-1987 21:05 147 topics Updated: 7-JUN-1988 09:00 Topic Author Date Repl Title -------------------------------------------------------------------------------- 1 EISNER::MEDLIN 24-MAR-1987 0 Welcome to the HARDWARE_HELP conference 2 EISNER::MEDLIN 24-MAR-1987 0 This topic reserved for future use by the moderator 3 EISNER::MEDLIN 24-MAR-1987 0 This topic reserved for future use by the moderator 4 EISNER::MEDLIN 24-MAR-1987 0 This topic reserved for future use by the moderator 5 EISNER::MEDLIN 24-MAR-1987 9 Live Video Problem Needs a Switch * 6 EISNER::PROVOST 25-MAR-1987 60 HARDWARE HINTS AND KINKS (HHK) * 7 EISNER::MAXWELL 25-MAR-1987 3 Information on Old Hardware * 8 EISNER::EISNER 26-MAR-1987 11 Hardware Look-a-loos * 9 EISNER::PROVOST 26-MAR-1987 10 HARDWARE WISH LIST 10 EISNER::CETRON 26-MAR-1987 2 gpx monitor cable extention * 11 EISNER::PROVOST 1-APR-1987 13 ECO/FCO's for common problems 12 EISNER::FRISBIE 11-APR-1987 1 Long access times on RD53/RQDX3 13 EISNER::KASPER 14-APR-1987 28 Backup Devices 14 EISNER::MCGLINCHEY 19-APR-1987 7 Not another DLVJ1 problem! 15 EISNER::LEFEBVRE 20-APR-1987 14 Cabling Solutions 16 EISNER::COAR 18-MAY-1987 1 VT52 parity problem 17 EISNER::PROVOST 19-MAY-1987 16 KA620 18 EISNER::KILLEEN 5-JUN-1987 4 DOCUMENTATION ERRORS - HARDWARE 19 EISNER::KILLEEN 10-JUN-1987 16 MODEM AND COMMUNICATION EQUIPMENT 20 EISNER::GARDNER 16-JUN-1987 2 RX50 drive differences 22 EISNER::KILLEEN 17-JUN-1987 14 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 23 EISNER::FRISBIE 22-JUN-1987 15 MMJ to DB25 help needed 24 EISNER::GOLDSTEIN 30-JUN-1987 2 VT240,2planes,sixels HELP 25 EISNER::KILLEEN 1-JUL-1987 20 PDP-11 CPU PERFORMANCE 26 EISNER::HORN 14-JUL-1987 0 DEBET/DEREP fiber question 27 EISNER::PROVOST 21-JUL-1987 26 POWER 28 EISNER::DELARISCH 5-AUG-1987 13 VT240 Problems 29 EISNER::HORN 20-AUG-1987 4 VT340 video output? 30 EISNER::HANRAHAN 24-AUG-1987 2 finding DEQNA rev. level 31 EISNER::KILLEEN 25-AUG-1987 16 VTXXX TERMINALS 32 EISNER::PINSLEY 25-AUG-1987 8 Bar Codes 34 EISNER::FRISBIE 27-AUG-1987 11 Ethernet transceivers and cables 35 EISNER::TILLMAN 31-AUG-1987 2 If I had more memory, I might not forget so much 36 EISNER::KILLEEN 31-AUG-1987 3 PORTABLE TERMINALS 37 EISNER::KILLEEN 1-SEP-1987 2 HARDWARE NEWS FLASHES 38 EISNER::NORTON 2-SEP-1987 6 DECserver/DELNI problem 39 EISNER::WALLIS 2-SEP-1987 2 Configuring a Microvax II File Server 40 EISNER::KILLEEN 5-SEP-1987 33 PERIPHERAL DEVICE ANNOUNCEMENTS 41 EISNER::KILLEEN 5-SEP-1987 3 PDP-11 SYSTEM ANNOUCEMENTS 42 EISNER::KILLEEN 5-SEP-1987 11 MICROVAX SYSTEM ANNOUCEMENTS 43 EISNER::KILLEEN 5-SEP-1987 21 VAX SYSTEM ANNOUNCEMENTS 44 EISNER::ETHINGTON 9-SEP-1987 7 RDxx FCT Format 45 EISNER::LEDERMAN 9-SEP-1987 12 No SIXEL through VT125 printer port? 46 EISNER::PRIGOT 9-SEP-1987 13 New MicroVAXen? 47 EISNER::BYRNE_C 10-SEP-1987 49 Can You Speed up an 11/70? 48 EISNER::ROSEBERRY 14-SEP-1987 5 FCO kit for TA/TU78 49 EISNER::FRISBIE 14-SEP-1987 2 microVAX-II to microVAX-3xxx CPU upgrade 50 EISNER::MERUSI 15-SEP-1987 2 DMZ-32 Bit Droppings 51 EISNER::FRISBIE 15-SEP-1987 12 RX33 for BA23 or BA123 boxes? 52 EISNER::WALLIS 17-SEP-1987 3 ?? Hot Sites ?? 53 EISNER::MACNEIL 18-SEP-1987 1 MUXserver 100 & DECMux II 54 EISNER::PRIGOT 21-SEP-1987 3 SA482? 55 EISNER::NORTON 21-SEP-1987 3 Problem with borrowed RA81 56 EISNER::WALLIS 25-SEP-1987 3 When is the HSC saturated? 57 EISNER::KASPER 25-SEP-1987 8 Laser Printer Questions 58 EISNER::TILLMAN 30-SEP-1987 1 Emulex performance 59 EISNER::DROPPERS 5-OCT-1987 10 How to update an HSC 60 EISNER::NORTON 6-OCT-1987 2 LG01 clone? 61 EISNER::HAHN 7-OCT-1987 4 STAT-MUX 62 EISNER::DROPPERS 15-OCT-1987 10 FCOs for Self/Third Party Maint. 63 EISNER::NORTON 16-OCT-1987 4 RA60 Data recovery? 64 EISNER::STAMERJOHN 16-OCT-1987 2 RQXD3, but no manual 65 EISNER::WALLIS 16-OCT-1987 2 6 MIPS + 6 MIPS = ? 67 EISNER::MCDOUGALL 21-OCT-1987 5 VT125 Info 68 EISNER::PERRY 28-OCT-1987 4 Need RA70 help for LAVC 69 EISNER::TILLMAN 29-OCT-1987 5 DEPCA information sought 70 EISNER::KENNEDY 30-OCT-1987 1 Help needed w/ RD53 / RQDX2 combination 71 EISNER::RHODE 4-NOV-1987 3 Need help interfacing GPX to plotter 72 EISNER::BYRNE_C 6-NOV-1987 9 Datascopes and Protocol Analyzers 73 EISNER::LEDERMAN 9-NOV-1987 1 DHU-11 problems and fixes 74 EISNER::KENNEDY 17-NOV-1987 18 BA23 power switch needed 75 EISNER::NORTON 18-NOV-1987 6 IRIS/VAX INTERCONNECTION 76 EISNER::DROPPERS 19-NOV-1987 2 HSC50 lost its NAME 77 EISNER::RENES 22-NOV-1987 1 VAX 8530 performance myth 78 EISNER::DOHERTY 28-NOV-1987 4 360K Floppies on VaxMates 79 EISNER::BYRNE_C 3-DEC-1987 29 IBM (!) PC configurations 80 EISNER::MCDOUGALL 8-DEC-1987 6 TU78 on VAX 8550? 81 EISNER::HORN 10-DEC-1987 1 VAXmate clock/battery 82 EISNER::RICE 13-DEC-1987 1 Interfacing an optical scanner 83 EISNER::MERUSI 16-DEC-1987 5 RD53 Problems 84 EISNER::BYRNE_C 16-DEC-1987 7 VAX Processor timings 85 EISNER::STAMERJOHN 17-DEC-1987 1 PC Lap-top's 86 EISNER::PERRY 17-DEC-1987 8 Need Q-bus help... 87 EISNER::OSUDAR 21-DEC-1987 17 tape drive hangs system 88 EISNER::FRISBIE 22-DEC-1987 0 VR260 documentation needed 89 EISNER::RICE 6-JAN-1988 9 VAX <==> Apple Laserwriter 90 EISNER::KILLEEN 7-JAN-1988 6 TK50, TK70 & THE VAXstation 3200 91 EISNER::KOZAM 9-JAN-1988 3 What Hardware Documents Do You use? 92 EISNER::NAGY 14-JAN-1988 3 What's in that box (RD32)? 93 EISNER::KENNEDY 16-JAN-1988 14 Low-end disk devices - The Digital difference 94 EISNER::BOWLER 18-JAN-1988 0 Synch BI-unibus devices 95 EISNER::NORTON 21-JAN-1988 5 Looking for displays 96 EISNER::LEDERMAN 22-JAN-1988 6 VT125 doesn't work at 19,200 Baud? 97 EISNER::RICE 31-JAN-1988 0 RX33 on a PRO? 98 EISNER::KILLEEN 1-FEB-1988 5 WHAT IS THE BEST WAY TO SELL OFF SOME HARDWARE? 99 EISNER::PERRY 5-FEB-1988 1 Neet KA620/630 SBC VAX help... 100 EISNER::ROBITAILLE 10-FEB-1988 3 Looking for a good key... 101 EISNER::OSUDAR 12-FEB-1988 2 LN03 weirdness 102 EISNER::TILLMAN 19-FEB-1988 4 Western Datacomm - Vadic 4224 103 EISNER::KENNEDY 23-FEB-1988 22 DF124 switch setting info needed 104 EISNER::BYRNE_C 25-FEB-1988 4 Ethernet Hardware Interface Chips 105 EISNER::FRISBIE 4-MAR-1988 1 M9301 Bootstrap info wanted 106 EISNER::RICE 7-MAR-1988 11 Lap_top Terminals 107 EISNER::DELARISCH 7-MAR-1988 12 LA-50 problems! 108 EISNER::DELARISCH 9-MAR-1988 9 Q-BUS Board for TU-80? 109 EISNER::FRISBIE 9-MAR-1988 2 uV3000 Machine Checks 110 EISNER::ROSEBERRY 15-MAR-1988 7 Third Party Maintenance 111 EISNER::KILLEEN 17-MAR-1988 6 DEC RS232 VS DEC RS423 112 EISNER::ROSEBERRY 25-MAR-1988 1 LN03 Toner Options 113 EISNER::RENES 7-APR-1988 2 2nd DW750 installtion woes 114 EISNER::KILLEEN 8-APR-1988 0 TK25 115 EISNER::KILLEEN 9-APR-1988 5 CSU/DSU DEVICES ON DDS TERMINATIONS 116 EISNER::KILLEEN 9-APR-1988 3 9600 BAUD DIALUP MODEMS 117 EISNER::OSUDAR 12-APR-1988 5 Emulex LAT server 118 EISNER::KILLEEN 13-APR-1988 3 TK25 BOOT ON A PDP-11/73A 119 EISNER::LEDERMAN 13-APR-1988 3 LQP03 a viable item? 120 EISNER::OSUDAR 19-APR-1988 20 VAX 62xx question(s) 121 EISNER::MAYHEW 19-APR-1988 11 The LN03 OPC is dead -- Long live the LN03 OPC 122 EISNER::HUTTO 21-APR-1988 4 Can a bad RA81 crash a good HSC50? 123 EISNER::NORTON 22-APR-1988 9 MXV11B question 124 EISNER::GARDNER 25-APR-1988 0 MSV11-SA help needed 125 EISNER::HAHN 25-APR-1988 1 mVAX test meanings 126 EISNER::DEC_NEWS_1 25-APR-1988 2 VAX System Positioning Announcements 127 EISNER::BORCHARDT 27-APR-1988 0 SI93C disk drives 128 EISNER::PINSLEY 27-APR-1988 1 LN03 AC/AD KITS 129 EISNER::DROPPERS 28-APR-1988 2 Keyswitch Crashes 8650... 130 EISNER::DEC_NEWS_1 30-APR-1988 4 How to use VAX 6200s 131 EISNER::DEC_NEWS_1 30-APR-1988 2 Hardware and Software Service Bundled 132 EISNER::RENES 5-MAY-1988 0 Versatec -BI/Unibus woes 133 EISNER::SEASTREAM 6-MAY-1988 3 PDP-11 / VAX Shared Memory 134 EISNER::JOE 11-MAY-1988 0 DV11 TIMING PROBLEM 135 EISNER::BRUCKER 11-MAY-1988 1 SID/XSID register layouts 136 EISNER::KOZAM 11-MAY-1988 2 Can you boot a VAxstation 2000 through the printer port? 137 EISNER::TILLMAN 13-MAY-1988 0 No MNP on -2574 138 EISNER::TILLMAN 23-MAY-1988 1 Support for the DEQNA will disappear 139 EISNER::HAHN 24-MAY-1988 0 Other Mailer services for repairs 140 EISNER::NORTON 24-MAY-1988 4 Callback modem use with KERMIT 141 EISNER::ZIMMERMAN 25-MAY-1988 4 Moving the LP27 and LP06 far away? 142 EISNER::ROSEBERRY 29-MAY-1988 3 Speed Up a VAX 11/780 ? 143 EISNER::BENDER 1-JUN-1988 3 Let's talk DECtalk 144 EISNER::KILLEEN 4-JUN-1988 12 18-BIT VS 22-BIT QBUS DEVICES 145 EISNER::KILLEEN 5-JUN-1988 2 MICRO NOTES 146 EISNER::HASSINGER 6-JUN-1988 2 CD-ROM questions 147 EISNER::KILLEEN 6-JUN-1988 3 RL02 PARTS  ================================================================================ Note 5.0 Live Video Problem Needs a Switch 9 replies EISNER::MEDLIN "Terry Medlin" 15 lines 24-MAR-1987 21:27 -------------------------------------------------------------------------------- Suppose you have an office of say 25 VT100/VT101s and you want to be able to see a live copy of any terminal by "flipping a switch". This would include both input and output. The terminals will be initially hooked via DZ11 boards to a VAX 11/750 running VMS. There is a partial software solution offered thru third party firms. I am not aware of a complete software solution A hardware solution is to run coax from each terminal to a coax patch panel and then a "switch" to connect any of the coazes to a separate monitor (or VT100). Anyone have a source for this "switch" or a better idea. ================================================================================ Note 5.1 Live Video Problem Needs a Switch 1 of 9 EISNER::EISNER "Dan L. Eisner" 14 lines 24-MAR-1987 23:55 -< Some starting points >- -------------------------------------------------------------------------------- A few questions. How far is it to from the terminals to the distribution panel/switch? Would it be easier to make up a RS-232 line monitor? (a bunch of diodes will do the trick) Since the computer echos all characters, looking at the transmitted output should show what is happening. Coaxes to a switch may get you in trouble if the line lengths are long. If the coax route is tried, be sure to use 75 ohm coax. RS-170 defines that the impedance is 75 ohm. Line switchs can be had from any company that sells to the TV industry. Start with the local cable company for a reference to a local distributer. Next is to call a TV studio and ask for the Cheif Engr. Explain your problem and he should really give you an ear full (most I have known love to talk trade). ================================================================================ Note 5.2 Live Video Problem Needs a Switch 2 of 9 EISNER::PROVOST 6 lines 25-MAR-1987 09:50 -< Alternatives >- -------------------------------------------------------------------------------- I believe there exists a solution if you connect all terminals via terminal servers, or if you force all users to log in over network. Not that computer does not echo all characters, such as passwords, ^C, etc. Tom Provost ================================================================================ Note 5.3 Live Video Problem Needs a Switch 3 of 9 EISNER::CETRON 5 lines 26-MAR-1987 12:41 -< software alternative >- -------------------------------------------------------------------------------- I believe that the solution from clyde digital here in utah is totally software and allows use to audit,monitor,break in and assist ...... -ed ================================================================================ Note 5.4 Live Video Problem Needs a Switch 4 of 9 EISNER::HASSINGER "Bob Hassinger" 14 lines 7-MAY-1987 00:57 -< VT100 as a video monitor??? >- -------------------------------------------------------------------------------- >> < Note 5.0 by EISNER::MEDLIN "Terry Medlin" > >> A hardware solution is to run coax from each terminal to >> a coax patch panel and then a "switch" to connect any >> of the coazes to a separate monitor (or VT100). For a separate monitor, yes, but have you ever seen this work going into a VT100? The monitor has to sync to the incoming video. All conventional monitors do this but I have never seen a VT100 sync to it's video input. They seem to free run, creating their own sync. I think you may find you can build a video switch without much trouble if you have more time than money and don't mind getting your hands a little dirty. As noted, passive, pushbutton video switchs are easy to buy in the professional video market. ================================================================================ Note 5.5 Live Video Problem Needs a Switch 5 of 9 EISNER::COAR "Ken A L Coar - OPUS/OPS" 9 lines 18-MAY-1987 14:59 -< AUDIT from CDS - good stuff >- -------------------------------------------------------------------------------- > I believe that the solution from clyde digital here in utah is totally > software and allows use to audit,monitor,break in and assist ...... I tried out the stuff from Clyde Digital, and, with the exception of a couple of minor problems involving RTxx terminals, it worked 100% as advertised - very slick. As close to a complete software solution as I have ever seen. #k ================================================================================ Note 5.6 Live Video Problem Needs a Switch 6 of 9 EISNER::TINKELMAN "Bob Tinkelman" 29 lines 2-JUN-1987 10:44 -< Problems using VT100 Video-In jack >- -------------------------------------------------------------------------------- << < Note 5.4 by EISNER::HASSINGER "Bob Hassinger" > << << >> < Note 5.0 by EISNER::MEDLIN "Terry Medlin" > << >> A hardware solution is to run coax from each terminal to << >> a coax patch panel and then a "switch" to connect any << >> of the coaxes to a separate monitor (or VT100). << << For a separate monitor, yes, but have you ever seen this work going << into a VT100? The monitor has to sync to the incoming video. All << conventional monitors do this but I have never seen a VT100 sync to << it's video input. They seem to free run, creating their own sync. We actually tried to run a cable from a VCR to the the Video-In jack on the back of the VT100 a couple of years ago and discovered that it would NOT work. The problem, as indicated by Bob in 5.4 is that the VT100 insists on generating its own synch pulses (think of it as the timing of when it wants to start each sweep the electron beam and when it wants to start the top line of the screen). It expects anyone feeding its Video-In jack to be listening to the VT100 Video-Out jack and to use those synch pulses for the signal coming in. Naturally, our "home VCR" did not have the ability to accept "external synch". I expect only real expensive professional equipment, like the king meant to run in a TV studio (all of which runs "in synch") would have such capability. In any case, VT100 to VT100 just won't work. ================================================================================ Note 5.7 Live Video Problem Needs a Switch 7 of 9 EISNER::HASSINGER "Bob Hassinger" 15 lines 2-JUN-1987 16:16 -< TBC... >- -------------------------------------------------------------------------------- > I expect only real expensive professional equipment, like the > king meant to run in a TV studio (all of which runs "in synch") would > have such capability. Actually, even the low end professional VCRs I have worked with will not sync up directly. In my experience the standard solution is a "time base corrector" that goes between the VCR and the destination that requires sync'ed video. A TBC typically contains enough memory to buffer a full video frame so it can deliver it starting at the time the external sync pulse calls for it. Low end TBCs are going for a few thousand $s these days. They offer a possible solution to some needs like this but I would think you would need to review resolution (bandwidth) specs, particularly on the real low end stuff, because 80x24 terminals use a lot more bandwidth than a typical normal video signal. ================================================================================ Note 5.8 Live Video Problem Needs a Switch 8 of 9 EISNER::HOUGHTON "John N Houghton - CompuServe" 26 lines 1-JUL-1987 17:47 -< Source for you-build-it video switch >- -------------------------------------------------------------------------------- > Suppose you have an office of say 25 VT100/VT101s and you > want to be able to see a live copy of any terminal by > "flipping a switch". This would include both input and > output. The terminals will be initially hooked via DZ11 > boards to a VAX 11/750 running VMS. > A hardware solution is to run coax from each terminal to > a coax patch panel and then a "switch" to connect any > of the coazes to a separate monitor (or VT100). > Anyone have a source for this "switch" or a better idea. Steve Ciarcia published plans for an active video switch in his Circuit Cellar column in Byte sometime in the past year, I believe. Unfortunately I don't have a set handy to get the exact reference. It won't deal with the reported sync problem (which I've never encountered, but I haven't tried running a VT100 to a VCR, only commercial monitors and other VT100s, which worked fine). The sync should work fine between VT100s even with the reported sync problem - try it with a length of cable first. If you do have a sync problem, there are cheaper fixes then the timebase corrector, but will require a hardware type (preferably someone who can build a Cable TV descrambler - the fix is similar to the fix for a suppressed sync scrambled signal). Of course, you'll need to find someone who can build the switch for you. ================================================================================ Note 5.9 Live Video Problem Needs a Switch 9 of 9 EISNER::HAHN 8 lines 17-AUG-1987 17:08 -< BYTE reference >- -------------------------------------------------------------------------------- BYTE February 1986 pp85-99 Build and Audio-and-Video multiplexer. The device uses a 8x8 cross point switch (74HC22106) any eight devices to anyother eight devices. If all you want is to select one of 25 and see it on one monitor, there are simpler switches. Pierre  ================================================================================ Note 6.0 HARDWARE HINTS AND KINKS (HHK) 62 replies EISNER::PROVOST 26 lines 25-MAR-1987 10:28 -------------------------------------------------------------------------------- This topic will be to discuss known glitches and fixes in DEC hardware. DRV11W - Latest rev (E) of DRV11W doesn't work with DEC uVMS driver. Older DRV11W's work fine. A new DRV11W can be made to work by installing jumpers W2, W3, W6. This information is from the Fall DECUS Symposium Hardware Hints and Kinks session. A customer purchased a DRV11W rev-E and found it indeed did not work. It came with jumpers W1, W3, and W5 installed. The jumpers are binary. Moving W5 to W6 did not help. Can anyone shed light on this? DRV11-WA - Apparently this device does not provide all the signals one is accustomed to having on a DRV11W. More information would be appreciated. uVAX CPU boards numbered M...H may produce machine checks 81, 82, 83. This results from same NEC chips which caused replacement of memory boards for uVAXen. Unfortunately we have been unable to determine the FCO which replaces these boards. Anyone know? VAXstation II - VCB01 boards do not understand multilevel interupts. When a higher priority device is on the BUS farther from the CPU than the VCB01, the VCB01 steals the interrupt grant. The driver then appologizes by making an entry entitled "ILLVEC" in the errorlog. This makes for poor real-tim performance. The problem can be avoided by putting the VCB01 farther out on the BUS than any foreign devices, since DEC doesn't use multi-level interrupts. tjp ================================================================================ Note 6.1 HARDWARE HINTS AND KINKS (HHK) 1 of 62 EISNER::MAXWELL "Gary Maxwell" 8 lines 25-MAR-1987 19:19 -< FPJ11 Availability/Reliability >- -------------------------------------------------------------------------------- What is the latest information on the FPJ11? I still have a "recent" (less than one year old) one that crashes M-Plus when a Fortran-77 program is executed from Batch (yeah, figure that one out, folks) So when and how can one get one that works? Are there really problems with the FPJ11 working with 3rd party memory boards? Which ones? ================================================================================ Note 6.2 HARDWARE HINTS AND KINKS (HHK) 2 of 62 EISNER::EISNER "Dan L. Eisner" 7 lines 26-MAR-1987 00:35 -< DRV11-WA info >- -------------------------------------------------------------------------------- DRV11-WA rev e. I fixed the hardware by correcting the software. The delta between the rev's were the addition of an interrupt enable flip-flop. The driver was corrected by ignoring the first interrupt from the hardware. If the ignore interrupt code is removed, then the hardware works ok. I will check the changes and document them here. I used a DRV11-WA straight out of the box and it worked ok for an application that used a DR11-W before. ================================================================================ Note 6.3 HARDWARE HINTS AND KINKS (HHK) 3 of 62 EISNER::CETRON 6 lines 26-MAR-1987 12:45 -< microvax v1.2 roms >- -------------------------------------------------------------------------------- uVAX II v1.2 roms still search disks in wrong order (2,1,0) and is fixed (finally) in v 1.3 roms, supposedly field circus will update you (without official fco/eco) if you are on contract. -ed ================================================================================ Note 6.4 HARDWARE HINTS AND KINKS (HHK) 4 of 62 EISNER::FRISBIE "Alan E. Frisbie" 28 lines 27-MAR-1987 00:50 -< DR11-W vs DRV11-W vs MDB "DR(V)11-W" >- -------------------------------------------------------------------------------- Consulting my manuals for the DR11-W (Unibus) and the DRV11-WA (Q-bus), I find the following differences: Pin DR11-W DRV11-W J1-D ACLO FNCT 2 H INIT V2 H J1-K BURST RQ L SINGLE CYCLE H J1-X END CYCLE H GND J1-Z CYCLE RQ B H GND J2-X GO GND I believe that J1-K is really the same signal with two different names, but you can't take anything for granted these days. My question is, why did DEC make them different? It would have been trivial to make them totally compatible. MDB makes a Q-bus board that has the exact same pinouts as the DR11-W Unibus board and, in 18-bit addressing mode, is programmed the same as the DR11-W. In 22-bit addressing mode, however, it adds the necessary BAE register and shifts all the registers to non-standard addresses. It also fails to make bits 4 & 5 of the CSR read-only in 22-bit mode like DEC did with the DRV11-W. This caused me to spend two days in wonderful downtown Freeport, Texas this week while I debugged the driver. The lesson is: READ THE MANUALS, AND DON'T TRUST WHAT YOU THINK YOU KNOW! Alan ================================================================================ Note 6.5 HARDWARE HINTS AND KINKS (HHK) 5 of 62 EISNER::EISNER "Dan L. Eisner" 66 lines 30-MAR-1987 22:55 -< XA Driver changes for DRV11-WA Rev E >- -------------------------------------------------------------------------------- DRV11-WA, Rev-E VMS changes. I removed the following lines in the driver at change level X-6 (DGB0141). After XA_CONTROL_INIT: After note about IE bit; Comment out: BBS #XA_CSR$V_IE...9$ BBSS #UCB$V_IGNORE_UNEXPT...9$ This takes care of part of the ignore of the 1st interrupt. After BLOCK_MODE: After label 35$: Comment out: CMPW (SP)...MPL BNEQ 37$ CMPB #DT$_...(R5) BEQL 37$ BBS #XA$V_LINK...40$ THIS takes care of some operation incomplete errors due to bus interference. The system seems to correct itself during the retry code. Your main line code should have retries in it. After XA_INTERRUPT: After or at 25$: Comment out; BBSC #UCB$V...30$ This takes care of "If we get an interrupt then it really was an interrupt". After XA_REGISTER: After 55$: Comment out; CMPB #DT$_XA....(R5) BEQL 57$ Leave in; BISB #,XA_CSR(R4) CLRW XA_CSR(R4) This will correct for the unexpected interrupt. If you use any of the ststus bits to reset the remote device, be sure to enable them in the previous MOVW instruction. If there are any questions feel free to contact me thru this conference. ================================================================================ Note 6.6 HARDWARE HINTS AND KINKS (HHK) 6 of 62 EISNER::PROVOST 28 lines 8-APR-1987 16:29 -< ILLVEC and VCB01 >- -------------------------------------------------------------------------------- In the HMS HHK section ofhe SIGs Newsletters, I mentioned a problem with the VCB01 board on the monochrome VAXstation II stealing interrupts intended for his superiors. I just got a call from Bill James, in Cambridge, England, who has encountered the problem on a VAXstation II GPX. The symptom is the appearance in the errorlog of ILLVEC bugchecks. I discovered that these disappeared when I removed the VCB01 (monitor board). I then checked the microfiches for the VCDRIVER, and found that the VCDRIVER makes these entries in the errorlog whenever it gets an interrupt that it thinks it should not have gotten. I suspected my weird devices might be interrupting at priority higher than 5, and that the VCB01 was not monitoring the higher level interrupt lines. This is a common problem with DEC devices on the Q-BUS. I checked with a contact at Digital, and found this to be the case. Case closed, and reported in the newsletter. Problem is that I believe Bill James' equipment all runs DEC software, so is unlikely to use higher level interrupts. So, what is the problem? The ILLVEC message can come from VCDRIVER, but perhaps it can also come from elsewhere. I have been told that when a BA123 VAXstation gets loaded down, it sometimes garbages the interrupt vector. This results from a timing problem in which the interrupt is removed from the lines before the CPU has latched it. This would result in unexplained interrupts, which uVMS might log as ILLVEC bugchecks. Unfortunately my source for this vector degeneration information seems to have dried up. Has anyone seen this kind of problem? Can anyone supply more information? tjp ================================================================================ Note 6.7 HARDWARE HINTS AND KINKS (HHK) 7 of 62 EISNER::PROVOST 9 lines 8-APR-1987 16:38 -< /REWIND may be required >- -------------------------------------------------------------------------------- AVIV 825 or AVIV 925 controllers with STC 2925 tape drives skip first save set when using BACKUP without /REWIND. These same controllers when used with STC 1953 or STC 2923 drives have no problem.ÿ I have seen this problem in the past with early versions of the TU81 and VMS software. I have not tried all combinations of switch settings in the controller. Since the main difference between the 2923 and the 2925 is that the 2925 has cache, I should try turning off the cache to see what happens. ================================================================================ Note 6.8 HARDWARE HINTS AND KINKS (HHK) 8 of 62 EISNER::EISNER "Dan L. Eisner" 6 lines 8-APR-1987 23:45 -< Non-reset due to BOT may be the culprit >- -------------------------------------------------------------------------------- It could be that the cache is not cleared at BOT. If there was an error in the read of the first block, you would not want to clear the cache. Need to check the logic of the controller. This would be the correct type of oversite error. Disable cache, and check if their are any switch settings to clear the cache at BOT. These are just some thoughts. ================================================================================ Note 6.9 HARDWARE HINTS AND KINKS (HHK) 9 of 62 EISNER::CETRON 12 lines 9-APR-1987 22:38 -< the gpx blues >- -------------------------------------------------------------------------------- I have a gpx which is an upgrade from an mvII. I have next to it the identical mvII (serial number + 1) except the gpx also has an rqdx3 and the mvII only an rqdx2. I run a small program which opens an unformatted file of info, actually 512 bytes per block of 512 blocks (actually a bitmap) reads the data 1 block at a time and then exits. this program executes on the mvII in about 10 secs. which is the same as 4 other mvII's around. The gpx runs the same program in about 19secs. EVEN WITHOUT THE VADRIVER OR GPX SOFTWARE running. maybe this is further proof that the vcb0x's snarf up unused AND used cycles on the bus ================================================================================ Note 6.10 HARDWARE HINTS AND KINKS (HHK) 10 of 62 EISNER::FRISBIE "Alan E. Frisbie" 10 lines 10-APR-1987 02:03 -< Faster is slower, sometimes >- -------------------------------------------------------------------------------- I believe that the difference in time could be accounted for by the fact that the RQDX2 requires sector interleaving, while the RQDX3 does not. The inter-sector delay on the RQDX2 systems is enough time for the program to issue a QIO before the disk gets to the next desired sector. On a RQDX3, the disk is already in the middle of the next desired sector, so it must wait an entire revolution. Newer and faster isn't always better! Alan ================================================================================ Note 6.11 HARDWARE HINTS AND KINKS (HHK) 11 of 62 EISNER::EISNER "Dan L. Eisner" 5 lines 10-APR-1987 11:38 -< Interleave the program for a test >- -------------------------------------------------------------------------------- I order to resolve the question. Why not change the}i 512 X{512 program to interleave. If the time factor decreases then the before mentioned problem is the culprit. Alan, could the problem be the Disk problem you ran into? ================================================================================ Note 6.12 HARDWARE HINTS AND KINKS (HHK) 12 of 62 EISNER::CETRON 2 lines 10-APR-1987 22:21 -< interleave what? >- -------------------------------------------------------------------------------- interleave what? the program jsut does 512 reads of 512 bytes... ================================================================================ Note 6.13 HARDWARE HINTS AND KINKS (HHK) 13 of 62 EISNER::EISNER "Dan L. Eisner" 4 lines 11-APR-1987 00:09 -< Interleave method >- -------------------------------------------------------------------------------- Read block 0 then block 2 then block 4 then block 6... at >512 switch to read block 1, block 3 ...etc. That way the software can read the disc faster. Maybe. I thought it may be worth a try to see if there is any differences. ================================================================================ Note 6.14 HARDWARE HINTS AND KINKS (HHK) 14 of 62 EISNER::FRISBIE "Alan E. Frisbie" 6 lines 11-APR-1987 02:36 -< Refer to new note for disk delay problem >- -------------------------------------------------------------------------------- Dan, I don't think that Ed's problem (with the disk!) is related to the disk delay problem that I had. For the benefit of the rest of the people, I will type it up as a separate note. Alan ================================================================================ Note 6.15 HARDWARE HINTS AND KINKS (HHK) 15 of 62 EISNER::MCGLINCHEY 4 lines 19-APR-1987 13:01 -< Everybody to the back of the Bus? >- -------------------------------------------------------------------------------- Tom - see my note on DLVJ1's - DEC is still configuring them into uV II's. ================================================================================ Note 6.16 HARDWARE HINTS AND KINKS (HHK) 16 of 62 EISNER::MCCARTHY 13 lines 12-MAY-1987 19:53 -< It's interleaving >- -------------------------------------------------------------------------------- re: .9,.10 and others on RQDX2/3 performance ... This is almost definitely caused by the interleaving. The RQDX3 is optimized around multi-block transfers, since VAXen typically do same. We noticed this when testing the RQDX3 for RSX. Also, I believe that the first sector to be read after the seek completes is used only to verify positioning, so the effective rotational latency for single block reads is 5 or so percent higher than the RQDX2s. Again this is noise for larger transfers. -Brian ================================================================================ Note 6.17 HARDWARE HINTS AND KINKS (HHK) 17 of 62 EISNER::CETRON 3 lines 15-MAY-1987 14:54 -< wrongo, its the vcbox;s >- -------------------------------------------------------------------------------- sorry guys, the 'other 4' mvII's also had rqdx3's... the ONLY slow machine is the one with the gpx board set... ================================================================================ Note 6.18 HARDWARE HINTS AND KINKS (HHK) 18 of 62 EISNER::PROVOST 2 lines 15-MAY-1987 15:02 -< VCB0x => ILLVEC >- -------------------------------------------------------------------------------- Have you checked you error log for "ILLVEC" errors? ================================================================================ Note 6.19 HARDWARE HINTS AND KINKS (HHK) 19 of 62 EISNER::COAR "Ken A L Coar - OPUS/OPS" 13 lines 18-MAY-1987 15:13 -< If RMS is involved, disk speed no longer matters >- -------------------------------------------------------------------------------- > ... I run a small program which opens > an unformatted file of info, actually 512 bytes per block of 512 > blocks (actually a bitmap) reads the data 1 block at a time and > then exits. Unformatted? This sounds like a FORTRAN program using OPENs and READs. If this is the case, RMS is buffering the data in, several blocks at a time, and interleaving would not help in any case. If the program is using $QIOs and accessing the data directly, interleaving might indicate something - but if RMS is involved, forget it. (Unless you want to SET RMS /MULTI=1..) #k ================================================================================ Note 6.20 HARDWARE HINTS AND KINKS (HHK) 20 of 62 EISNER::PROVOST 21 lines 4-JUN-1987 15:44 -< E.O.T. problems >- -------------------------------------------------------------------------------- We have a VAXstation II running uVMS. We acquire data in real time from an MBD CAMAC branch driver on a UNIBUS connected via an ABLE Microverter to the Q-Bus. The data is written to an STC 1953 tape drive (GCR) via an AVIV controller. We write to the E.O.T. strip. When that is detected, we write 5 E.O.F.'s and rewind the tape. We then read the tape and analyze the data. When we read the logical end of tape (3 EOTs) we rewind and start over. This sequence fails intermittently. Symptom is read process reading tape off end of reel. Sometimes the sequence succeeds. Some tapes never fail. Some never succeed. Some are intermittent. We have found that success is independent of tape vendor, tape quality, and tape size. We have found success/failure to be independent of hardware. (We have a spare VAXstation with a full set of equipment.) Any ideas out the? ================================================================================ Note 6.21 HARDWARE HINTS AND KINKS (HHK) 21 of 62 EISNER::PROVOST 2 lines 4-JUN-1987 15:46 -< Also indep. of blksize >- -------------------------------------------------------------------------------- I forgot to mention... we tried a variety of block sizes... no differences. ================================================================================ Note 6.22 HARDWARE HINTS AND KINKS (HHK) 22 of 62 EISNER::CETRON 13 lines 4-JUN-1987 22:21 -< ....a baggy a day... >- -------------------------------------------------------------------------------- first question, what kind of tape drive. I will just guess that it is a tsvo5 or similar. We have a cipher m890 or some such which is the cachetape 1600/3200 version of the tsv05 and I had some real problems reading SDC tapes since they were so shiny and would convince the drive that it had read EOT. So as any hardware hacker would do, I took a piece of the plastic bag from the doc set and covered the eot/bot sensors. woops!!!! it cut out too much light and I had funny problems similar to yours. I put a baggy over the sensors and all was well. wierd huh?? -ed ================================================================================ Note 6.23 HARDWARE HINTS AND KINKS (HHK) 23 of 62 EISNER::EISNER "Dan L. Eisner" 14 lines 5-JUN-1987 01:36 -< Possible EOT fix >- -------------------------------------------------------------------------------- Top, I had the same problem a few years back. It turned out to be the controller. The controller would sample the EOT sensor at the end of the block when it was read. If you write over the EOT tape with data, the data may be longer than the EOT spot. The write command sees the EOT since it finds it via the edge detection. The read command looks for a level. I fixed it by making longer EOT sensors, you can get the tape in HI-FI shops, Calculate the longest block + the interecord gap*2. in inches and make it bigger than that. The problemis that you have to go to the EOT to add the tape. The other fix was hardware. I added a Flip-flop to the tape transport that was turned on when the EOT was found by the light and then turned off due to reverse direction and EOT being on. You will also need to add a reset to insure that power-on resets it. If you want, give me a call and I will discuss in more detail. ================================================================================ Note 6.24 HARDWARE HINTS AND KINKS (HHK) 24 of 62 EISNER::PROVOST 8 lines 5-JUN-1987 10:06 -< I'LL TRY IT. >- -------------------------------------------------------------------------------- Tape drive is STC (Storage Technology Corporation) 1953 tri-density 125 ips. The controller emulates a TS11. I think I'll try the lllllloooooooooonnnnnngggggg EOT trick to verify that the problem is similar. Since we write about 2000 2400' reels per year, I don't think adding EOT strips is a final solution. However, if I can prove that the problem is understood, I should be able to get the hardware modified by AVIV or STC. ================================================================================ Note 6.25 HARDWARE HINTS AND KINKS (HHK) 25 of 62 EISNER::MAXWELL "Gary Maxwell" 9 lines 5-JUN-1987 16:53 -< two more bits worth on EOT problem >- -------------------------------------------------------------------------------- We have had similar problems with our 1953's and we have a System Industries controller. Our 1953's are real old (very small serial numbers). Since we don't write to EOT very often, we don't see the problem very often. BRU never gives us a problem at EOT. It's always caused by some silly application. :-} ================================================================================ Note 6.26 HARDWARE HINTS AND KINKS (HHK) 26 of 62 EISNER::PROVOST 22 lines 9-JUN-1987 10:00 -< RQDX3 LUN jumper locations >- -------------------------------------------------------------------------------- I just stumbled over a problem, so I thought I'd pass the word so you don't fall on the same one. I wanted a second disk drive on my VAXstation II temporarily. Knowing that the documentation was inadequate and erroneous concerning the switch settings on RD disks added to an RQDX3, I decided to save myself some trouble by moving the RQDX3 along with its RD53. (I have spare systems.) I then discovered that the documentation is inadequate and erroneous concerning the jumper settings on the RQDX3! When adding a second RQDX3 to a VAXstation II, you must specify use of LUNs 4-7, instead of 0-3. This is done with a jumper. No indication of what jumper does what. The jumpers for CSR address could be interpreted from the old CSR seting, but the old LUN setting was no jumpers. Looking at older documentation, I found that the RQDX2 had the LUN jumpers clearly indicated on the board diagram. Unfortunately the RQDX2 looks nothing like an RQDX3. I called my friend Jeff Killeen, and he informed me the right jumper was W12. Hope this saves someone some delay someday. Tom ================================================================================ Note 6.27 HARDWARE HINTS AND KINKS (HHK) 27 of 62 EISNER::PROVOST 14 lines 9-JUN-1987 11:26 -< Who needs documentation? >- -------------------------------------------------------------------------------- < < Note 6.26 by EISNER::PROVOST > < -< RQDX3 LUN jumper locations >- < I wanted a second disk drive on my VAXstation II temporarily. < Knowing that the documentation was inadequate and erroneous < concerning the switch settings on RD disks added to an RQDX3, < I decided to save myself some trouble by moving the RQDX3 along < with its RD53. (I have spare systems.) While my back was turned, someone connected the second RD53 directly to the RQDX3. It works with no switches changed. So much for following directions! Tom ================================================================================ Note 6.28 HARDWARE HINTS AND KINKS (HHK) 28 of 62 EISNER::KILLEEN 8 lines 9-JUN-1987 21:12 -< RQDX3 DOCUMENTS >- -------------------------------------------------------------------------------- Two very valuable and cheap documents are: RQDX3 CONTROLLER MODULE USER'S GUIDE (EK-RQDX3-UG(-001)) RQDXE EXPANDER MODULE USER'S GUIDE (EK-RQDXE-UG-(-002)) The number at the end are the rev levels ================================================================================ Note 6.29 HARDWARE HINTS AND KINKS (HHK) 29 of 62 EISNER::KILLEEN 25 lines 9-JUN-1987 21:43 -< RD,RX DRIVE SELECT USAGE >- -------------------------------------------------------------------------------- > I wanted a second disk drive on my VAXstation II temporarily. > Knowing that the documentation was inadequate and erroneous > concerning the switch settings on RD disks added to an RQDX3, For the most part the drive select switches on the RDn drives are ignored. When the drive is mounted in a BA123 box the cable from the M9058 SIGNAL DISTRIBUTION BOARD controls the units drive number designation. As a matter of practice the drive select should be set to 3. DRIVE SELECT USAGE TABLE BA23 BA123 EXTERNAL BOX -DA, -RA RD51,52,53,54, USED IGNORED IGNORED 31,32 SEL3=DRIVE 1 SET TO SEL3 SET TO SEL3 SEL4=DRIVE 2 RX50 N/A N/A N/A RX33 SEL1=DRIVE 1 SEL1=DRIVE 1 SEL=DRIVE 1 SEL2=DRIVE 2 SEL2=DRIVE 2 SEL=DRIVE 2 ================================================================================ Note 6.30 HARDWARE HINTS AND KINKS (HHK) 30 of 62 EISNER::KILLEEN 60 lines 9-JUN-1987 22:29 -< MICRO-11 DOCUMENTS >- -------------------------------------------------------------------------------- MICRO-11 USEFUL DOCUMENTION - WILL WORTH THE COST ZYAAB-GZ - DOC SET (BA123 BOX) MICRO PDP11 SYSTEMS TECH MANUAL (BA123 BOX) AZ-GLHAA-MC MICRO PDP11 OWNER'S MANUAL (BA123 BOX) AZ-GLIAA-MC ZYAAA-GZ - DOC SET (BA23 BOX) MICRO PDP11 SYSTEMS TECHNICAL MANUAL (BA23) EK-MIC11-TM MICRO PDP11/53 SYSTEMS SUPPLEMENT MANUAL (BA23) AZ-GPTAA-MC MICRO PDP11 OWNER'S MANUAL (BA23) AZ-FI13A-MC EK-MIC11-OM MICRO PDP11 HARDWARE INFO ADDENDUM EK-O06AA-AD RQDX3 CONTROLLER MODULE USER'S GUIDE EK-RQDX3-UG RQDXE EXPANDER MODULE USER'S GUIDE EK-RQDXE-UG RQDX1 CONTROLLER MODULE USER'S GUIDE EK-RQDX1-UG RQDX2 CONTROLLER MODULE USER'S GUIDE EK-RQDX2-UG RD51-D,-R FIXED DISK DRIVE SUBSYSTEM OWNER'S MANUAL EK-LEP02-OM RD52-D,-R FIXED DISK DRIVE SUBSYSTEM OWNER'S MANUAL EK-LEP04-OM RD53-D,-R FIXED DISK DRIVE SUBSYSTEM OWNER'S MANUAL EK-LEP06-OM RD54-D,-R FIXED DISK DRIVE SUBSYSTEM OWNER'S MANUAL EK-LEP07-OM **** THE LAST FOUR HAVE ERRORS AND ARE OF QUESTIONABLE VALUE TK50 TAPE DRIVE SUBSYSTEMS USER'S GUIDE EK-OTK50-UG TK50 TAPE DRIVE SUBSYSTEM TECHNICAL MANUAL EK-OTK50-TM KDF11-BA CPU MODULE USER'S GUIDE (11/23 QUAD HEIGHT) EK-KDFEB-UG KDJJ11-A CPU MODULE USER'S GUIDE (11/73 DUAL HEIGHT) EK-KDJ1A-UG ZJADX-GZ - XXDP DOC SET XXDP USER'S MANUAL AZ-GNJAA-MC PDP-11 MICROCOMPUTER INTERFACES HANDBOOK EB-23144 MICRO/PDP-11 HANDBOOK EB-24944 PDP-11 ARCHITECTURE HANDBOOK EB-23657 MICROSYSTEMS HANDBOOK EB-26085 HAVE FUN! ================================================================================ Note 6.31 HARDWARE HINTS AND KINKS (HHK) 31 of 62 EISNER::PROVOST 5 lines 12-JUN-1987 10:40 -< TA78 rev's coming? >- -------------------------------------------------------------------------------- TA78 fails on some tapes. HSC50 at software level v350. There will be a substancial set of revisions available about July to detune drives to make them more tolerant of errors on tapes that are marginal w.r.t. ANSI standard. tjp ================================================================================ Note 6.32 HARDWARE HINTS AND KINKS (HHK) 32 of 62 EISNER::KILLEEN "Jeff Killeen" 5 lines 13-JUL-1987 17:33 -< FYI >- -------------------------------------------------------------------------------- Digital is having a fire sale on RD51-D's. They are selling for $480 MLP. These drives can be converted into RD53-DA, RD53-RA, RD54-DA and RD54-RA disk drives by buying a clone RD53 and RD54. This will save you half the price over the DEC drives. To my knowledge this is the only way to do it. ================================================================================ Note 6.33 HARDWARE HINTS AND KINKS (HHK) 33 of 62 EISNER::KENNEDY "Terry Kennedy" 20 lines 13-JUL-1987 19:46 -< More FYI >- -------------------------------------------------------------------------------- >> Digital is having a fire sale on RD51-D's. They are selling for $480 MLP. >> These drives can be converted into RD53-DA, RD53-RA, RD54-DA and RD54-RA >> disk drives by buying a clone RD53 and RD54. This will save you half >> the price over the DEC drives. To my knowledge this is the only way to >> do it. Some notes about the above - first, they can/will not ship it via UPS because the box is *HUGE*. The will only ship it motor freight, which can run up some big $ rapidly if you're not near Maynard. Next, you'll probably need the RQDXE -AA expander to use the external box on a Micro PDP-11. This costs $250.00 MLP. Upon consideration of all this, you might be better off getting a third-party replacement controller that will support other drive geometries and won't re- quire the RQDXE expander. This is especially the case for those who are running RQDX1 or RQDX2 controllers. Lastly, if you want the correct nameplate (the strip that says |d|i|g|i|t|a|l| RD5x), you'll have to buy 10 and wait 3 months because of minimum order/special order considerations. I know this is silly, but I heard of a manager who insisted the extra $ be spent for a 'real' RD54-DA because it had the right nameplate. tmk ================================================================================ Note 6.34 HARDWARE HINTS AND KINKS (HHK) 34 of 62 EISNER::KILLEEN "Jeff Killeen" 9 lines 13-JUL-1987 19:54 -< I HAVE CUSTOMERS WHO WOULD LIKE THE RIGHT STRIP >- -------------------------------------------------------------------------------- >>> Lastly, if you want the correct nameplate (the strip that says >>> |d|i|g|i|t|a|l| RD5x), you'll have to buy 10 and wait 3 months >>> because of minimum order/special order considerations. Seriously Terry.... Do you have the part numbers for the RD53 and RD54 strips? You also do not need the strips for the -RA models since you throw the skins away! ================================================================================ Note 6.35 HARDWARE HINTS AND KINKS (HHK) 35 of 62 EISNER::KILLEEN "Jeff Killeen" 9 lines 13-JUL-1987 20:03 -< FUN WITH DEC FS >- -------------------------------------------------------------------------------- >>> Lastly, if you want the correct nameplate (the strip that says >>> |d|i|g|i|t|a|l| RD5x), you'll have to buy 10 and wait 3 months >>> because of minimum order/special order considerations. Want to drive a DEC FS guy nuts? Take a RX50-DA and replace the RX50-AA with two RX33-AA's. We had scraped the RX33's from 11/53's. It works fine, it is all DEC hardware, but there is no equivalent DEC part number. And of course it said RX50 on the front. I thoroughly enjoyed watching them trying to price it. ================================================================================ Note 6.36 HARDWARE HINTS AND KINKS (HHK) 36 of 62 EISNER::KENNEDY "Terry Kennedy" 13 lines 13-JUL-1987 20:22 -< Yes, we have the numbers >- -------------------------------------------------------------------------------- Well, it just so happens: Strip for RD51 - 74-29647-01 " " RD52 - 74-29647-02 In any sane society, the RD53 should be -03 and the RD54 -04, but you had better call and check. 1-603-884-5001 (DEC unlisted parts) You'll need to get a 'product demand quote' which takes about 30 days, and then send it back with a check prepaying the order. About 90 days from then, you might receive the parts. By the way, if you're upgrading a RD52, I'd love to steal your nameplate. terry ================================================================================ Note 6.37 HARDWARE HINTS AND KINKS (HHK) 37 of 62 EISNER::KILLEEN "Jeff Killeen" 4 lines 30-JUL-1987 08:51 -< DEQNA ECO >- -------------------------------------------------------------------------------- The current ECO level for the DEQNA is REV "E". Without this ECO the DEQNA will not correctly support PDP-11 4MB systems or DECOM's. There is another ECO in the works for the DEQNA. The FCO reference for REV "E" is DEQNA-R-001 or EQ-01418-01. ================================================================================ Note 6.39 HARDWARE HINTS AND KINKS (HHK) 39 of 62 EISNER::KENNEDY "Terry Kennedy" 20 lines 6-AUG-1987 03:13 -< Note 6.38, take 2... >- -------------------------------------------------------------------------------- < Note 6.38 by EISNER::KENNEDY "Terry Kennedy" > -< Would you believe Rev. J4? >- >> The current ECO level for the DEQNA is REV "E". Without this ECO the >> DEQNA will not correctly support PDP-11 4MB systems or DECOM's. There >> is another ECO in the works for the DEQNA. The FCO reference for REV "E" >> is DEQNA-R-001 or EQ-01418-01. I just received from DEC a new DEQNA which is Rev. H4. This is, however, out of rev for RSTS 9.4, which requires DEQNA's of Rev. J4 or later. This info is direct from the RSTS implementor. Bear in mind that this rev letter is the one stamped on the module handle, not the one on the packing list. Never trust the packing list! What used to be the second paragraph of my reply has bee removed as I was told by DEC that although the information was derived from a published document of Digital that I would be in violation of an agreement I have with them by mentioning the hardware product in question. tmk ================================================================================ Note 6.40 HARDWARE HINTS AND KINKS (HHK) 40 of 62 EISNER::KENNEDY "Terry Kennedy" 14 lines 6-AUG-1987 21:24 -< RQDX1/RQDX2 assistance request >- -------------------------------------------------------------------------------- I am having a problem with a system containing a board which Digital claims is a RQDX2. The board identification is M8639-YB. DEC's doc- umentation claims that a M8639 is a RQDX1. DEC's add-on and insulting center told me that what I needed was the FCO kit EQ-01361-01, also known as M8639-R-007. Now, this kit makes references to M8639-00 and M8639-YA modules, but no M8639-YB. The ROMS in the new kit are part #s 23-172E5-00 and 23-173E5-00. The ROMS on the board I have are 23-178E5-00 and 23-179E5-00. My questin is, am I taking a step backwards or forwards with this EQ kit? If it is the wrong kit, does anyone know what the right kit is? If it's any help, the rev stamped onto the handle is C1, and the board artwork is 50-15649-02 A1. ================================================================================ Note 6.41 HARDWARE HINTS AND KINKS (HHK) 41 of 62 EISNER::TABOR "Bill Tabor" 8 lines 6-AUG-1987 22:44 -< YA=RQDX1 YB=RQDX2 >- -------------------------------------------------------------------------------- < I am having a problem with a system containing a board which Digital < claims is a RQDX2. The board identification is M8639-YB. DEC's doc- < umentation claims that a M8639 is a RQDX1. It is my understanding that the M8639-YA is the RQDX1 and the M8639-YB is the RQDX2. The FCO will change the YA to YB. ================================================================================ Note 6.42 HARDWARE HINTS AND KINKS (HHK) 42 of 62 EISNER::KILLEEN "Jeff Killeen" 18 lines 6-AUG-1987 22:50 -< CALL SCOTT & TOM >- -------------------------------------------------------------------------------- >>> I am having a problem with a system containing a board which Digital >>> claims is a RQDX2. The board identification is M8639-YB. DEC's doc- >>> umentation claims that a M8639 is a RQDX1. Each time DEC added a new drive they changed the frimware on the RQDX1/2 controller. I believe the firmware works on both the RQDX1 and RQDX2 controllers. I would strongly suggest you contact Scott Deno or Tom Trimarchi at Trimarchi and Associates (814) 234-5659. They know exactly what firmware goes with what controller and drive. They can supply you faster and more accurately than DEC. They stock the firmware. Now to your question M8639 (RQDX1) M8639-YB (RQDX2) M8639-YA (RQDX1) ================================================================================ Note 6.43 HARDWARE HINTS AND KINKS (HHK) 43 of 62 EISNER::KENNEDY "Terry Kennedy" 19 lines 7-AUG-1987 21:31 -< RQDX story - part 2 >- -------------------------------------------------------------------------------- Well, after hashing out the whole RQDX deal, I've determined the following: 1) The EQ kit is not for a RQDX2 - it is for RQDX1 only 2) It is not possible to upgrade a RQDX1 to a RQDX2 - you can get support for some of the newer drives, but you'll still have an RQDX1 3) My RQDX is an RQDX2 and it *is* broken Given all of the above, and the fact that DECmailer wanted $650 to fix the board and that DEC Field Service wanted $1400, I decided to splurge and get an RQDX3 (faster, better, etc). With the discount level my organization has, this came out to $1521.50. Now, you may ask why I bought another DEC board when brand 'x' is faster, cheaper, or better in some other way. The answer is that the system this is from is heavily into the DEC way of running RD disk drives - RQDXE expander, cab kit, external RD52-D, etc. and it simply was too hard & expensive to replace all of that with some other vendor's 'simple' cabling. I hope this helps anyone else who is having the same problem. tmk ================================================================================ Note 6.44 HARDWARE HINTS AND KINKS (HHK) 44 of 62 EISNER::KILLEEN "Jeff Killeen" 14 lines 7-AUG-1987 22:23 -< DOES IT WORK >- -------------------------------------------------------------------------------- >>> I decided to splurge and get an RQDX3 (faster, better, etc). Don't forget you have to hardware reformat your existing drives. >>> The answer is that the system this is from is heavily into the DEC way >>> of running RD disk drives - RQDXE expander, cab kit, external RD52-D, etc. >>> and it simply was too hard & expensive to replace all of that with some >>> other vendor's 'simple' cabling. RQDXE? - I thought the RQDX01/02 only worked with RQDX1-E and the RQDX03 only worked with RQDXE. The manual says you can't use RQDXE with the RQDX01/02 and the RQDX1-E can't be used with then RQDX03. Does it work? ================================================================================ Note 6.45 HARDWARE HINTS AND KINKS (HHK) 45 of 62 EISNER::KENNEDY "Terry Kennedy" 23 lines 8-AUG-1987 00:54 -< Does it work? - I hope so >- -------------------------------------------------------------------------------- >> Don't forget you have to hardware reformat your existing drives. The symptom of my controller problem is that the format was being wiped and I have to reformat weekly to remove the 1,000's of bad blocks which develop, do this isn't a problem. >> RQDXE? - I thought the RQDX01/02 only worked with RQDX1-E and the RQDX03 >> only worked with RQDXE. The manual says you can't use RQDXE with the >> RQDX01/02 and the RQDX1-E can't be used with then RQDX03. Does it work? From EK-RQDXE-UG-002: "DESCRIPTION This guide shows you how to install and configure the RQDXE dual-height expander module. The RQDXE (Digital part number M7513) is a replacement for the RQDX1-E extender module (Digital part number M7512). The RQDXE lets you expand the RQDXn (RQDX1, RQDX2 and RQDX3) controller modules to support external drives..." I suspect that your experience was with the older RQDXE unit. After all of the ECOing, I guess it now works properly (almost). ================================================================================ Note 6.46 HARDWARE HINTS AND KINKS (HHK) 46 of 62 EISNER::KILLEEN "Jeff Killeen" 10 lines 8-AUG-1987 08:10 -< MORE INFO >- -------------------------------------------------------------------------------- !>> Don't forget you have to hardware reformat your existing drives. ! !The symptom of my controller problem is that the format was being wiped !and I have to reformat weekly to remove the 1,000's of bad blocks which !develop, do this isn't a problem. NO - What I meant was when you go from a RQDX01/02 to a RQDX03 you have to hardware reformat all the disks. ================================================================================ Note 6.47 HARDWARE HINTS AND KINKS (HHK) 47 of 62 EISNER::KENNEDY "Terry Kennedy" 13 lines 8-AUG-1987 17:07 -< reformat vs. reformat >- -------------------------------------------------------------------------------- !The symptom of my controller problem is that the format was being wiped !and I have to reformat weekly to remove the 1,000's of bad blocks which !develop, do this isn't a problem. > NO - What I meant was when you go from a RQDX01/02 to a RQDX03 you have > to hardware reformat all the disks. When I said I have to reformat weekly I mean getting out XXDP and running the ZRQB?? RDnn format program. I know that for the RQDX3 the format util- ity is ZRQC??, but that isn't very different from what I am doing weekly. Or am I missing the point again? tmk ================================================================================ Note 6.48 HARDWARE HINTS AND KINKS (HHK) 48 of 62 EISNER::KENNEDY "Terry Kennedy" 15 lines 1-SEP-1987 23:15 -< RQDX3 report >- -------------------------------------------------------------------------------- I've just installed the RQDX3 upgrade and thought some of my experiences might be helpful to others: 1) The RQDX3-M upgrade kit does *not* include a manual. You should order manual EK-RQDX3-UG as well. 2) If using an RQDXE extender, the cable supplied with the extender is too short to install the board side-by-side. You will have to install them in adjacent slots. 3) The minimum revision of the formatter is C0 (as in ZRQCC0). Earlier revs will not handle the second-source RD52, RD54, RX33, or RD31 drives. The formatter will still say 'pass 1 aborted for this drive', but you can ignore it. The performance is much better (on Micro PDP-11/23) and my problems with spurious disk errors seem to have gone away. ================================================================================ Note 6.49 HARDWARE HINTS AND KINKS (HHK) 49 of 62 EISNER::KENNEDY "Terry Kennedy" 13 lines 3-SEP-1987 20:27 -< RQDXn in non-BA23/123 box? >- -------------------------------------------------------------------------------- Now that I have a surplus RQDX1 *and* a surplus RQDX2, I thought I'd install one of them in one of my other Q-Bus systems. Surprise! - DEC doesn't support them in anything but the BA23/BA123 boxes. Surely some- one out there must know how to cable the RQDX to a drive without going through the BA23 distribution panel, ready switch, etc. I tried ordering the humorously-named 'BA23 Field Maintenance Print Set' - MP-01651-00 - 54 bucks gets you four pages and no information. Boo, hiss! If anyone has any ideas, I'd appreciate it... Terry Kennedy ================================================================================ Note 6.50 HARDWARE HINTS AND KINKS (HHK) 50 of 62 EISNER::TABOR "Bill Tabor" 2 lines 3-SEP-1987 21:06 -< Try External Drive Cable Kit >- -------------------------------------------------------------------------------- You should be able to connect it by using an external (table top) box cable. ================================================================================ Note 6.51 HARDWARE HINTS AND KINKS (HHK) 51 of 62 EISNER::KILLEEN "Jeff Killeen" 14 lines 3-SEP-1987 21:19 -< RQDX? AND BA-11'S >- -------------------------------------------------------------------------------- > Now that I have a surplus RQDX1 *and* a surplus RQDX2, I thought I'd > install one of them in one of my other Q-Bus systems. Surprise! - DEC The only legal way to use these controllers in a BA-11 box is with a RQDX1-E or RQDXE extender board. Of course you must have RDxx-RA or RDxx-DA disk drive. This is supported config. Trimarchi in State College P.A. has built cables that go directly from the RQDX? controllers to the drives. Give Tom or Scott a call at 814-234-5659. You could build the cables yourself by getting a copy of the RQDX1 or RQDX2 controller manual and a copy of the drive interface standard. ================================================================================ Note 6.52 HARDWARE HINTS AND KINKS (HHK) 52 of 62 EISNER::KENNEDY "Terry Kennedy" 11 lines 5-SEP-1987 00:55 -< Street legal? >- -------------------------------------------------------------------------------- > The only legal way to use these controllers in a BA-11 box is with a > RQDX1-E or RQDXE extender board. Of course you must have RDxx-RA or > RDxx-DA disk drive. This is supported config. Well, I don't quite think that this system would be legal anyway - it is an old 16-bit box I've converted to 22-bit, and I've redesigned the power supply to give me more current. I will try calling the people you suggest. It *should* be possible to figure out the cables, but DEC isn't being very helpful... ================================================================================ Note 6.53 HARDWARE HINTS AND KINKS (HHK) 53 of 62 EISNER::KOZAM 24 lines 3-FEB-1988 21:47 -< No Sweat! >- -------------------------------------------------------------------------------- If you're the type of person who gets beads of sweat (yes, perspiration) on their forehead as you frantically try to get something to work, then you've probably experienced this phenomenon. Yesterday, one such bead of sweat fell into a VT125 through one of the air vents. This made the graphics board somewhat unhappy. Realizing what had happened, I used a trick I learned when working in a hospital biomedical lab (where all sorts of bodily fluids find their way to circuit components). I pulled the graphics board out, let it cool, then somewhat liberally poured 70% isopropyl alcohol (rubbing alcohol) over the area of the board most likely to have been sweated on. After fanning the board to evaporate the residual alcohol, I plugged it back in, and everything was fine! Isopropyl alcohol is my solvent of choice, since it is cheap and fairly harmless. It seems to be able to dissolve the salts in sweat, yet evaporates quickly, leaving little residue. This method is also good for the Coke/Pepsi/Coffee-in-the-keyboard syndrome. Hope this helps someone. Marc Kozam ================================================================================ Note 6.54 HARDWARE HINTS AND KINKS (HHK) 54 of 62 EISNER::ERSKINE "Chris Erskine" 7 lines 4-FEB-1988 07:26 -< Prefered Solvent >- -------------------------------------------------------------------------------- < Isopropyl alcohol is my solvent of choice, since it is cheap < and fairly harmless. It seems to be able to dissolve the salts in < sweat, yet evaporates quickly, leaving little residue. I prefer Vodka since if it does not work, you can use the rest of the bottle and not care about the hardware. ================================================================================ Note 6.55 HARDWARE HINTS AND KINKS (HHK) 55 of 62 EISNER::GARDNER "Tim Gardner" 20 lines 4-FEB-1988 08:18 -< I like my vodka a little stronger >- -------------------------------------------------------------------------------- Re: < Note 6.54 by EISNER::ERSKINE "Chris Erskine" > -< Prefered Solvent >- >< Isopropyl alcohol is my solvent of choice, since it is cheap >< and fairly harmless. It seems to be able to dissolve the salts in >< sweat, yet evaporates quickly, leaving little residue.. > > I prefer Vodka since if it does not work, you can use the rest of > the bottle and not care about the hardware. Personally, I prefer reagent grade ethanol. Unfortunately, its pretty hard to come by these days. I have used a water bath followed by a liberal dousing of ethanol to clean a keyboard that got a Slice bath. The water washes away most of the sugar and the ethanol dissolves and removes the water. tg ================================================================================ Note 6.56 HARDWARE HINTS AND KINKS (HHK) 56 of 62 EISNER::HAHN 5 lines 4-FEB-1988 10:41 -< a slightly different mixture >- -------------------------------------------------------------------------------- I go along with the .-1 note, I even add detergent to warm water in case the device has had a chance to dry with the coffee, cream and sugar. It works fine for keyboards. Pierre ================================================================================ Note 6.57 HARDWARE HINTS AND KINKS (HHK) 57 of 62 EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" 12 lines 4-FEB-1988 13:47 -< Try a little (hic!) ethanol... >- -------------------------------------------------------------------------------- When I used to work for the University of Texas Health Science Center at San Antonio, we used to use ethanol to clean any board we suspected of contamination. This was usually the case when we'd get a keyboard that was suffering from kkkkkeybouncccce. We'd just pull out the board & dip it (not whip it as in DEVO). The ethanol was great in disolving hair, skin flakes, soft drinks, (this is getting gross). No after effects to the board, either. Bob ================================================================================ Note 6.58 HARDWARE HINTS AND KINKS (HHK) 58 of 62 EISNER::KENNEDY "Terry Kennedy" 4 lines 4-FEB-1988 19:13 -< Try this... >- -------------------------------------------------------------------------------- For a nice clean keyboard, try adding some Polident denture cleaner to the water - this will clean the keycaps nicely. One place I used to work cad a competition for the grodiest keyboard - whoever won got to clean *everybody's* keyboards. ================================================================================ Note 6.59 HARDWARE HINTS AND KINKS (HHK) 59 of 62 EISNER::MERRIMAN "George Merriman - CCA/NY" 2 lines 5-FEB-1988 13:02 -< A warning >- -------------------------------------------------------------------------------- Be careful of cleaners containing TCE (Trichloroethane, I think). TCE disolves LK201's. ================================================================================ Note 6.60 HARDWARE HINTS AND KINKS (HHK) 60 of 62 EISNER::KILGALLEN "Larry Kilgallen" 5 lines 6-FEB-1988 12:55 -< What a great idea !!! >- -------------------------------------------------------------------------------- > Be careful of cleaners containing TCE (Trichloroethane, I think). > TCE disolves LK201's. Larry Kilgallen TECO fan ================================================================================ Note 6.61 HARDWARE HINTS AND KINKS (HHK) 61 of 62 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 8 lines 1-JUL-1988 11:13 -< try high test alcohol >- -------------------------------------------------------------------------------- When we use rubbing alcohol for keyboards we just grab the stuff we keep for tape heads. It is the 92 or 93% (claims not to rust your needles, if anyone cares) rather than the 70% regular stuff, and it dries faster as it has less water in it. Most drug stores have it, but some discount stores (like the 2 near our office) may not bother to stock it as it is a lower sales volume item. ================================================================================ Note 6.62 HARDWARE HINTS AND KINKS (HHK) 62 of 62 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 19 lines 1-JUL-1988 23:15 -< For the fanatics... >- -------------------------------------------------------------------------------- >> -< try high test alcohol >- >> Most drug stores have it, but some discount stores (like the 2 near >> our office) may not bother to stock it as it is a lower sales >> volume item. Most major cities have chemical supply houses where you can get the good stuff. Here in Glendale (Los Angeles), I can get 99% Isopropyl or 95% Ethanol/5% Isopropyl. The price is about $5 per 16 oz bottle. If you can't find a local source, try mail order from: Tri-Ess Sciences, Inc. 622 W. Colorado Street (818) 247-6910 Glendale, CA 91204 (213) 245-7685 P.S. It's also a good place to get cotton swabs on 6" sticks for getting into small places. Alan  ================================================================================ Note 7.0 Information on Old Hardware 8 replies EISNER::MAXWELL "Gary Maxwell" 14 lines 25-MAR-1987 19:14 -------------------------------------------------------------------------------- I hope I am not being too presumptuous in writing a note asking for information on ancient hardware that some people are still using somewhere in the Universe. ( This note has been moved from the NEW_CONFERENCE_IDEAS conference. ) Does anyone know about a "PDP Maxima" computer system that uses the CSP Language? They apparently exist in Europe. Does anyone have documentation for a DKC11-AB (a W9512 module)? This is an old bus foundation module. I have heard from Bruce Mitchell that there is nothing more written about it other than an overview document. I guess this was a piece of WYSIWYG hardware. ================================================================================ Note 7.1 Information on Old Hardware 1 of 8 EISNER::FRISBIE "Alan E. Frisbie" 10 lines 27-MAR-1987 01:05 -< Need print set for RX02 >- -------------------------------------------------------------------------------- If anyone has a documentation set for an RX02 and RXV21 controller, I would like to borrow it for a few days. The manuals that I need most are: MP-00629-00 RX02 Print set EK-ORX02-TM RX02 Technical Manual EK-ORX02-IP RX02 Illustrated Parts Breakdown (IPB) EK-13077-IP 70-13077 Flexible Disk Drive IPB Alan ================================================================================ Note 7.2 Information on Old Hardware 2 of 8 EISNER::PROVOST 6 lines 27-MAR-1987 13:42 -< FISCHY PRINTES >- -------------------------------------------------------------------------------- <<< If anyone has a documentation set for an RX02 and RXV21 <<< controller, I would like to borrow it for a few days. <<< The manuals that I need most are: I have the ufiche for RX02 and RXV11, but you can probably get something hard copy and up-to-date. Else, call. ================================================================================ Note 7.3 Information on Old Hardware 3 of 8 EISNER::FRISBIE "Alan E. Frisbie" 29 lines 28-MAR-1987 01:31 -< Answer for DCK11-AB question >- -------------------------------------------------------------------------------- Gary, The W9512 is simply the bare dual-width Q-bus wire-wrap prototype module. The data sheet I have is dated March 1978 and shows a photograph, an outline drawing and a listing of the VCC and GND pins. All this should be obvious if you have the board, but I can send you a copy if you need it. The DCK11-AB (not DKC), on the other hand, is a W9512 module plus the following Q-bus interface chips: DC003 (1) Interrupt logic DC004 (1) Protocol logic DC005 (4) Bus Transceivers DC006 (2) Word Count / Bus Address DC010 (1) Bus Master logic This information is from Fall 1978 copy of Digital Catalog Sales, pages 58-61. I can send you a copy of this if you wish. I also have (somewhere under the many stacks of paper in my office) a copy of the Chipkit User's Guide for the above chips. I can send you a copy of this 16-page guide, plus a later micro-note on extending the DMA logic from 16- to 18-bit addressing (remember, this was back in the dark ages of the Q-bus). If any of these documents would help you, let me know. Alan ================================================================================ Note 7.4 Information on Old Hardware 4 of 8 EISNER::HAHN 2 lines 8-JUN-1988 23:16 -< RX02 belts >- -------------------------------------------------------------------------------- Need part number for RX02 belt, my drive failed; the belt was laying in the bottom of the tray! Where do I get a few? Pierre ================================================================================ Note 7.5 Information on Old Hardware 5 of 8 EISNER::MAYHEW "Bill Mayhew" 8 lines 8-JUN-1988 23:34 -< RX02 belt part number and price >- -------------------------------------------------------------------------------- < Note 7.4 by EISNER::HAHN > > Need part number for RX02 belt, my drive failed; the belt was laying > in the bottom of the tray! Where do I get a few? Pierre According to my DEC Self-Maintenance Services book, part number 12-14005-00 is $7, assuming you have a 60Hz drive. Otherwise, you want the -01 variant, same price. Should be orderable through DECdirect. ================================================================================ Note 7.6 Information on Old Hardware 6 of 8 EISNER::PROVOST 15 lines 30-NOV-1988 13:44 -< NEED PRO350 HARDWARE HELP >- -------------------------------------------------------------------------------- Given: 1 PRO-350 with 512K memory. Needed: More memory without sacraficing slots. I am told this can be done by changing chips on the mother board. There was an article on how to do this in a past SIG Newsletter. Can anyone point me more directly at the article? Alternatively, can you tell me what must be done? Thanks, Tom ================================================================================ Note 7.7 Information on Old Hardware 7 of 8 EISNER::HAHN "DECUServe Vice Chair" 1 line 30-NOV-1988 17:10 -< move to PRO_300 note 31.* >- -------------------------------------------------------------------------------- For an answer to the above check with PRO_300 conference 31.* ================================================================================ Note 7.8 Information on Old Hardware 8 of 8 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 6 lines 23-AUG-1989 12:05 -< Found RX02 print set after two years >- -------------------------------------------------------------------------------- > The manuals that I need most are: > MP-00629-00 RX02 Print set Well, I finally found most of it (all of the boards for both RX01 and RX02 drives and Unibus controllers). If anyone else needs it, you know where to find it now.  ================================================================================ Note 8.0 Hardware Look-a-loos 11 replies EISNER::EISNER "Dan L. Eisner" 1 line 26-MAR-1987 02:01 -------------------------------------------------------------------------------- I am looking for a TU-56 and controller for a PDP-11. Dual DEC-tape. ================================================================================ Note 8.1 Hardware Look-a-loos 1 of 11 EISNER::PROVOST 8 lines 26-MAR-1987 09:41 -< Here's one, under my desk >- -------------------------------------------------------------------------------- <<< I am looking for a TU-56 and controller for a PDP-11. Dual DEC-tape. I have one I'm not using. We stole the power controller and used the rack, but the electronics should still work. It is officially owned by the U.S. Dept. of Energy, but I think we may have declared it scrap. In any case we can scrap it and maybe get it to you. We might be able to liberate the power controller, but the rack is pretty and useful. Lets follow up by phone and mail. ================================================================================ Note 8.2 Hardware Look-a-loos 2 of 11 EISNER::KILLEEN 2 lines 17-JUN-1987 19:06 -< RX50-DA, RX50-RA >- -------------------------------------------------------------------------------- I am looking for a RX50-DA or RX50-RA unit. You keep the RX50 all I need is the mounting unit part. ================================================================================ Note 8.3 Hardware Look-a-loos 3 of 11 EISNER::RICE "Gary Rice" 7 lines 18-JUN-1987 10:24 -< Parts is parts >- -------------------------------------------------------------------------------- > You keep the RX50 all I need is the mounting unit part. If you are referring to the "skid plate" that bolts onto the bottom of the drive, I have several of them. Call me if you want to deal. Gary (714)952-6582 ================================================================================ Note 8.4 Hardware Look-a-loos 4 of 11 EISNER::KILLEEN 11 lines 18-JUN-1987 11:51 -< LOOKING FOR LEPRECHANS >- -------------------------------------------------------------------------------- > You keep the RX50 all I need is the mounting unit part. >>>If you are referring to the "skid plate" that bolts onto the bottom >>>of the drive, I have several of them. Call me if you want to deal. No I am looking for the external mounting and power supply. Or to put it another way I am looking for a leprechan box with a RX50 distribution panel. ================================================================================ Note 8.5 Hardware Look-a-loos 5 of 11 EISNER::KILLEEN 2 lines 22-JUN-1987 21:02 -< FREE - YOU PAY SHIPPING >- -------------------------------------------------------------------------------- We are about to scrap some RK06s and RK07s, plus some RK611 and RK711 controllers. Is anyone interested for free. ================================================================================ Note 8.6 Hardware Look-a-loos 6 of 11 EISNER::KILLEEN 3 lines 24-JUN-1987 15:06 -< Rx-DA or Rx-RA >- -------------------------------------------------------------------------------- I am also looking for a RD50-DA, RD50-RA, RD51-DA, or RD51-RA. Again you can keep the drive if you want. I am just looking for the box. ================================================================================ Note 8.7 Hardware Look-a-loos 7 of 11 EISNER::KENNEDY "Terry Kennedy" 11 lines 30-JUN-1987 18:55 -< We'd love them >- -------------------------------------------------------------------------------- >> We are about to scrap dome RK06s and RK07s, plus some RK611 and >> RK711 controllers. Is anyone interested for free. Yes - We love 'em (and we're a charitable organization, too...) Please call: Terry Kennedy St. Peter's College Jersey City, NJ (201) 435-0252 ================================================================================ Note 8.9 Hardware Look-a-loos 9 of 11 EISNER::HOUGHTON "John N Houghton - CompuServe" 21 lines 1-JUL-1987 18:15 -< RD52 Leprechaun box - minus a few parts >- -------------------------------------------------------------------------------- > I am also looking for a RD50-DA, RD50-RA, RD51-DA, or RD51-RA. > Again you can keep the drive if you want. I am just looking for > the box. I've got an empty RD52 leprechaun box right here in my grubby paws, but it's missing some pieces - the RD52 was moved into the cab of a different system. What's missing is the online/write protect switch assembly (livable - the drives default to online writeable when the switch is missing), and whatever connectors plugged into the back holes marked J1 J2 and J3. I have no idea why these got taken, because an internally mounted drive would use ribbon cables, not shielded round cables with moby connectors on it. If you've a friendly Field Circus rep to help you figure which parts you'll need to get this up to spec, I can send it off to you for shipping cost. John Houghton CompuServe Data Technologies (formerly Software House) 1105 Massachusetts Ave Cambridge MA 02138 (617)661-7620 ================================================================================ Note 8.10 Hardware Look-a-loos 10 of 11 EISNER::KENNEDY "Terry Kennedy" 6 lines 1-JUL-1987 23:15 -< Censorship? >- -------------------------------------------------------------------------------- RE: -< VAXstation conversion kit for sale >- Has censorship reared its ugly head? My copy of the VAX Notes UG says: "The SET NOTE/HIDDEN command lets the moderator hide a note" What's up? ================================================================================ Note 8.11 Hardware Look-a-loos 11 of 11 EISNER::KILLEEN 12 lines 1-JUL-1987 23:30 -< POLICY NOT CENSORSHIP >- -------------------------------------------------------------------------------- The DECUServe commercialism policy is that there will be no overtly self serving commercial use of the DECUServe system. Selling something, even for one dollar, is commercial and overtly self serving. Note: As stated this is the least restrictive commercialism policy in DECUS. Having read the note I am sure the author thought he was doing the users of the system a favor. However, we have to draw a line no matter how noble the attempt. If you do not set a clear policy you get a lot of "well this case is just like that case" comments.  ================================================================================ Note 9.0 HARDWARE WISH LIST 12 replies EISNER::PROVOST 15 lines 26-MAR-1987 09:52 -------------------------------------------------------------------------------- Sometimes one needs a device which may or may not exist. This topic is for a wish list for additional functionality in a product, or for a product about which the requester may not be aware. We have need for emergency messages to come out of our public address system when any of 30 to 60 panic buttons are pushed. It must operate on UPS, be easily verified, and have cold standby (spares?) We have looked at boards which issue a digitized recording from PROMs. These are reliable but have no flexibility and would need outside prioritization and other logic. We have considered DECtalk on a Rainbow or clone, but it is expensive to start, and difficult to verify and spare. We'd also be inventing a wheel, with room for lots of mistakes. I'd like to find someone who has already addressed this problem successfully. tjp ================================================================================ Note 9.1 HARDWARE WISH LIST 1 of 12 EISNER::EISNER "Dan L. Eisner" 26 lines 26-MAR-1987 11:47 -< A method for the PA problem. >- -------------------------------------------------------------------------------- Some questions: How long does the text have to be (in time)? Would you be willing to make your own with some tech help? Do you have the labor pool to make it yourself? Do you need ASCII to sound? or will sound to sound work? (i.e. does it need to be like DECtalk? or will talk into and produce talk output work?) Most of the hardware should be available from your scrap heap. You need an A/D and a D/A with lots of memory. Some fortran code to control the data. A floppy disc or two. I worked on a project in the early 70's that used a SDC 930 with 16K om memory that sang. The vioce quality was based on sample interval which can be low. Data compression can be used if the data can be preprocessed. How about a DRV11-WA hocked to a RT System which is driven by a A/D and D/A. The memory requirements need to be looked at. There are some questions but I think it is due-able. 2nd thought, How about an IBM PC? seems to me that there are lots of cards available to do the job there. You will need a connection to the PA system, mic input and the relay logic to turn on the sound. The control leads of the DRV11 should do it. You could put a tone decoder on it and have it rigged up the receive phone call-ins. Many thoughts... ================================================================================ Note 9.2 HARDWARE WISH LIST 2 of 12 EISNER::HASSINGER "Bob Hassinger" 6 lines 7-MAY-1987 01:07 -< Why fancy digital? >- -------------------------------------------------------------------------------- Tom, Does it need to be digital? How about something simple - cassette tape recorders from Radio Shack, etc... Bob H ================================================================================ Note 9.3 HARDWARE WISH LIST 3 of 12 EISNER::PROVOST 7 lines 11-MAY-1987 13:42 -< TRIED IT! >- -------------------------------------------------------------------------------- <<< Does it need to be digital? How about something simple - cassette tape <<< recorders from Radio Shack, etc... We have used cassette tape recorders in the past. To assure immediate audible message, we keep the recoreders running at all times. The tapes wear out.I believe we now have about 36 recorders, and the messages usually don't work when tested. Our need is expanding, and we are looking for a better way. ================================================================================ Note 9.4 HARDWARE WISH LIST 4 of 12 EISNER::HASSINGER "Bob Hassinger" 8 lines 11-MAY-1987 14:49 -< How about an Amiga for voice output? >- -------------------------------------------------------------------------------- Okay Tom, if you can't get tape recorders to do what you want then the cheapest voice output I have come across is in the Amiga. 68000 CPU, floppy disk, synthetic voice, digitized audio, multitasking OS, etc, etc (and it is fun). I think the new model 500 may be priced under 1K for a package that should work for you - take a look. Bob H ================================================================================ Note 9.5 HARDWARE WISH LIST 5 of 12 EISNER::TINKELMAN "Bob Tinkelman" 23 lines 2-JUN-1987 11:05 -< Digital Pathways SLC-II >- -------------------------------------------------------------------------------- << We have need for emergency messages to come out of our >> << public address system when any of 30 to 60 panic buttons are >> << pushed. It must operate on UPS, be easily verified, and have >> I looked into (but never used) a product from Digital Pathways some years ago. Their model number is SLC-II. It's now fairly old technology as far as voice synthesis goes. (Vocabulary is limited to about 300 words.) However, it is programmable and, from their glossy brochures, seems designed to do the job you want. In summary, the SLC-II has the following ports: RS232 (4), RJ11, "external amplifier". The basic configuration seems to be to have the SLC-II "watch" an EIA line between a computer and a terminal, a line which is transparently routed through two of its EIA ports. It "traps" on certain conditions (things it sees on the line) and then takes actions. These usually involve synthesizing speach but can also (although not appropriate for your needs) include dialing a series of phone numbers to, say, wake up the system manager to tell him/her that the VAX is down. If you want to give them a call to see if they still make this product, etc., I have their phone number (or the one they had several years ago). It's 415-493-5544 (Mountain View, CA. ================================================================================ Note 9.6 HARDWARE WISH LIST 6 of 12 EISNER::FRISBIE "Alan E. Frisbie" 28 lines 4-JUN-1987 01:06 -< Another voice peripheral >- -------------------------------------------------------------------------------- << We have need for emergency messages to come out of our >> << public address system when any of 30 to 60 panic buttons are >> << pushed. It must operate on UPS, be easily verified, and have >> Tom, I don't know this will serve your needs, but I just got a flyer today from Western Peripherals for their DigiVoice DV-1000 Record/Playback Audio Digitizer: "The DigiVoice acts as a random access solid-state tape recorder, utilizing modified ADPCM speech algorithms to reproduce clear natural speech for voice response applications. Using on-board RAM or EPROMs, the DV-1000 stores user-recorded messages in digital format with user-selected numerical files." Specifications: Maximum number of files - 9999 40 seconds to 40 minutes of audio storage (solid-state) Instant random access Built-in sequencer to combine messages for longer announcements Built-in repeat and pause functions File upload and download to host computer Microphone or line audio input Serial or parallel interface etc. Let me know if you want more information. Alan ================================================================================ Note 9.7 HARDWARE WISH LIST 7 of 12 EISNER::PROVOST 14 lines 4-SEP-1987 16:00 -< voice => VMS file digitized >- -------------------------------------------------------------------------------- Friend has audio cassette containing 50 to 100 segments of voice, each segment less than 15 seconds, typical segment 7 seconds. Wants segment digitized and stored as VMS file on uVAX, so that he can mix voice with data and transmit on cable TV. Wants filter to less than 4KHz, digitize at 8KHz. Output should go to a magnetic tape. This kind of thing can be done cheaply on a PC, but the results are difficult to integrate with a VAX application. Anyone got one? Does anyone provide such a service? Tom ================================================================================ Note 9.8 HARDWARE WISH LIST 8 of 12 EISNER::HASSINGER "Bob Hassinger" 11 lines 4-SEP-1987 17:30 -< Sounds easy, what am I missing? >- -------------------------------------------------------------------------------- > -< voice => VMS file digitized >- I don't quite understand the constraints Tom. Description sounds like one time A/D. If so, do it on anything and transfer resulting files - should not be hard - something like KERMIT plus maybe a HEXify/deHEXify program pair. Output via almost any D/A on uVAX and maybe a simple low pass filter. If it is a Qbus system that should be easy and fairly cheap. Don't know what to do on a 2000. What am I missing in your need? Bob ================================================================================ Note 9.9 HARDWARE WISH LIST 9 of 12 EISNER::HAHN 8 lines 4-SEP-1987 20:28 -< Should not be too hard >- -------------------------------------------------------------------------------- As you indicated a PC is inexpensive, so digitze it there, then Kermit (file type binary) back to your VAX you then have your data. You can talk to anyone in the biomedical research field that does EMG or EKG digitizing. We use a PDP11 running RT to digitize, 8Khz of one channel is not too difficult. Talk to me off line if you wnat more information 516/444-1362. Pierre ================================================================================ Note 9.10 HARDWARE WISH LIST 10 of 12 EISNER::EISNER "Dan L. Eisner" 10 lines 8-SEP-1987 10:23 -< Voice to Tape solution. >- -------------------------------------------------------------------------------- > Wants filter to less than 4KHz, digitize at 8KHz. Output should go >to a magnetic tape. Tom, I would think an a/d convertor from one of the lab workstations around would do the job. You should be able to mod the XA driver to collect the data. Dan ================================================================================ Note 9.11 HARDWARE WISH LIST 11 of 12 EISNER::PROVOST 39 lines 9-JUN-1988 11:04 -< VME & ELN ? >- -------------------------------------------------------------------------------- Below is a request for information concerning a possible product. Since we at Bates Linear Accelerator Center might also be interested in such a product, I thought I'd pass the request on to you. Anyone have any rumors, contacts, information? The source is INFO-VAX, but since the author is clearly looking for information from any possible source, I am quoting his request without first consulting him. Tom Provost [:) (tech hat) =========================================================================== We are using VAX-ELN for a nuclear-physics data-acquisition system (based on Camac) and for a cyclotron control system. The advantage of VAX-ELN is an excellent software-development environment and a very good integration in our VAX/VMS environment. Disadvantages are that there is not very much (cheap) peripheral hardware (like ADCs, DACs) available for Q-bus, Unibus or BI-bus systems, and that parallel multi-processing is possible only with rather expensive BI-bus systems. Other institutes often use VME systems with M68000 processors in such cases, which has the advantages of having many vendors supplying peripheral hardware, and of a rather cheap solution for multi-processing. A disadvantage is usually a rather poor software developing environment, which weights heavily in our small computer group, because of shortage of manpower. I think that it would be a perfect solution if we could combine the advantages of both systems, which would be the case if DEC would produce a VAX processor board which could be placed in a VME crate and supported under ELN. A few days ago I heard a rumor that DEC is indeed developing something like that. Is there anybody who can confirm this rumor? Where could I get more information on this topic. My sales-representative does not know anything about it. F. Zwarts Phone: (+31)50-633619 Kernfysisch Versneller Instituut Bitnet/Earn: ZWARTS@HGRRUG51 Zernikelaan 25 Surfnet: KVIANA::ZWARTS 9747 AA Groningen Telefax: (+31)50-634003 The Netherlands Telex: 53410 rugro nl ================================================================================ Note 9.12 HARDWARE WISH LIST 12 of 12 EISNER::PROVOST 9 lines 6-JUN-1989 15:57 -< Schematic capture >- -------------------------------------------------------------------------------- We have about 100 schematic drawings done by hand. We have P-CAD and AUTOCAD on clones. We need a mechanism (hardware-software or service organization) which can scan hand drawn schematics and yield machine-readable information understandable to the PC software/hardware. Any hope? Tom Provost  ================================================================================ Note 10.0 gpx monitor cable extention 2 replies EISNER::CETRON 7 lines 26-MAR-1987 12:49 -------------------------------------------------------------------------------- I have found a way to move the monitor of the GPX type systems (mine is 8-bit color) away from the host microvax (mine is about 30 feet away.....) reply for more info - if there is interest, I will post the method -ed ================================================================================ Note 10.1 gpx monitor cable extention 1 of 2 EISNER::FRISBIE "Alan E. Frisbie" 6 lines 27-MAR-1987 00:17 -< Details, please >- -------------------------------------------------------------------------------- I am very interested. I only have the non-GPX version, but several of my clients would like to remote their GPX monitors. The problem that I see is not in the video signals, but in the mouse connection. Alan ================================================================================ Note 10.2 gpx monitor cable extention 2 of 2 EISNER::POOLE "Clyde T. Poole" 39 lines 16-MAY-1987 17:05 -< VAXStation and GPX monitor extension cables >- -------------------------------------------------------------------------------- The following note was moved by the moderator (ctp, EISNER::POOLE). ------------------------------------------------------------------- <<< EISNER::DUA0:[NOTES$LIBRARY]HARDWARE_HELP.NOTE;1 >>> -< HARDWARE_HELP >- ================================================================================ Note 16.0 VAXStation and GPX monitor extension cables No replies EISNER::CETRON 29 lines 16-MAY-1987 14:50 -------------------------------------------------------------------------------- Ok, heres the trick to extend your gpx cables. Since the cable already supplied is full of things, and impedances are everything, the idea is to use the cable and to build an extender for it. first, build a db-15 extension cable for the keyboard and mouse/tablet connections. Note, if you use twisted pair cables, try to assign the twisted pairs intelligently - like don't twist keyboard tx with keyboard rx - the pin outs are in the hardware manual shipped with the unit. This cable will then extend from the end of the dec supplied cable to the connector on the back of the mv or vs II box. DO NOT EXTEND THE VIDEO CONNECTIONS. To extend the video cables is slightly tougher. The first step is to build three coax cables (75 ohm - this is critical) as long as the db-15 cable with bnc plugs on each end. End one will plug into the monitor directly (not into the orignal cable) and the other end into similar receptacles to be built. Using small pieces of the 75-ohm coax, connect three bnc receptacles to the video connections within the db-15 connnector that already has the other digital connections (keyboard and tablet/mouse). plug both ends of the db-15 connections in (one side with only digital signals to the end of the original cables, the side with video AND digital to the connector on the back of the vax) the video already into the monitor and you are complete. c'est tout!  ================================================================================ Note 11.0 ECO/FCO's for common problems 22 replies EISNER::PROVOST 12 lines 1-APR-1987 10:17 -------------------------------------------------------------------------------- This TOPIC is intended to report ECO/FCO information for common problems. Some uVAX memory chips did not meet spec. Since there is memory on the CPU board as well as on the memory boards, an ECO is required to both. Symptoms involve MCHK #81,82,83. The following FCO's are mandatory, and involve a board-swap (no cost). uVAX II CPU Boards with module #'s "M... AH" require FCO EQ-01358-01 uVAX II 1 MB memory bds module #'s "M... BH" require FCO EQ-01358-02 uVAX II 2 MB memory bds module #'s "M... BH" require FCO EQ-01359-02 uVAX II 4 MB memory bds module #'s "M... BH" require FCO EQ-01359-02 ================================================================================ Note 11.1 ECO/FCO's for common problems 1 of 22 EISNER::FRISBIE "Alan E. Frisbie" 5 lines 2-APR-1987 00:13 -< No cost? >- -------------------------------------------------------------------------------- When you say that the FCO is mandatory and that there is no cost, does that include those systems that are not under a current maintenance agreement? Alan ================================================================================ Note 11.2 ECO/FCO's for common problems 2 of 22 EISNER::PROVOST 41 lines 2-APR-1987 09:17 -< FREE FCO??? >- -------------------------------------------------------------------------------- -< No cost? >- There are several categories of FCO, and I should look them up for reference, but here's a quick review as I remember them: The highest level of FCO is for safety factors. I believe Digital is required to provide it free to all purchasers of the product. Next is MANDATORY, which is free if you ask for it, and must be installed for contract customers. Most third party maintenance contracts also require installation. If you don't have a contract, you should still get the kit or boardswap free, but you must cover installation. Next level is for infrequent problems, and is required if symptoms appear. Again, no cost. Next level is for improved functionality or reliability, but problem doesn't bring device below design spec. This is free to contracts, but can be costly to others. Note the RA81 glue problem. Next is a level for which you must ask for the FCO, although it is free to contract customers when requested. Next is a level for which you may request the change, but you pay the bill event though you may be on contract. There are FCO's and ECO's. ECO's are Engineering Change Orders, and are installed in the manufacturing process. FCO's are Field Change Orders, and are installed in the field. Most people don't realize this distinction. Sometimes an ECO can have a corresponding FCO, sometimes not. ECO's usually change the REVISION LEVEL, although REV LEVEL can refer to Board Revision or to Etch Revision. If someone has a formal writeup on all this, he should enter it here. Otherwise I promise to run it down and enter a new reply. tjp ================================================================================ Note 11.3 ECO/FCO's for common problems 3 of 22 EISNER::PROVOST 29 lines 15-MAY-1987 12:28 -< KDJ11-A ECO >- -------------------------------------------------------------------------------- DEVICE KDJ11-A PROBLEM Bus contention problemÿ when DCJ11 exits from an FPJ11 provoked stall while a DMA bus transaction is current CORRECTION ECO # M8192-TW007 dated 7-OCT-1986 Rework boards and update documentation IDENTIFICATION Module at Rev Level D1 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE KDJ11-A PROBLEM Incorrect data results after some MULF and MULD instrictions which calculate a zero result and are followed by one of several FP instructions which use address moode zero CORRECTION ECO # M8192-TW009 dated 11-NOV-1986 new FPJ11 chip placed on module IDENTIFICATION New part number is 21-21858-05 Requires module etch rev level of C1 or greater ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ================================================================================ Note 11.4 ECO/FCO's for common problems 4 of 22 EISNER::PROVOST 24 lines 15-MAY-1987 12:42 -< KA630 ECO >- -------------------------------------------------------------------------------- DEVICE KA630 (M7606-AH) PROBLEM MicroVax II and VAXstation II running ULTRIX 32M V1.1 exhibit syystem crash bug. Error is logged on system console and in ultrix system message file machine check 80 CORRECTION ECO # M7606-ML006 dated 04-APR-1986 Replaced old memory chip Upgrade Kit EQ-01358-02 IDENTIFICATION Module number M7606-AS Module at etch rev level H2 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE KA630 (M7606-AH) M7606-ZS Reworked M7606-AH with NEC chips (P/N 21-21413-02) M7606-AS New build with NEC chips (P/N 21-21413-02) M7606-ZF Reworked M7606-AH with Hitachi chips (P/N 21-21414-02) M7606-ZC Reworked M7606-AH with Fujitsu chips (P/N 21-21415-02) M7606-ZAP New build with Mitsubishi chips (P/N 21-22422-02) M7606-ZP Reworked M7606-AH with Mitsubishi chips (P/N 21-22422-02) ================================================================================ Note 11.6 ECO/FCO's for common problems 6 of 22 EISNER::PROVOST 23 lines 15-MAY-1987 13:02 -< MS630-B ECO >- -------------------------------------------------------------------------------- DEVICE MS630-B (M7608-AH/BH) PROBLEM MicroVax II and VAXstation II running Ultrix 32M V1.1 exhibit system crash bug. Error is logged on system console and in Ultrix system message file Machine check 80 CORRECTION ECO # M7608-ML003 dated 04-APR-1986 Replaced old memory chip Upgrade Kit EQ-01359-01/02 IDENTIFICATION Module number M7608-AS Module at etch rev level B2 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE MS630-B (M7608-AH) M7608-AS/BS New build with NEC Chips (P/N 21-21413-02) M7608-AP/BP New build with Mitsubishi Chips (P/N 21-22422-02) M7608-AC/BC Reworked M7608-AH/BH with Fujitsu chips (P/N 21415-02) M7608-AF/BF Reworked M7608-AH/BH with Mitsubishi chips (P/N 21-22422-02) ================================================================================ Note 11.7 ECO/FCO's for common problems 7 of 22 EISNER::PROVOST 38 lines 15-MAY-1987 13:04 -< MS630-A ECO >- -------------------------------------------------------------------------------- DEVICE KA630 (M7606-AH) PROBLEM Low process yield for DC333 chip P/N 21-20887-01 circuit schematic typos CORRECTION ECO # M7606-ML007 dated 10-NOV-1986 New DC333 part number 21-20887-05 Upate ciruit schematic IDENTIFICATION Module at etch rev level H3 Document B-DD-M7606-0 at Rev j ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE MS630-A (M7607-AH) PROBLEM MicroVax II and VAXstation II running Ultrix 32M V1.1 exhibit system crash bug. Error is logged on system console and in Ultrix system message file Machine check 80 CORRECTION ECO # M7607-ML003 dated 26-MAR-1986 Replaced old memory chip Upgrade Kit EQ-01358-02 IDENTIFICATION Module number M7607-AS Module at etch rev level A3 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE MS630-A (M7607-AH) M7607-AC Reworked M7607-AH with Fujitsu chips (P/N 21415-02) M7607-AF Reworked M7607-AH with Hitachi chips (4/N 21415-02) M7607-AS New build with NEC Chips (P/N 21-21413-02) M7607-AP New build with Mitsubishi Chips (P/N 21-22422-02) ================================================================================ Note 11.8 ECO/FCO's for common problems 8 of 22 EISNER::PROVOST 42 lines 15-MAY-1987 13:19 -< RQDX3 ECO >- -------------------------------------------------------------------------------- DEVICE RQDX3 (M7555) PROBLEM Wrong part number on parts list - APGA - EPROMs CORRECTION ECO # M7555-ML002 dated 18-MAR-1986 change P/N 21-24144-01 to 21-21444-02 (E8) change P/N 23-166E5-00 to 23-216E5-00 (E21) change P/N 23-167E5-00 to 23-217e5-00 (E22) IDENTIFICATION Module at etch rev level C1 etch serial number 5916674-01-C1 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE RQDX3 (M7555) PROBLEM Internal (DEC) test jumpers needed IRQ detect logic not needed for new DPGA no support for RX33 and RD54 CORRECTION ECO # M7555-ML003 dated 09-JUN-1986 E21 install 23-243E5-00 microcode ROM E22 install 23-244E5-00 microcode ROM add jumpers for FF333 transfer IDENTIFICATION Module at etch rev level D1 etch serial number 5016674-01-D1 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE RQDX3 (M7555) PROBLEM Wrong revision written on documentation for the RQDX3 (B-DD-M7555-0) CORRECTION ECO # M7555-ML03A dated 09-JUN-1986 corrected typo created with ECO # M7555-ML003 IDENTIFICATION NONE ================================================================================ Note 11.9 ECO/FCO's for common problems 9 of 22 EISNER::PROVOST 14 lines 15-MAY-1987 14:13 -< RQDXE ECO >- -------------------------------------------------------------------------------- DEVICE RQDXE (M7513) PROBLEM When poser is lost due to a system shutdown or a power failure, the external RD drives can become corrupted. The result is seen as bad blocks, or loss of drive format. CORRECTION ECO # M7513-I001 dated 19-DEC-1986 Replace old module with a new M7513 Eq. Kit No. EQ-01456-01 IDENTIFICATION Module M7513 at etch rev level B1 E13 has P/N 74HC32 ================================================================================ Note 11.10 ECO/FCO's for common problems 10 of 22 EISNER::PROVOST 47 lines 15-MAY-1987 14:30 -< DEQNA ECO >- -------------------------------------------------------------------------------- DEVICE DEQNA (M7504) PROBLEM Data corruption and/or lock-up when DEQNA is busy with transmit pre-fill and receive message comes in, system reset is required. CORRECTION ECO # M7504-MK005 dated 10-AUG-1986 Replace M7504 with rev E1 module update documentation IDENTIFICATION Module at etch rev level E1 or greater E11 has P/N 23-087K4 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE DEQNA (M7504) PROBLEM Chip E4 (P/N 23-045L3) will occasionally enter into a race conditions causing corrupt data E4 socket is needed CORRECTION ECO # M7504-MK0ÿû5A dated 10-AUG-1986 Build a new chip add socket for E4 IDENTIFICATION Module at etch rev level E2 E4 has P/N 23-052L3-00 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE DEQNA (M7504) PROBLEM Rework module, too many wires module difficult and costly to build additional changes to documentation CORRECTION ECO # M7504-MK006 dated 6-SEP-1986 module at etch rev level F1 add changes to documentation IDENTIFICATION Verify that module has no wires at location E11 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ================================================================================ Note 11.11 ECO/FCO's for common problems 11 of 22 EISNER::PROVOST 50 lines 15-MAY-1987 14:41 -< TQK50 ECO >- -------------------------------------------------------------------------------- DEVICE TQK50 (M7546) PROBLEM Numerous solder shorts and solder defects. Board is difficult to rework. CORRECTION ECO # M7546-SH002 dated 04-APR-1986 Rework the board to remove 11 staking pins IDENTIFICATION Module at etch rev level B2 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE TQK50 (M7546) PROBLEM Update EPROMs with new microcode CORRECTION ECO # M7546-SH003 dated 17-JUN-1986 Replace E29 (P/N 23-170E5-00) with (P/N 23-180E5-00) Replace E28 (P/N 23-171E5-00) with (P/N 23-181E5-00) IDENTIFICATION Module at etch rev level C1 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE TQK50 (M7546) PROBLEM TK50 when running with an MSV11-J memory module will hang the system CORRECTION ECO # M7546-SH004 dated 25-SEP-1986 Replace FPLS Q-bus state machine with new version to fix the resetting of the 186 timer Replace E3 (P/N 23-051L1-00) with (P/N 23-055L1-00) IDENTIFICATION Module at etch rev level D1 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE TQK50 (M7546) PROBLEM MDM host command timeout error CORRECTION ECO # M7546-SH005 dated 2-FEB-1986 Update controller EPROMs Replace E29 (P/N 23-180E5-00) with (P/N 23-208E5-00) Replace E28 (P/N 23-181E5-00) with (P/N 23-209E5-00) IDENTIFICATION Module at etch rev level E1 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ================================================================================ Note 11.12 ECO/FCO's for common problems 12 of 22 EISNER::PROVOST 30 lines 15-MAY-1987 14:47 -< BA23 ECO >- -------------------------------------------------------------------------------- DEVICE BA23-A PROBLEM connector for D.D. power cable being derated by component eng. Amperage requirements of the unit exceeds this reduced rating. CORRECTION ECO # BA23-A-MK003 dated 26-AUG-1986 change to a cable that meets the required amperage rating IDENTIFICATION BA23-A at rev level C1 New power cable P/N 17-01311-01 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ DEVICE BA23-A PROBLEM Connector for disk power harness derated by component eng. and no longer meets power requirements. CORRECTION ECO # BA23-A-MK004 dated 23-FEB-1986 change to a power harness with a connector that meets thepower requirements. IDENTIFICATION BA23-A at rev level D1 New HARNESS P/N 17-01356-02 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ================================================================================ Note 11.13 ECO/FCO's for common problems 13 of 22 EISNER::PROVOST 4 lines 15-MAY-1987 14:59 -< ECO CREDITS >- -------------------------------------------------------------------------------- I would like to thank Jack Toto from Digital who gave a presentation at Nashville from which I lifted most of the ÿmaterial in the preceding ÿnotes. tjp ================================================================================ Note 11.14 ECO/FCO's for common problems 14 of 22 EISNER::PROVOST 24 lines 15-JUN-1988 12:28 -< BA23-C, BA23-D >- -------------------------------------------------------------------------------- ECO UPDATE DECUS 1988 BA23-C and BA23-D FCO number: BA23-1002 Problem: DC power cable 70-20450-02 burns during normal use in the BA23C/D box. The current rating of the components have been derated resulting in that the cable no longer meets our needs. Correction: It has been replaced with DC Power Cable 17-01312-01. Quick Check: Part number located on tag attached to the cable. FCO Kit Ordering: EQ-01467-01 FCO, Parts FA-04754-01 FCO only Thanks to O.Perez/J. Serbia ================================================================================ Note 11.15 ECO/FCO's for common problems 15 of 22 EISNER::PROVOST 24 lines 15-JUN-1988 12:28 -< BA123 >- -------------------------------------------------------------------------------- ECO UPDATE DECUS 1988 BA123 FCO number: BA123-1001 Problem: DC power cable 17-00865-01 overheats during normal use in the BA123 box. The current rating of the components have been derated and the cable no longer meets our needs. Correction: It has been replaced with DC Power Cables 17-01311-02. (Quantity 2) Quick Check: Part number located on tag attached to the cable. FCO Kit Ordering: EQ-01468-01 FCO, Parts FA-04755-01 FCO only Thanks to O.Perez/J. Serbia ================================================================================ Note 11.16 ECO/FCO's for common problems 16 of 22 EISNER::PROVOST 30 lines 15-JUN-1988 12:29 -< DELUA-M >- -------------------------------------------------------------------------------- ECO UPDATE DECUS 1988 DELUA-M FCO number: DELUA-1002 Problem: 1. DELUA in a 8300 receives a UNIBUS ACCESS TEST failure during power-up self-test and causes a system crash. 2. Power-up self-test failure of DMA UNIBUS ACCESS on DELUA when in a PDP 1184. 3. LAVC Clusters keep breaking connections due to host time-outs 4. Terminal servers disconnect due to host time-outs. Correction: Upgrade all DELUA-M at revision "E2", "E3", or "E4" exhibiting problems/symptoms as described above. Quick Check: Verify that E31 is P/N 23-451E4-00 and that E56 is P/N 23-452E4-00. FCO Kit Ordering: EQ-01501-01 FCO, Parts FA-04796-01 FCO only Thanks to O.Perez/J. Serbia ================================================================================ Note 11.17 ECO/FCO's for common problems 17 of 22 EISNER::PROVOST 23 lines 15-JUN-1988 12:29 -< RRD50 >- -------------------------------------------------------------------------------- ECO UPDATE DECUS 1988 RRD-50 FCO number: RRD50-R001 Problem: 1. Intermittent, undetected data compare errors. 2. Needs new microcode release to meet MSCP and DUP specs. Correction: Replace Q-Bus Controller (M7552 Rev C1) Quick Check: Look for DC7005C (P/N 21-23862-02) at E32 FCO Kit Ordering: EQ-01462-01 FCO, Parts FA-04748-01 FCO only Thanks to O.Perez/J. Serbia ================================================================================ Note 11.18 ECO/FCO's for common problems 18 of 22 EISNER::PROVOST 30 lines 15-JUN-1988 12:29 -< RA81X >- -------------------------------------------------------------------------------- ECO UPDATE DECUS 1988 RA81X FCO number: RA81X-R016 Problem: The Arrowhart L5-30P power plug has insufficient clamping force to supply the amount of current needed to support a full cabinet configuration. The insufficient clamping force with increased current causes contact resistance to rise, resulting in excess heat to develop. In time, this heat reaches a point of melting the Arrowhart L5-30P power plug. Correction: Replace 874-D Power Controller with Revision "D" or higher. Quick Check: RA81-CA/CE, DA/DE, EA/EE, Three High H9642 Cabinet contains power controller 874-D and is revision "D" or higher. FCO Kit Ordering: EQ-01473-01 FCO, Parts FA-04762-01 FCO only Thanks to O.Perez/J. Serbia ================================================================================ Note 11.19 ECO/FCO's for common problems 19 of 22 EISNER::PROVOST 28 lines 15-JUN-1988 12:30 -< RQDX2 >- -------------------------------------------------------------------------------- ECO UPDATE DECUS 1988 RQDX2 (M8639-YB) FCO number: RQDX2-R001 Problem: The accumulation of recoverable (soft) errors on RD53 drives. The development of uncorrectable (hard) errors on RD53 drives. The need to reformat RD53 drives every six weeks to three months. Correction: Replace the cable P/N 17000681-01 or BC05W-02 with the cable P/N 17-01520-01. Quick Check: The presence of P/n 17-01520-01 (shielded) cable connecting M8639-YB to M9058. FCO Kit Ordering: EQ-01489-01 FCO, Parts FA-04782-01 FCO only FA-04782-02 FCO only Thanks to O.Perez/J. Serbia ================================================================================ Note 11.20 ECO/FCO's for common problems 20 of 22 EISNER::PROVOST 29 lines 15-JUN-1988 12:30 -< RQDX3 >- -------------------------------------------------------------------------------- ECO UPDATE DECUS 1988 RQDX3 (M7555) FCO number: RQDX3-R001 Problem: The accumulation of recoverable (soft) errors on RD53 drives. The development of uncorrectable (hard) errors on RD53 drives. The need to reformat RD53 drives every six weeks to three months. Correction: Replace the cable P/N 17000681-01 or BC05W-02 with the cable P/N 17-01520-01. Quick Check: The presence of P/n 17-01520-01 (shielded) cable connecting M7555 to M9058. FCO Kit Ordering: EQ-01489-01 FCO, Parts FA-04782-01 FCO only FA-04782-02 FCO only Thanks to O.Perez/J. Serbia ================================================================================ Note 11.21 ECO/FCO's for common problems 21 of 22 EISNER::PROVOST 38 lines 15-JUN-1988 12:30 -< TQK50 >- -------------------------------------------------------------------------------- ECO UPDATE DECUS 1988 TQK50 (M7546) FCO number: TQK50-R002 Problem: Failure of the TK50/TQK50 Subsystem: 1. When performing Backup/Verify on RSX, RSTS/E, and DSM 11. 2. When recovering from an ABORT operation (CTL C). 3. High soft error rate on VAX 3500 and VAX 3600 systems. 4. Eliminates the Non-existent memory (NXM)/Parity errors seen with single word transfers. 5. When writing multiple small files using the PIP command (Positioning lost when searching for tape marks in reverse). 6. MDM timeout problems (MDM test 12). 7. False or unknown drive errors when soft read/write errors were reported in the error log. 8. Inability to read ANSI tapes that were created under RSX by a VMS system. Correction: Update M7546 microcode and module revision per FCO. Quick Check: Examine the error log for controller hardware REV 8 or REV 9 which contain the change. Also check module for REV J1 or greater. FCO Kit Ordering: EQ-01494-01 FCO, Parts FA-04788-01 FCO only ** Drive has to be Revision E1 or greater. Thanks to O.Perez/J. Serbia ================================================================================ Note 11.22 ECO/FCO's for common problems 22 of 22 EISNER::PROVOST 7 lines 15-JUN-1988 12:32 -< ECO Credits >- -------------------------------------------------------------------------------- The last 8 NOTES were made possible by a session in Cincinnati sponsored by the HMS SIG. The contents are taken from a handout prepared by O. Perez and J. Serbia of Digital. Thanks folks. Tom  ================================================================================ Note 12.0 Long access times on RD53/RQDX3 3 replies EISNER::FRISBIE "Alan E. Frisbie" 56 lines 11-APR-1987 03:12 -------------------------------------------------------------------------------- After I had my VAXstation II for a few months, I decided to learn how to use TPU. Imagine my distress when it took 58 seconds before the screen was first displayed! I called Telephone Support (Digital Dial-a-Prayer) to ask if this was normal and they didn't think so, but couldn't reproduce the problem (where have I heard that before?). To make a looooong story short, it turned out that certain blocks on the RD53/RQDX3 had very long access times. The READY light would go out completely for seconds at a time when trying to read a single block, as with DUMP. The Diagnostics and the Error Logger reported no problems at all. A simple-minded program to do a histogram of access times of an entire disk (read sequentially, one block at a time) produced the following: Time (sec) Quantity 0.00 - 0.01 33447 0.01 - 0.02 104376 0.02 - 0.03 754 0.03 - 0.04 13 . . (45 blocks 0.05 - 0.24 sec, . . reason not known, but seem . . to be "normal" variation) 0.23 - 0.24 4 15.48 - 15.49 1 (Note jump in time!!!) 15.50 - 15.51 1 15.51 - 15.52 3 . . (17 blocks 15.48 - 15.55 sec.) . . . . 23.30 - 23.31 1 (4 blocks 23.24 - 23.31 sec.) The program reported the LBNs of all long access times and they were 100% repeatable. Naturally, when the Field Circus guy came out, he couldn't understand what I was complaining about! The disk was reformatted, but had the exact same problem. Of course, when VMS was restored to the disk, the bad blocks were now in different files. TPU now came up fast, but Directory took 25 seconds! Fortunately, I had already contacted DEC Engineering and put Field Service on the phone with them. The drive and controller were replaced and the bad ones sent back for analysis. This was in mid-January and I have not heard what the results were. I have heard that this problem has been seen very rarely, but that it does show up. I shudder to think what a non-technical customer would have gone through to get this resolved. "The diagnostics don't show anything wrong!" Alan ================================================================================ Note 12.1 Long access times on RD53/RQDX3 1 of 3 EISNER::FRISBIE "Alan E. Frisbie" 12 lines 1-JUN-1987 13:53 -< Update (?) on RD53/RQDX3 problem >- -------------------------------------------------------------------------------- I just got a call from the Product Support Supervisor for the RD53 (and other buy-out disks). A marketing-type person had just told him that I had described this problem at the Spring Symposium and was looking for an answer. (Isn't modern instantaneous communications wonderful?) The P.S.S. collected all the data, names and dates from me and promised to get back in a few days with an answer. More information as I receive it... Alan ================================================================================ Note 12.2 Long access times on RD53/RQDX3 2 of 3 EISNER::WICHMAN "Jon Wichman" 5 lines 14-APR-1989 12:02 -< Is there a cure for this? >- -------------------------------------------------------------------------------- Anything further on this? I think I've seen the same thing on a system (which has other problems too - I'll be posting a note on the other problems shortly). Thanks! ================================================================================ Note 12.3 Long access times on RD53/RQDX3 3 of 3 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 9 lines 17-APR-1989 11:49 -< Another DEC black hole problem >- -------------------------------------------------------------------------------- RE: RQDX3/RD53 15-second access time problem >> Anything further on this? No, I never heard anything more from DEC about this. I suspect that any fixes were just bundled into later revisions of the firmware, but have no way of knowing. I have not seen the problem in the last two years and have never met anyone who suffers from it.  ================================================================================ Note 13.0 Backup Devices 41 replies EISNER::KASPER "Beverly T. Kasper" 5 lines 14-APR-1987 11:03 -------------------------------------------------------------------------------- This topic is intended as a place to discuss relative merits of, and known problems with, various backup devices. Obviously, vendor names will be mentioned; please remember to keep the content of your notes technical and non-commercial. ================================================================================ Note 13.1 Backup Devices 1 of 41 EISNER::KASPER "Beverly T. Kasper" 10 lines 14-APR-1987 11:09 -< Super drive? >- -------------------------------------------------------------------------------- The April issue of the Dec Professional has an ad (pp 100-101) for a drive which claims to fit 5.2 gigabytes on a standard T-120 cassette. It also claims a transfer rate of 4 megabytes per second. This sounds too good to be true. VHS backups have supposedly had very slow transfer rates up til now; I thought that was a constraint of the design of the media. Anyone out there know more about this? ================================================================================ Note 13.2 Backup Devices 2 of 41 EISNER::CETRON 5 lines 14-APR-1987 19:29 -< I think the Super drive missed the bus >- -------------------------------------------------------------------------------- sounds bad news to me..... the transfer rate seems way to high (and is quite useless since only the bi bus will go that fast) -ed ================================================================================ Note 13.3 Backup Devices 3 of 41 EISNER::RICE "Gary Rice" 27 lines 15-APR-1987 10:33 -< Hot off the press >- -------------------------------------------------------------------------------- > The April issue of the Dec Professional has an ad (pp 100-101) for > a drive which claims to fit 5.2 gigabytes on a standard T-120 > cassette. It also claims a transfer rate of 4 megabytes per second. > Anyone out there know more about this? I saw the same ad and called Honeywell. I talked to the chief engineer on the project. Here are some of his information about the drive: - YES, a transfer rate of 4 mb is there (in the same way that Ethernet is a 10 mb cable) - The tpae transport travels 4 times the speed of a VHS VCR to achieve that rate - Current VMS support is via a DMF32 synchronous printer port (he didn't know anything about software support though) - SCSI ("skuzzy") port like on the Apple Macintosh is also available now - Current development work is on an HSC interface - Next year (assuming no more legal battles with DEC) is a BI bus versionÿûýé - Price & availability were not his area of expertise. I expect to get a product description packet in the mail this week with that type of data. Will keep you informed Gary ================================================================================ Note 13.4 Backup Devices 4 of 41 EISNER::MAXWELL "Gary Maxwell" 11 lines 15-APR-1987 19:22 -< More poop >- -------------------------------------------------------------------------------- < - Price & availability were not his area of expertise. < I expect to get a product description packet in < the mail this week with that type of data. Will < keep you informed My latest issue of Mini-Microsystems said that evaluation units would available in July '87. They won't be cheap - $20,000 at OEM pricing. The major application is for huge data acquisition experiments (oil exploration, ocean-floor mapping), and that is the initial target market, although they may have a real hit on their hands. ================================================================================ Note 13.5 Backup Devices 5 of 41 EISNER::RICE "Gary Rice" 17 lines 17-APR-1987 10:39 -< Price update >- -------------------------------------------------------------------------------- > My latest issue of Mini-Microsystems said that evaluation units would > available in July '87. They won't be cheap - $20,000 at OEM pricing. For us end users, $20K seems like a bargain. The letter I got from the product manager said: "...volume production deliveries in July will come with a SCSI controller with a Unibus controller available in October. Prices depend on configuration, but the tape unit list price will start at $39,000 plus controller at $3,000-$5,000 more." Pretty "Pricey" if you ask me. Maybe I will look harder at optical disks. Gary ================================================================================ Note 13.6 Backup Devices 6 of 41 EISNER::MAXWELL "Gary Maxwell" 5 lines 17-APR-1987 17:07 -< Maybe we should leave out the $$$$$ >- -------------------------------------------------------------------------------- Maybe we should be more careful about quoting prices. My fault, I guess I started it.... Gary. ================================================================================ Note 13.7 Backup Devices 7 of 41 EISNER::RICE "Gary Rice" 7 lines 18-APR-1987 10:08 -< NIX on no prices >- -------------------------------------------------------------------------------- > Maybe we should be more careful about quoting prices. If I read my ethics discussions correctly, prices ARE allowed in the context we have just used them in. Arguments anyone? Gary ================================================================================ Note 13.8 Backup Devices 8 of 41 EISNER::NAGY "Frank J. Nagy, VAX Guru" 19 lines 23-AUG-1987 10:28 -< DigiData Gigastore system >- -------------------------------------------------------------------------------- Relative to the discussion on the pricey Honeywell Test Instruments VHS tape system: We were looking at an enhancement for BACKUP on our LAVC (currently 2 RA81s, going to 4) which currently uses TK50s (sigh). We saw the Honeywell ads and talked to them, quickly realizing that they were in a different ballpark than us. We came across ads (in several rags) for DigiData's Gigastore tape subsystem. This system also uses VHS tapes, some of the specifications are: - capacity of 2.5 Gbytes per T120 VHS cartridge - transfer rates of about 120KB/sec (i.e. 4xTK50) - error rates of 1 in 10**23 bits Yes, thats 1 in many, many bits. According to DigiData, the system has been optimized for backup and archival needs. Interfaces for both QBus and UNIBUS are available along with VMS drivers. I would like to say more, but our purchase order for a Gigastore system has floundered in the latest budget crisis here. If we ever get one, I'll add a report on our experience with it. ================================================================================ Note 13.9 Backup Devices 9 of 41 EISNER::NAGY "Frank J. Nagy, VAX Guru" 3 lines 23-AUG-1987 10:30 -< More on Gigastore >- -------------------------------------------------------------------------------- Forgot to say in my previous note: the Gigastore is considerably less pricey than the Honeywell system. Roughly twice the cost of a TK50 plus controller. (See, I did it without prices!) ================================================================================ Note 13.10 Backup Devices 10 of 41 EISNER::KILLEEN "Jeff Killeen" 18 lines 23-AUG-1987 17:12 -< PRICE IS OK >- -------------------------------------------------------------------------------- > Forgot to say in my previous note: the Gigastore is considerably less > pricey than the Honeywell system. Roughly twice the cost of a TK50 > plus controller. (See, I did it without prices!) SURPRISE! - You *CAN* use price on DECUServe. The policy is you can use price as long as you have no vested commercial interest in the product e.g. you own stock, manaufacture, service, or sell it. If you had a vested interest and even through you did not metion price you would have been questioned on this note and the last one. Jeff Killeen Executive Committee I would assume the price is about $7,500? ================================================================================ Note 13.11 Backup Devices 11 of 41 EISNER::NAGY "Frank J. Nagy, VAX Guru" 4 lines 24-AUG-1987 07:58 -< Gigastore $ >- -------------------------------------------------------------------------------- I think I remember the price being about $6,200 on the PR. My only vested interest in the Gigastore will (hopefully :-) be as a satisfied user. ================================================================================ Note 13.12 Backup Devices 12 of 41 EISNER::LEDERMAN "Bart Z. Lederman" 26 lines 25-AUG-1987 08:19 -< It's not just initial cost >- -------------------------------------------------------------------------------- We've been looking into the problem for some time. The entire field seems to be heating up. There is also a unit announced that uses the smaller 8mm video tapes (also in the Gigabyte range), and HP and Sony have announced a joint venture to put data on the DAT (Digital Audio Tape) format, for which provisions were made in the original specification. That would also be in the 3 Gigabyte range, and the cartidges are small enough to fit into portable and even laptop machines (talk about competition for the TK50!). I don't know if anyone besides Honeywell is ready to deliver now, though. Certanly the DAT tapes are not on the market yet (I don't know for certain if Honeywell has actually delivered yet either.) I have two reactions. First, with additional products on the market and competition, Honeywell's price may well come down (as it probably would anyway if it caught on and production ramped up if past experience is anything to go by). Second, we propose to put large amounts of information on which operation of the entire company depends onto tapes like these for storage. We would be willing to pay extra to get our devices from a company with a secure background which we can trust to be in business for the next 15 to 20 years, which is at least how long we would have to keep these things running. We can't bet the corporation on a startup company with no background and no known future. If I were buying a unit with my own money for home use I might feel differently. ================================================================================ Note 13.13 Backup Devices 13 of 41 EISNER::FRISBIE "Alan E. Frisbie" 15 lines 25-AUG-1987 15:49 -< DAT decks in three weeks >- -------------------------------------------------------------------------------- >> Certanly the DAT tapes are not on the market yet On Friday (21-Aug), Marantz announced that their DAT units will go on sale in about three weeks. The price I heard was in the $1200 range. I don't know how much additional hardware will be necessary to connect it to a computer. The price will probably remain high for quite a while. These units will NOT have the circuitry that prevents copying Compact Disks, so they will be in great demand. All this reminds me of early hobby computers that used cassette audio tapes for storage. Have we come full circle? :-) Alan ================================================================================ Note 13.14 Backup Devices 14 of 41 EISNER::DELARISCH 27 lines 25-AUG-1987 18:33 -< DAT w/o DES so you can COPY >- -------------------------------------------------------------------------------- >> < Note 13.13 by EISNER::FRISBIE "Alan E. Frisbie" > >> -< DAT decks in three weeks >- >> >> >> Certanly the DAT tapes are not on the market yet >> >> On Friday (21-Aug), Marantz announced that their DAT units >> will go on sale in about three weeks. The price I heard >> was in the $1200 range. I don't know how much additional >> hardware will be necessary to connect it to a computer. >> According to my information (My Brothers are in the Audio Trade) the initial DAT's will be selling in the $2000 range initially. >> The price will probably remain high for quite a while. >> These units will NOT have the circuitry that prevents copying >> Compact Disks, so they will be in great demand. Yup, the prices will be HIGH for a Loooooooong while! The unit will be shipped for Digital to Digital copying .... the copy prevention circuitry is currently unnecessary since CD players don't have Digital outputs (only Analog ones). More Latter .... as the situation digitizes! ;-}} -Arnold ================================================================================ Note 13.15 Backup Devices 15 of 41 EISNER::HANRAHAN "Jamie Hanrahan" 6 lines 26-AUG-1987 14:05 -< nitpicking >- -------------------------------------------------------------------------------- This should be moved to rec.audio on Usenet, but as a sometime audiophile I can't resist... > CD players don't have Digital outputs (only Analog ones). In fact, many of the better CD players have digital outputs. ================================================================================ Note 13.16 Backup Devices 16 of 41 EISNER::DELARISCH 15 lines 26-AUG-1987 17:26 -< Whom? >- -------------------------------------------------------------------------------- >> < Note 13.15 by EISNER::HANRAHAN "Jamie Hanrahan" > >> -< nitpicking >- >> > CD players don't have Digital outputs (only Analog ones). >> In fact, many of the better CD players have digital outputs. Really??? WHOM? -Arnold (Audiofool or is that Audiopile ;-} ) ================================================================================ Note 13.17 Backup Devices 17 of 41 EISNER::LEDERMAN "Bart Z. Lederman" 16 lines 26-AUG-1987 21:38 -< Music recorders won't help us much. >- -------------------------------------------------------------------------------- Though interesting (Music, and the proper reproduction thereof, is of great interest to me), I don't think it's really applicable. The format of data storage on a DAT tape is going to be different than the format for music (though compatible with the head / scanning scheme) just as the format for data storage on video tape is going to be different than for storing video. There has been, for several years, devices which turn a data stream into an RS-170 compatible signal that can be recorded on an ordinary video recorder, but they were never very popular and the devices I see on the market now have very much higher data rates (though you ought to look at the bandwidth of a video signal sometime: it's really amazing that you can buy a recorder now in New York for well under $200 that will record and play back video, but you can't get an ordinary 9 track tape drive for anything under several thousands of dollars). ================================================================================ Note 13.18 Backup Devices 18 of 41 EISNER::KENNEDY "Terry Kennedy" 14 lines 26-AUG-1987 23:11 -< More thoughts >- -------------------------------------------------------------------------------- > Who makes a CD player with digital output? Sony ES series (extra squirrely) has a digital coaxial connector on the back. It is one of the major useless extras in the series. [CDP-550ES] > ... it's really amazing that you can buy a recorder now in > New York for well under $200 that will record and play back video, but > you can't get an ordinary 9 track tape drive for anything under several > thousands of dollars). Ah - but the number of video recorders sold in one week doesn't equal the number of 9-track drives built in a year. Economies of scale (or lack thereof). Also, the native bit error rates on the two devices (before ex- ternal error correction) are quite different ================================================================================ Note 13.19 Backup Devices 19 of 41 EISNER::KILLEEN "Jeff Killeen" 10 lines 26-AUG-1987 23:31 -< THIS AND DAT >- -------------------------------------------------------------------------------- > Sony ES series (extra squirrely) has a digital coaxial connector on the > back. It is one of the major useless extras in the series. [CDP-550ES] According to Video Review the issue with DAT Tape is not CD to DAT recording. The synch tone on the digital connectors for DAT and CD's are at different MHZ. If you digitally record CD to DAT you get a hum in the copy. The issue is DAT to DAT. What is interesting is if the record companies never make DAT tapes there is no problem. Also you can get the same quality as DAT by using PCM with your VCR - and a lot less bucks. ================================================================================ Note 13.20 Backup Devices 20 of 41 EISNER::KILLEEN "Jeff Killeen" 22 lines 27-AUG-1987 23:24 -< ALMOST >- -------------------------------------------------------------------------------- > According to Video Review the issue with DAT Tape is not CD to DAT recording. > The synch tone on the digital connectors for DAT and CD's are at different > MHZ. If you digitally record CD to DAT you get a hum in the copy. The issue > is DAT to DAT. What is interesting is if the record companies never make > DAT tapes there is no problem. Also you can get the same quality as DAT > by using PCM with your VCR - and a lot less bucks. Well I was almost right - the danger of recalling from memory.... According to Video Review Aug 1987 "The digital sampling rates of CDs and DAT - the intervals which a recorder measures an audio signal - are different, making the two digital codes incompatible." You would need to build a digital to digital converter box. -------------------------------------------------------------------------------- Now that I have sinned with you all - audio discussions are not for DECUServe. ================================================================================ Note 13.21 Backup Devices 21 of 41 EISNER::MCPHERSON "Mike McPherson" 17 lines 9-SEP-1987 18:43 -< Gigastore data rates >- -------------------------------------------------------------------------------- < Reply to 13.8 > I had a chat with one of the folks from DigiData this afternoon, and he said something that got me thinking (no comments please). He said that the Gigastore drive streams data to tape at about 120Kb per second (we knew that already), but that if the computer driving it can't generate a data stream at that rate, the drive *inserts filler* to make up the difference! Is this a surprise to anyone but me? Since my 750 can't even keep a TU81 streaming, it doesn't have a prayer of streaming this guy! It seems to me that an incremental BACKUP of an RA81 is going to produce a multi-volume backup tape set, most of which contains *nothing*! If I have this all wrong, please let me know. I was hoping that this might be the answer to my backup problems, but now I'm not so sure. ================================================================================ Note 13.22 Backup Devices 22 of 41 EISNER::FRISBIE "Alan E. Frisbie" 10 lines 15-NOV-1987 03:15 -< Article in Digital Review >- -------------------------------------------------------------------------------- The current (09-Nov-87) issue of Digital Review has an article on Data Digital Audio Tape (DDAT) (page 118). It covers many of the same issues we have covered here. The projected transfer rate is only 200 KB/sec, but the capacity is expected to be over 1.0 GB (Yes, Gigabytes). Unfortunately, they don't expect them in any quantity until 1989. Alan ================================================================================ Note 13.23 Backup Devices 23 of 41 EISNER::NAGY "Frank J. Nagy, VAX Guru" 56 lines 13-FEB-1988 10:09 -< Gigastore usage and performance report >- -------------------------------------------------------------------------------- We have been using Digi-Data "Gigastore" VHS-based tape drives since the middle of November for archival storage purposes on our LAVC. The Gigastores are connected to a pair of MicroVAX-IIs which are the boot and core nodes (each serves a pair of RA81s to the entire cluster). We experienced one initial problem common to both drives -- occasionally, BACKUP would fail in the middle of a save signalling a message like "device offline or not in configuration". Digi-Data shipped us a new drive which we swapped with the most-failure-prone drive of the two. Since we have done so (a few weeks ago) we have not seen a recurrence of this failure on either drive! Both drives are functioning well enough that we are now depending upon them for our system backups. We offer no explanation for the reason that such errors have abated from the formerly less-failure-prone drive (and, indeed, the second drive as well). We have 4 RA81s as system disk and user file store. A pair of RA81s is served by each of the core MicroVAX-IIs (each equipped with a Gigastore drive). A batch job runs at 3 AM and saves a pair of RA81s onto a cassette in each Gigastore; the system manager replaces the cassettes in the morning with a fresh set. Of course, we organized operations so the RA81s saved by a particular MicroVAX are those RA81s directly connected to that MicroVAX (thus eliminating any unnecessary Ethernet overhead). Discounting the files marked /NOBACKUP, we are currenting saving about 360 MB to each of the Gigastores (total of roughly 720 MB in use). From MONITOR records made during running of these jobs, we find that the CPU utilization (on the MicroVAX-II) is in the range of 20-30%. The backups are done /NOCRC (in order to reduce loading on the MicroVAX.) We see the disk I/O rate averaging on the order of about 15 requests per second. This is in the middle of the night and the systems are otherwise idle. All this results in an effective useable I/O bandwidth to the Gigastore of about 60-65 KBytes/sec. (Digi-Data claims -- and there is no evidence to the contrary -- that the Gigastore is capable of sustaining 120 KBytes/sec). A test, saving all 4 RA81s to a single Gigastore cassette, saved 713 Megabytes in 3 hours and 9 minutes -- an effective rate of 64 KBytes/sec. The Gigastore is serving our basic needs even though we are a bit disappointed in the overall performance. BACKUP seems incapable of pumping data out fast enough for the Gigastore even though its usage of the CPU and disk are very modest. The Gigastore looks like a Pertec tape drive and is connected to a Dilog controller which emulates a TS11. A modified version of the standard TS11 driver is used as some timeouts in the software had to be adjusted. This is a bit of a concern and we are wondering what experience people are having with Exabyte systems on TMSCP controllers. Does anyone have experience with tape systems using the Exabyte drive which uses the 8 mm cassettes? The Exabyte system is claimed to run at 250 KBytes/second. What type of performance are you seeing? ================================================================================ Note 13.24 Backup Devices 24 of 41 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 65 lines 25-FEB-1988 12:05 -< Someone HAS used Exabyte. >- -------------------------------------------------------------------------------- RE: < Note 13.23 by EISNER::NAGY "Frank J. Nagy, VAX Guru" > -< Gigastore usage and performance report >- | Does anyone have experience with tape systems using the Exabyte drive | which uses the 8 mm cassettes? The Exabyte system is claimed to run | at 250 KBytes/second. What type of performance are you seeing? Here's a note I pulled from the Pageswapper that describes some experiences with Exabyte. Hope this helps. <<< NODE::DISK$PAGESWAPPER:[000000]IO.NOTE;3 >>> -< Submissions to the Pageswapper INPUT/OUTPUT Column >- ================================================================================ Note 913.0 Looking for Exabyte 8mm users No replies NODE::US151570 "KEVIN J. KUREK" 49 lines 23-FEB-1988 15:15 -------------------------------------------------------------------------------- How many of you have tried the 8mm 2.3 Gb drive from Exabyte? What has been your experience with the drive? I would like to get your feedback? Recently, I purchased one of these drives from a value added reseller. The drive was easy to install and took less then 30 minutes on a VAX 11/780. The drive works great. It truely holds 2.3Gbytes. To prove this I successfully backed up 5 RA81s that were 95+% full to the drive. Performance of the drive has proven to be adequate for my environment. I have seen transfer rates between 70 and 90 Kbytes/sec. I did expect higher transfer rates, but since backup is done during the night I really do not care how long it takes. My test showed that initializing (INIT dev: vol_name:) a tape on the average takes 2.25 min. Mounting (MOUNT/FOR dev:) an initialized tape took approximately 33 seconds (it takes 30 sec. to get the drive spinning). And the BACKUP/IMAGE of an RA81 took approximately 1.25 hours. The backup command used did have CRC turned on and used the default buffer count. The disk was very fragmented and contains many small files. (average block size is 35). One of my initial concerns about the drive was media reliability and data integrity. After reading several articles about the drive I discovered that the drive employs both ECC and Reed Solomon error checking. This gives the drive a non-recoverable error rate of less than on bit in 10**13, or about the same as the rates for a Winchester disk. Thus, I am no longer as concerned about data integrity. I am still concerned about media reliablity. The only problem that I have had so far is the reliability of the drive itself. The first drive I received lasted only 2 days before breaking while the 2nd drive lasted 2 weeks before breaking. I am currently on my 3rd drive and have had it for about 5 weeks without a problem. Considering that SONY makes 90% of the drive I believe that it is just coincidental. Lastly, I have found one major difference between an Exabyte drive and a Digi-data VHS drive. It is my understanding that whenever the VHS drive must stop and start that approximately 1 Mbyte of tape space is lost. On the other hand this transition on the 8mm tape only loses 8 Kbytes of tape space. If you would like to discuss this drive with me further please feel free to call. Kevin J. Kurek 1-414-929-4713 ================================================================================ Note 13.25 Backup Devices 25 of 41 EISNER::MILLER_R 74 lines 26-APR-1988 20:18 -< Exabyte experience >- -------------------------------------------------------------------------------- < Note 13.24 by EISNER::TILLMAN "Brian Tillman, Smiths Industries." > -< Someone HAS used Exabyte. >- RE: < Note 13.23 by EISNER::NAGY "Frank J. Nagy, VAX Guru" > -< Gigastore usage and performance report >- | Does anyone have experience with tape systems using the Exabyte drive | which uses the 8 mm cassettes? The Exabyte system is claimed to run | at 250 KBytes/second. What type of performance are you seeing? Here's a note I pulled from the Pageswapper that describes some experiences with Exabyte. Hope this helps. <<< NODE::DISK$PAGESWAPPER:[000000]IO.NOTE;3 >>> -< Submissions to the Pageswapper INPUT/OUTPUT Column >- ================================================================================ Note 913.0 Looking for Exabyte 8mm users No replies NODE::US151570 "KEVIN J. KUREK" 49 lines 23-FEB-1988 15:15 -------------------------------------------------------------------------------- How many of you have tried the 8mm 2.3 Gb drive from Exabyte? What has been your experience with the drive? I would like to get your feedback? Recently, I purchased one of these drives from a value added reseller. The drive was easy to install and took less then 30 minutes on a VAX 11/780. The drive works great. It truely holds 2.3Gbytes. To prove this I successfully backed up 5 RA81s that were 95+% full to the drive. Performance of the drive has proven to be adequate for my environment. I have seen transfer rates between 70 and 90 Kbytes/sec. I did expect higher transfer rates, but since backup is done during the night I really do not care how long it takes. My test showed that initializing (INIT dev: vol_name:) a tape on the average takes 2.25 min. Mounting (MOUNT/FOR dev:) an initialized tape took approximately 33 seconds (it takes 30 sec. to get the drive spinning). And the BACKUP/IMAGE of an RA81 took approximately 1.25 hours. The backup command used did have CRC turned on and used the default buffer count. The disk was very fragmented and contains many small files. (average block size is 35). One of my initial concerns about the drive was media reliability and data integrity. After reading several articles about the drive I discovered that the drive employs both ECC and Reed Solomon error checking. This gives the drive a non-recoverable error rate of less than on bit in 10**13, or about the same as the rates for a Winchester disk. Thus, I am no longer as concerned about data integrity. I am still concerned about media reliablity. The only problem that I have had so far is the reliability of the drive itself. The first drive I received lasted only 2 days before breaking while the 2nd drive lasted 2 weeks before breaking. I am currently on my 3rd drive and have had it for about 5 weeks without a problem. Considering that SONY makes 90% of the drive I believe that it is just coincidental. Lastly, I have found one major difference between an Exabyte drive and a Digi-data VHS drive. It is my understanding that whenever the VHS drive must stop and start that approximately 1 Mbyte of tape space is lost. On the other hand this transition on the 8mm tape only loses 8 Kbytes of tape space. If you would like to discuss this drive with me further please feel free to call. Kevin J. Kurek 1-414-929-4713 We've done some testing with the Exabyte. We backup an RA81 using BACKUP with NOCRC but with VERIFY in 1 to 1 1/2 hours on a microVAX II. More to come. We'll start using the Exabyte heavily in the next three months using two Exabyte drives to backup a VAXcluster disk farm and an LAVC disk farm. ================================================================================ Note 13.26 Backup Devices 26 of 41 EISNER::LAMB 5 lines 1-JUN-1988 18:23 -< Summus-Gigatape or CY-2000 >- -------------------------------------------------------------------------------- Has anyone had experiance with the Summus-Gigatape from Summus Computer Systems or the CY-8200 fr Contemporary Cybernetics? My field of f choice has been pretty much narrowed down to one of these two. These drives will be primarily used for backup of one or two RA81s attached to a uVAX II. ================================================================================ Note 13.27 Backup Devices 27 of 41 EISNER::TANNENBAUM "Saul Tannenbaum" 5 lines 1-JUN-1988 19:09 -< VMS 5.0 and backup devices >- -------------------------------------------------------------------------------- This question is probably premature, but... How are all these foreign VCR backup devices surviving the migration to VMS Version 5? Does any vendor yet offer a VMS 5.0 compatible driver? Has anyone actually used one on a VMS 5.0 system? ================================================================================ Note 13.28 Backup Devices 28 of 41 EISNER::NAGY "Frank J. Nagy, VAX Guru" 3 lines 2-JUN-1988 07:31 -< Emulate and win! >- -------------------------------------------------------------------------------- At our LUG meeting last night, one DECUS attendee reported that one of the V5 Field Test sites said they had successfully used the Summus controller (emulates a TK50) under VMS V5 (FT2) with no problems. ================================================================================ Note 13.29 Backup Devices 29 of 41 EISNER::MERRIMAN "George Merriman - CCA/NY" 43 lines 2-OCT-1988 17:11 -< Exabyte tape drive >- -------------------------------------------------------------------------------- I have extracted the following question and comments on Exabyte tape units from WHO_AM_I: <<< EISNER::DUA0:[NOTES$LIBRARY]WHO_AM_I.NOTE;1 >>> -< A Repository of People's Background >- ================================================================================ Note 203.0 Dorn Williams 6 replies EISNER::WILLIAMS_D "Dorn" 22 lines 28-SEP-1988 13:15 -------------------------------------------------------------------------------- . . . we are going to be buying a 850Meg CDC hard disk and hopefully an Exabyte tape drive. We are getting the CDC drive with the hope that its faster seek and transfer times will improve performance on the MVII. If anyone has had experience (good or bad) with the Exabyte drive please add a reply. . . <<< EISNER::DUA0:[NOTES$LIBRARY]WHO_AM_I.NOTE;1 >>> -< A Repository of People's Background >- ================================================================================ Note 203.1 Dorn Williams 1 of 6 EISNER::ESBENSEN 16 lines 29-SEP-1988 00:58 -< Exabyte drives >- -------------------------------------------------------------------------------- Hi, We have been testing Exabyte drives for a few months. We have been through MANY rev levels. We have tried Megatapes and Dialogs sub-systems. Using a Tape Accelerator that we have, we were able to write out at a sustained 220kB/sec for a backup of 5700 files in 71 directories...on our MVII. Although it makes me a bit nervous about new rev levels each month, I have recommeded that we buy one. Our MVII has 1.2GB of storage...and we do full backups each day (14 tapes). With Exabyte unit we can do the backups unassisted each night. This looks like it will free up 30% of our CPU and I/Os during the day (from BACKUPs running!!). Dan E. ================================================================================ Note 13.30 Backup Devices 30 of 41 EISNER::BURNET "John Burnet" 13 lines 31-JAN-1989 11:07 -< Does an 8mm drive with an Ethernet interface exist? >- -------------------------------------------------------------------------------- Does anyone know of an 8mm or DAT drive with an Ethernet interface that can be configured as a "backup server" and used from more than one system on the Ethernet? (Only one system at a time, of course. ;-) This would certainly seem to be an idea whose time has come... the transfer rates of VHS, 8mm, and DAT devices (50-250 KB/sec?) fit in pretty well with the throughput that's attainable with DEC's controllers on a lightly-loaded Ethernet. Like a lot of other sites, we have more than one VAX that we'd like to hang an 8mm drive on, but we can't afford more than one drive and the VAXen (in our case) are a Unibus machine and a BI machine, so it's not likely that we'd be able to move the drive from one computer to the other. Thanks in advance for any leads that you may be able to supply. ================================================================================ Note 13.31 Backup Devices 31 of 41 EISNER::CHADWICK "Keith Chadwick, Fermilab" 6 lines 31-JAN-1989 12:04 -< It's called a microVAX... 8-). >- -------------------------------------------------------------------------------- > < Note 13.30 by EISNER::BURNET "John Burnet" > > -< Does an 8mm drive with an Ethernet interface exist? >- > > Does anyone know of an 8mm or DAT drive with an Ethernet interface that can be > configured as a "backup server" and used from more than one system on the > Ethernet? (Only one system at a time, of course. ;-) ================================================================================ Note 13.32 Backup Devices 32 of 41 EISNER::MAYHEW "Bill Mayhew" 8 lines 31-JAN-1989 13:12 -< Whatta concept... >- -------------------------------------------------------------------------------- We (that's the "royal" we, meaning "all of us" :-} ) now have TWO devices that we'd like to be able to hang on an Ethernet, with just enough attached intelligence to serve them to Real Computers: an 8mm backup device, and a CD-ROM. Now, if only either one (or both?!) could be made to work on a VAXstation 2000, the cost of using a "MicroVAX" as a source of server intelligence might actually be bearable... -Bill ================================================================================ Note 13.33 Backup Devices 33 of 41 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 33 lines 31-JAN-1989 15:37 -< one 8MM, >one CPU, existing controllers >- -------------------------------------------------------------------------------- > Like a lot of other sites, we have more than one VAX that we'd > like to hang an 8mm drive on, but we can't afford more than one drive and > the VAXen (in our case) are a Unibus machine and a BI machine, so it's > not likely that we'd be able to move the drive from one computer to > the other. We plan to share one 8mm drive between existing computers on EXISTING controllers. Ethernet would be great, and I am not sure about controller issues on BI machines, BUT for Q and U there is a simple answer. We have been using an electronic tape drive switch that works with any tape drive with the somewhat popular Pertec interface. It has 4 knobs each representing a tape drive (or several daisy chained that are ok to switch together), and each knob has 4 positions for 4 CPUs. There are other geometrys available, and we bought the box with a good knob selection (small $ increase) but only bought the cards inside for what we initially needed to save $s. This switch box is the one Iverson markets, and is actually made by someone else. We did NOT buy the east coast tape switch brand even though they are physically near us. If you are seriously looking and want to know my reasons, send me mail. There is at least one brand of 8MM drive that offers Pertec rather than SCSI, and we could run such a tape off our switch off our existing Dilog tape controllers. On the other hand, has anyone seen such a switch for SCSI peripherals? ================================================================================ Note 13.34 Backup Devices 34 of 41 EISNER::BURNET "John Burnet" 24 lines 31-JAN-1989 18:34 -< Ethernet, BI, Unibus, argggh >- -------------------------------------------------------------------------------- > < Note 13.31 by EISNER::CHADWICK "Keith Chadwick, Fermilab" > > -< It's called a microVAX... 8-). >- Great concept -- I wish we could go that route, but the Unibus machine is a 750, and before too long the 750's users would start to resent the fact that their tape server has more CPU power than their timesharing machine (not to mention four times the resale value). What we really *should* do is jettison the 750, but you know how that goes... a drive & controller purchase is a lot easier to justify than a CPU switch. This is also one reason we don't have a LAVc -- the 750 would have to be the boot node, since it doesn't have the ROM intelligence to Etherboot. > < Note 13.33 by EISNER::BRUCE "Barton F. Bruce - CCA/MA" > > We have been using an electronic tape drive switch that works > with any tape drive with the somewhat popular Pertec interface. Very interesting... thanks for the info. We may consider that solution, but (1) one of the machines *is* a BI machine (with, at this point, only a KLESI-B and a TU81+ :-( as far as tape devices are concerned); and (2) we were kinda hoping for some software-switchable method so that we could fire up a batch job to back up both systems unattended overnight. (I guess a solenoid and a D/A board would work... 8-) ================================================================================ Note 13.35 Backup Devices 35 of 41 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 27 lines 3-FEB-1989 03:26 -< 8MM w/ Pertec costs >- -------------------------------------------------------------------------------- > We plan to share one 8mm drive between existing computers on > EXISTING controllers. > There is at least one brand of 8MM drive that offers Pertec > rather than SCSI, and we could run such a tape off our switch > off our existing Dilog tape controllers. This sounds ok until you get the prices. The 8mm drive with a Pertec interface is ~$7500. without any controller. The same company sells a SCSI version WITH a Q bus controller for around $5000. Why the difference in $s? First, they are the only 8mm with Pertec, but another reason is that the 8mm drives take 10 sec to write a file mark, and some drivers (e.g. MS:) may timeout, so they built a 1.5 meg buffer in front so the drive can keep the driver happy by taking more commands (and fibbing about the success of prior ones...). This lets them use ANYONES Pertec controller on any brand computer. TMSCP doesn't have this problem, so their SCSI system uses CMD's TMSCP SCSI controllers. > On the other hand, has anyone seen such a switch for SCSI > peripherals? But am I not allowed to put >1 controller on a SCSI bus! ================================================================================ Note 13.36 Backup Devices 36 of 41 EISNER::KENNEDY "Terry Kennedy" 6 lines 3-FEB-1989 04:46 -< Technically feasible, but... >- -------------------------------------------------------------------------------- > But am I not allowed to put >1 controller on a SCSI bus! Indeed. SCSI allows multiple controllers. Whether DEC drivers support such configurations (or even know about them) is another issue. After all, some DEC operating systems still have problems with dual-ported RM and RP drives! ================================================================================ Note 13.37 Backup Devices 37 of 41 EISNER::BURNET "John Burnet" 10 lines 15-JUN-1989 21:41 -< Looking for opinions on Megatape 8mm >- -------------------------------------------------------------------------------- Does anyone have any experience with or knowledge of the Megatape 8mm products? We're considering a Megatape 8mm drive with a Unibus controller; maybe we'll spring for a second drive and a BI controller from the same company later on, if DEC's lawyers don't get there first. Is there anything in particular that makes the Megatape hardware better or worse than, say, Exabyte or Dilog or any of the other 8mm products on the market? Any info, war stories, etc. would be greatly appreciated. ================================================================================ Note 13.38 Backup Devices 38 of 41 EISNER::NAGY "Frank J. Nagy, VAX Guru" 8 lines 16-JUN-1989 08:14 -< 8mm transports by Sony, interface by Exabyte >- -------------------------------------------------------------------------------- >>Is there anything in particular that makes the Megatape hardware better or >>worse than, say, Exabyte or Dilog or any of the other 8mm products on the >>market? Everyone should be aware that all the 8mm drives are manufactured by Sony, then refitted by Exabyte (who adds a SCSI interface board) and sold with a controller by these myriad of other companies. The tape transport and interface are single-source products. ================================================================================ Note 13.39 Backup Devices 39 of 41 EISNER::KENNEDY "Terry Kennedy" 5 lines 16-JUN-1989 09:42 -< Wangtek? >- -------------------------------------------------------------------------------- > Everyone should be aware that all the 8mm drives are manufactured by > Sony, then refitted by Exabyte (who adds a SCSI interface board) ... I have a brochure (somewhere) from Wangtek with (I think) an 8MM drive they do the logic for. Then again, it may have been a DAT drive... ================================================================================ Note 13.40 Backup Devices 40 of 41 EISNER::BURNET "John Burnet" 15 lines 16-JUN-1989 22:40 -< 8mm drives, VAXBI controllers, Megatape >- -------------------------------------------------------------------------------- > Everyone should be aware that all the 8mm drives are manufactured by > Sony, then refitted by Exabyte (who adds a SCSI interface board) and > sold with a controller by these myriad of other companies. The tape > transport and interface are single-source products. Very interesting. Then that would mean that the major differences are price, service (e.g., whether DEC will let the customer put the drive and controller on his hardware service contract), and the particular company's reputation and life expectancy. On the other hand... what about that Megatape VAXBI controller? I noticed that Megatape had none on display at DEXPO. I wondered at the time if that had anything to do with DEC's legal muscle-flexing re. SI's DSA drives. Does anyone use the Megatape BI controller? I'm curious to know who makes it and whether it's likely to remain on the market. ================================================================================ Note 13.41 Backup Devices 41 of 41 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 8 lines 20-JUN-1989 10:08 -< Some spec differences between vendors. >- -------------------------------------------------------------------------------- > Everyone should be aware that all the 8mm drives are manufactured by > Sony, then refitted by Exabyte (who adds a SCSI interface board) and > sold with a controller by these myriad of other companies. The tape > transport and interface are single-source products. At least one company (Summus) lists an error rate one or two orders of magnitude better than the others. Do they do something funda- mentally different?  ================================================================================ Note 14.0 Not another DLVJ1 problem! 7 replies EISNER::MCGLINCHEY 22 lines 19-APR-1987 12:58 -------------------------------------------------------------------------------- When, of when will DEC get rid of the DLV-11/J? Or its alias, the DLVJ1? I jsut reviewed a system configuration for a MicroVAX II, as configured by whatever the sales people do to get a configuraiton, and in the quotation I saw: 1) a DLVJ1 2) 2 DZQ11's. I calle dthe sales person and she said that the DLVJ1's were in there because the customer asked for RS422 ports (honest!) and the DZQ's were in there instead of a DHV11 because (she said) it would be cheaper. I decided it was pointless to discuss the bus treachery of the DLV-11/J, but pursued the DZQ business a tad. 2 dZQ's are NOT cheaper the 1 DHV11, I replied. And then she admitted that DC starts with a packaged system and adds to it whatever is necessary to meet the customer's request. Oh, dear. ================================================================================ Note 14.1 Not another DLVJ1 problem! 1 of 7 EISNER::CETRON 6 lines 21-APR-1987 02:32 -< yes another problem >- -------------------------------------------------------------------------------- but jimmy me boy, thats the irish way.... (or is that whey???) ================================================================================ Note 14.2 Not another DLVJ1 problem! 2 of 7 EISNER::MCCARTHY 4 lines 12-MAY-1987 20:06 -< What DLVJ1? >- -------------------------------------------------------------------------------- > When, of when will DEC get rid of the DLV-11/J? > Or its alias, the DLVJ1? Oh, please PLEASE could we ? We thought you guys would object. ================================================================================ Note 14.3 Not another DLVJ1 problem! 3 of 7 EISNER::FRISBIE "Alan E. Frisbie" 31 lines 12-MAY-1987 22:13 -< Why users want the DLV11-J (DLVJ1) >- -------------------------------------------------------------------------------- The reason that many people want DLV11-J boards is so that they can write special drivers for their RS-232 widgits using a cheap (relatively), simple device. One look at the DZ- or DH-type manuals is enough to scare off the novice driver writer. I think what they really want is a simple-to-program serial device. This means a single channel per controller (multiplexors are difficult to program). If it has multiple controllers per board, that is a bonus. It should ALLOW interrupt-per-character operation, but DMA for both input and output would be a big plus. Naturally, it needs a DEEP silo to prevent data overruns. Besides the usual (and higher) baud rates, it should support an external clock for the more bizzare applications. Since it will be used with non-terminal devices, some form of hardware handshaking/flow control should be provided (DTR, CTS or whatever). I know that the DHQ11 does most of this. However, the average user is scared off by the prospect of writing a driver for an 8-line multiplexor when something simple like the DLV11-J exists. Of course, RSX users can always use McGlinchey's driver for the DZ or DH, but 99.99% of them don't know that it exists, and it doesn't help RT-11 or VMS users. Alan ================================================================================ Note 14.4 Not another DLVJ1 problem! 4 of 7 EISNER::MCGLINCHEY "Cape Malleum Majorem" 32 lines 17-MAY-1987 11:43 -< Dig a hole and bury it. >- -------------------------------------------------------------------------------- >> The reason that many people want DLV11-J boards is >> so that they can write special drivers for their >> RS-232 widgits using a cheap (relatively), simple >> device. >> I think what they really want is a simple-to-program >> serial device. AYE AYE!! >> It should ALLOW interrupt-per-character >> operation, but DMA for both input and output would >> be a big plus. What happened to 'simple'? DMA is NOT simple on a PDP-11, particularly for a novice! >> Naturally, it needs a DEEP silo >> to prevent data overruns. This, plus the single-interrupt-level nature of the DLVJ1, make it the miserable device it has always been. Brian, I'll be the first to vote for the deletion of the DLVJ1 from the product line. Jim. ================================================================================ Note 14.5 Not another DLVJ1 problem! 5 of 7 EISNER::FRISBIE "Alan E. Frisbie" 25 lines 17-MAY-1987 14:46 -< Clarification on DLVJ1 replacement >- -------------------------------------------------------------------------------- || >> It should ALLOW interrupt-per-character || >> operation, but DMA for both input and output would || >> be a big plus. || || What happened to 'simple'? DMA is NOT simple on a PDP-11, || particularly for a novice! 'Simple' is why it should allow interrupt-per-character operation -- for people that don't feel comfortable with DMA. This inefficient mode of operation is why it also needs a deep silo on input. DMA is for the people that know how to use it. || I'll be the first to vote for the deletion of || the DLVJ1 from the product line. I'll be the second to make the same vote, plus I'll be first in line to buy the new DFQ11 (The "F" is for Frisbie :-) ). Grant Technology has a new serial board with DMA input, but it has so many bells and whistles that it is far from simple. Plus, it is on a quad board. There is so much empty space between the chips that you could lease it out to graze cattle. Alan ================================================================================ Note 14.6 Not another DLVJ1 problem! 6 of 7 EISNER::CETRON 14 lines 17-MAY-1987 16:44 -< f for finally >- -------------------------------------------------------------------------------- > Plus, it is on a quad board. There is so much empty space >between the chips that you could lease it out to graze cattle. goody, we can go cow-tipping..... > in line to buy the new DFQ11 (The "F" is for Frisbie :-) ). actually the F stands for 'finally' I would buy several and also would love to be present to give the dlv11-j (so i'm old) a formal viking burial.... ================================================================================ Note 14.7 Not another DLVJ1 problem! 7 of 7 EISNER::FRISBIE "Alan E. Frisbie" 24 lines 23-MAY-1987 00:08 -< ANOTHER DLV11-J problem (and solution) >- -------------------------------------------------------------------------------- I got a call this morning from a customer whose printer was printing gibberish (other than the normal application program gibberish). It was an LA100 connected to the second serial port on the KDF11-B (11/23) CPU board. The operating system was RT-11. Field service had already checked baud rates, cables, etc. and had keyed in a short test program to print the alphabet. It worked perfectly. TSC suggested that the RT-11 driver was corrupted and should be replaced with one from a backup. No joy. What I discovered was that someone (everyone denies that they even opened the box) had installed a DLV11-J at the same CSR and Vector (176500/300) as the serial port on the CPU board. The port on the CPU board was printing about every forth or fifth character. What I can't figure out is how it EVER worked, since they claim that nobody even opened the cabinet. (The solution: since nothing was connected to the DLV11-J, I just unplugged it -- the best thing to do to it.) Alan  ================================================================================ Note 15.0 Cabling Solutions 17 replies EISNER::LEFEBVRE "Kenneth LeFebvre" 7 lines 20-APR-1987 10:49 -------------------------------------------------------------------------------- I often find that I have little problems with cables and connectors and the like that turn into days of work. Unfortunately, I often find that the solution was simple and has been discovered before by somebody else, too. How about if we share some of those "little" solutions here. Particularly, cabling-type problems. ================================================================================ Note 15.1 Cabling Solutions 1 of 17 EISNER::LEFEBVRE "Kenneth LeFebvre" 5 lines 20-APR-1987 10:52 -< DEC Printers to IBM Computers >- -------------------------------------------------------------------------------- I have a non-DEC computer (IBM PC) that I'd like to use my DEC printers (esp., LA75 & LN03) with. Just plugging them in doesn't work... Any ideas? ================================================================================ Note 15.2 Cabling Solutions 2 of 17 EISNER::SALLEE 8 lines 20-APR-1987 11:03 -< Tight fitting RJ-11 connectors >- -------------------------------------------------------------------------------- Subject brand X RJ-11 (telephone) connectors... are becoming popular for terminal hook up. I had a problem with some being a very tight fit. They would relieve themselves of the cable before disconnecting from the socket. solution: a flat mill file used to attack and round about .015 inch from the four edges parallel to the connection face. ================================================================================ Note 15.3 Cabling Solutions 3 of 17 EISNER::FRISBIE "Alan E. Frisbie" 87 lines 22-APR-1987 13:35 -< LP(V)11 to Centronics cable >- -------------------------------------------------------------------------------- People often ask me if it is possible to connect a printer with the Centronics (tm) parallel interface to a DEC computer. Not only is it possible, but DEC used to sell them! Although the signals for the Data Products line printers and the Centronics printers are different, they are similar enough that by connecting one wire in the cable to +5 or to ground the same interface can be used with either. The interface simply switches its control logic from Data Products mode to Centronics/LA-180 mode. This works with the Q-Bus LPV11 and I believe it is also true with the Unibus LP11. The trick to remember is that the DEC LA-180 uses timing and control signals similar to the Centronics printers. The DEC part number for the Centronics-compatible cable is 70-09087. Unfortunately, this has been an "obsolete" part for several years, so you may only be able to find it through third-party cable vendors. If you wish to build your own, the following table should tell you everything you need to know. You might consider taking an LA-180 cable and replacing the Berg connector with a Centronics-style connector. I have been using this for two years with my PDP-11/73 and a Toshiba P1351 printer with no problems and great results. If you have any problems, please let me know. Connections for DEC LPV11 (M8027) to Centronics printer Ribbon DEC (Berg) Centronics Signal Cable Connector Connector Name wire # pin # pin # 1 A 2 B 3 C 27 Ground 4 D 9 Data 8 (MSB) 5 E 29 Ground 6 F 11 Busy 7 H 8 J 9 K 10 L 11 M 12 N 13 On Line 13 P 14 R 32 Fault L 15 S 16 "LA180" (Ground) 16 T 17 U 18 V 14 "Connected" 19 W 28 Ground 20 X 10 Demand (ACKNLG L) 21 Y 22 Z 23 AA 22 Ground 24 BB 4 Data 3 25 CC 26 Ground 26 DD 8 Data 7 27 EE 23 Ground 28 FF 5 Data 4 29 HH 20 Ground 30 JJ 2 Data 1 (LSB) 31 KK 21 Ground 32 LL 3 Data 2 33 MM 34 NN 35 PP 25 Ground 36 RR 7 Data 6 37 SS 24 Ground 38 TT 6 Data 5 39 UU 19 Ground 40 VV 1 Strobe L DEC (Berg) connector pins (Looking at male connector on board) Pin "A" connects to ribbon cable wire #1 B D F J L N R T V X Z BB DD FF JJ LL NN RR TT VV A C E H K M P S U W Y AA CC EE HH KK MM PP SS UU Alan Frisbie Flying Disk Systems, Inc. ================================================================================ Note 15.4 Cabling Solutions 4 of 17 EISNER::PROVOST 4 lines 22-APR-1987 15:57 -< GREAT, but DMF? >- -------------------------------------------------------------------------------- GREAT!! Now, what if you have a DMF32? My sources have told me that the hardware manual has the wrong pin-out for the printer port. tjp ================================================================================ Note 15.5 Cabling Solutions 5 of 17 EISNER::EISNER "Dan L. Eisner" 4 lines 22-APR-1987 22:36 -< Yea, What about the DMF-32? >- -------------------------------------------------------------------------------- Can yo add a DMF-32 to a Centronics type printer? What driver do you load? Where do you point it? What are the commands to make it work (if the hardware can be hooked to it)? I would like to know since I have a 730 and a DMF-32. ================================================================================ Note 15.6 Cabling Solutions 6 of 17 EISNER::FRISBIE "Alan E. Frisbie" 5 lines 23-APR-1987 01:46 -< So you want DMF-32 to Centronics? >- -------------------------------------------------------------------------------- OK guys! If someone will send me a copy of the DMF-32 manual, I will figure out the differences (if any) and post another note. (This is how I build up my manual collection!) :-) Alan ================================================================================ Note 15.7 Cabling Solutions 7 of 17 EISNER::WELBORNE 7 lines 6-MAY-1987 23:53 -< DEC printer on a IBM PC >- -------------------------------------------------------------------------------- Ken we have our dec printers connected using the serial interface. We have had some problems with flow control since the IBM serial interface does not support x-on x-off.. In addition we had to jumper a couple of pins. I don't recall which at the moment but I'll get them tomorrow and let you know. Jim ================================================================================ Note 15.8 Cabling Solutions 8 of 17 EISNER::FRISBIE "Alan E. Frisbie" 8 lines 12-MAY-1987 00:50 -< I now have a DMF-32 manual >- -------------------------------------------------------------------------------- OK, Jim English called my bluff and sent me a DMF-32 manual. I haven't looked at it yet, so I don't know if I'll be able to come up with a Centronics-compatible cable. I will let you know in a few days. Watch this space... Alan ================================================================================ Note 15.9 Cabling Solutions 9 of 17 EISNER::PROVOST 15 lines 4-JUN-1987 15:10 -< Tomorrow yet? >- -------------------------------------------------------------------------------- < < Note 15.7 by EISNER::WELBORNE > < -< DEC printer on a IBM PC >- < < Ken we have our dec printers connected using the serial interface. < We have had some problems with flow control since the IBM serial < interface does not support x-on x-off.. In addition we had to jumper < a couple of pins. I don't recall which at the moment but I'll get < them tomorrow and let you know. < < Jim Is it tomorrow yet? We have been trying to hook LN03+ to IBM PC clone using the serial interface. We're awaiting your solution. Patient Tom ================================================================================ Note 15.10 Cabling Solutions 10 of 17 EISNER::FRISBIE "Alan E. Frisbie" 11 lines 4-JUN-1987 16:54 -< DMF-32 manual still needed >- -------------------------------------------------------------------------------- >> OK, Jim English called my bluff and sent me a DMF-32 manual. >> I haven't looked at it yet, so I don't know if I'll be >> able to come up with a Centronics-compatible cable. >> I will let you know in a few days. Well, the manual turned out to be the Owner's Manual which has lots of information for the I/O Driver writer, but not enough for the hardware hacker. What I need is the Technical Manual, or perhaps even the print set (I hope not). Alan ================================================================================ Note 15.11 Cabling Solutions 11 of 17 EISNER::ROCHELEAU 15 lines 14-AUG-1987 17:19 -< RA81 to HSC cables CHEAP >- -------------------------------------------------------------------------------- *** BC26V SDI Cables *** I recently upgrade a large number of RA81 disks to four high cabinets. This left me with a large pile (40 disks) internal SDI cables. I pondered what to do with these cables for a while and finally found a solution. Send then to a cable manaufacturer and have them modified to be standard SDI cables ... about four feet in length. I figure the cost of this to be about $15 per cable, thats a lot better that spending $350 or more on a 12 foot cable and coiling up 8 feet under the disk cabinet. I not going to advertise for any body but of you can't find a vendor, I have one that has build / repaired cables for me that work with no problems. My F.E. is hacked off because he can't tell the real DEC cables from the third party. ================================================================================ Note 15.12 Cabling Solutions 12 of 17 EISNER::LEFEBVRE "Kenneth LeFebvre" 53 lines 26-OCT-1987 15:43 -< Using an LN03 with an IBM PC/XT >- -------------------------------------------------------------------------------- > < Note 15.9 by EISNER::PROVOST > Tom, I'm sorry I haven't followed this conference enough to respond earlier. Here is the solution that I came up with to my own question! This is everything, I think. If I have forgotten something (as you will surely find out if I did, when you try it...) please let me know so I can update both this note and the little file I made to upload these instructions. ---------------------------A T T A C H M E N T---------------------------------- Using an LN03 with an IBM PC/XT 1. Add the following lines to AUTOEXEC.BAT: MODE COM1:48,N,8,1,P MODE LPT1:=COM1: 2. Make a connector as follows: PC LN03 ---- ---- 1 1 2 3 3 2 4 5 5 11 7 7 6 PC-8 -- 8 PC-20 -- 20 PC-6 -- 3. Set the DIP switches as follows on the LN03: Switch Pack One (SP1) 1 2 3 4 5 6 7 OFF OFF ON ON OFF OFF OFF Switch Pack Two (SP2) 1 2 3 4 5 6 7 OFF OFF OFF OFF ON ON OFF 4. Power everything down and back up. ================================================================================ Note 15.13 Cabling Solutions 13 of 17 EISNER::KOZAM 18 lines 12-DEC-1987 00:07 -< LP(V)11 as a partial DR(V)11 substitute >- -------------------------------------------------------------------------------- On the subject of the LPV11, has anyone used the LPV11 as a general-purpose parallel output interface? I realize that it won't have the features of a DRV11, but I think it would provide 8 output lines (bits 1 - 8) and four input lines (ONLINE, BUSY, DEMAND, and FAULT). DEMAND and FAULT can even cause interrupts on transition. My preliminary poking around leads me to believe that the LPV11 only pulses the data, while a DRV11 sets the voltages and holds them. Is this true? (I don't have an oscilloscope to look at the output waveform - otherwise I'd have the answer.) Can I make the LPV11 hold the voltages? If anyone has made the LPV11 work as a general-purpose board, I'd like to hear about did it. Marc Kozam ================================================================================ Note 15.14 Cabling Solutions 14 of 17 EISNER::ERSKINE "Chris Erskine" 17 lines 14-DEC-1987 06:27 -< Strobes >- -------------------------------------------------------------------------------- I have not played with the LPV-11 but after building LP-11 interfaces and designing LP interfaces for other systems, the voltages are just pulsed. What happens is the data is loaded into the output register and then the data strobe is fired. This is used to load the data into the line printer. After the data strobe goes low, the output register is cleared for the next byte to be loaded. If you want the interface, signals to be latched, you could add your own latches on the output lines and use the data strobe to store the data. Note that some of the input lines are used for hardware interlocking of the output lines. As I recall, you can not output data when the printer is in the BUSY state. In fact it might of even been that you had to raise DEMAND to output the next byte. A lot of this is dependent on what the hardware designer did when he designed the board. What you might want to do is to look at the logic prints and see what does happen. ================================================================================ Note 15.15 Cabling Solutions 15 of 17 EISNER::CONROY 25 lines 30-JUN-1988 15:31 -< Reverse LAT (DECServer to PDP-11) >- -------------------------------------------------------------------------------- I spent a couple hours on this, so I thought I'd share it (if anyone's in the same boat). To connect a DECServer-200/MC to a PDP-11, using modem control: The DB25s should have the following pins straight through: 2,3,4,5,6,7,8 In addition, pin 20 should be connected to pin 6 on the PDP side. Also, pin 22 should be connected to 4, 5, AND 8 on the PDP side. NOTE: This (and only this) worked on an Emulex distribution panel. One would hope that Able and DEC distribution panels would work in a similar fashion, but I know better than to make those assumptions. Also (and I know this is not really cable-related), here are the Server port settings needed: Access remote AltSpeed none Autobaud dis Dsrlogout dis modem control ena dtrwait ena (But all this is in the Server manual) And don't forget to set the RSTS/E terminal setting to modem (I don't know the RSX or RT11 equivalent). Other cable configurations "worked", but this one provided an automatic prompt when the PDP was connected to, and also disconnected when you logged out. All in all, the reverse LAT worked great this way! ================================================================================ Note 15.16 Cabling Solutions 16 of 17 EISNER::KENNEDY "Terry Kennedy" 12 lines 30-JUN-1988 18:25 -< Question, pointers... >- -------------------------------------------------------------------------------- > Also, pin 22 should be connected to 4, 5, AND 8 on the PDP side. Are you *sure* about this part? You've got Ring Indicator going to Request to Send, Clear to Send and Carrier Detect... By the way, DEC had an article in the RSTS/E Dispatch (last year, in June I think) giving the correct cable pinouts for DEC panels and the DS200 settings as well. Lastly, DEC announced (in Cinci) that 'real' LAT support is coming to RSTS in the near future. If you have RSTS-specific cable or setup questions, you might try the RSTS_OS conference under the Terminal Service topic. ================================================================================ Note 15.17 Cabling Solutions 17 of 17 EISNER::CONROY "Alan Conroy" 20 lines 18-JUL-1988 13:53 -< Believe it or not... >- -------------------------------------------------------------------------------- > Are you *sure* about this part? You've got Ring Indicator going to >Request to Send, Clear to Send and Carrier Detect... Yes! As weird as it appears, that was the only way to get it to work. That's why it took me a couple hours with a break-out box to figure it out. It's otherwise a simple connection, but the Emulex dist. panel is apparently brain damaged. > By the way, DEC had an article in the RSTS/E Dispatch (last year, in June >I think) giving the correct cable pinouts for DEC panels and the DS200 >settings as well. I remember it. It had the {standard" connections for this kind of application, but as I noted, the Emulex panel didn't respond well to the standard. > Lastly, DEC announced (in Cinci) that 'real' LAT support is coming to >RSTS in the near future. If you have RSTS-specific cable or setup questions, >you might try the RSTS_OS conference under the Terminal Service topic. It's about time! :-)  ================================================================================ Note 16.0 VT52 parity problem 1 reply EISNER::COAR "Ken A L Coar - OPUS/OPS" 12 lines 18-MAY-1987 15:28 -------------------------------------------------------------------------------- I have a user who has a VT52, which he claims is hardwired to 7-bit parity. Unfortunately, he needs to access a machine for which the inbound ports are set to 8-bit. The ports CANNOT be changed; he can't get another terminal. I have heard a rumour to the effect that a VT52 *can* be enabled for 8-bit use. Can anyone tell me if this is true, and if so, how it's done?? Thanx.. #k ================================================================================ Note 16.1 VT52 parity problem 1 of 1 EISNER::FRISBIE "Alan E. Frisbie" 30 lines 18-MAY-1987 15:57 -< VT52's can be set for 8-bit, no parity >- -------------------------------------------------------------------------------- >> I have heard a rumour to the effect that a VT52 *can* be enabled >> for 8-bit use. Can anyone tell me if this is true, and if so, how >> it's done?? It is trivial. Tip the VT52 back to expose the bottom cover and remove it. At the far left side of the printed circuit board are two slide switches. The left one (closest to the edge of the board) is unused, the right one selects parity on or off. About two inches below (towards the rear of the terminal) the switch is the only 40-pin IC on the board. Just to the right of it are two sets of jumper pads. The one closest to the chip is called W6 and the other is called W5. They select the flavor of the eighth (parity) bit according to the following table: Parity Switch ON Parity Switch OFF (towards front (towards rear of terminal) of terminal) W5 IN xxx 8th bit SPACEing (0) W5 OUT xxx 8th bit MARKing (1) W6 IN ODD parity xxx W6 OUT EVEN parity xxx This is all documented in the sticker on the bottom of the VT52. If yours is missing, I could send you a copy of mine. It will probably cause some funny looks as I set my VT52 on top of the Xerox machine, however. Alan  ================================================================================ Note 17.0 KA620 16 replies EISNER::PROVOST 14 lines 19-MAY-1987 09:18 -------------------------------------------------------------------------------- This is a continuation of the technical discussion of the KA620 chip. This discussion began in the VMS conference, and newcomers might want to review previous discussion there. The issue at hand is whether it is indeed a new chip, and whether VMS could be modified to run on it. This is a better place to discuss the technical differences between the KA620 and the KA630. An extremely good technical session was given at our LUG by engineers involved in the design of the new chip. It is a new chip. I will attempt to get some technical documentation for you.ÿ tjp ================================================================================ Note 17.1 KA620 1 of 16 EISNER::FRISBIE "Alan E. Frisbie" 30 lines 19-MAY-1987 19:53 -< The needle in the stack >- -------------------------------------------------------------------------------- With both Brian and Tom hoping to fleece the sucker (me), I figured that a closer look was in order. I found the difference between the KA630 and the KA620 (besides the SID Register value). The key is in Chapter 4 of the KA620 (KA630) User's Guide, in the paragraph, "P0 Region Address Translation" (4.8.5.1 for the KA620, 4.8.4.1 for the KA630): For the KA630, the P0BR (P0 Base Register) contains the SYSTEM VIRTUAL address of the P0 Page Table. For the KA620, the P0BR contains the the PHYSICAL address of the P0 Page Table. The same is true for the P1 region address translation. This means that the process Page Tables must be physically contiguous if a process is more than 64K bytes long (one page of Page Table Entries). This will make dynamic allocation of process page table space very awkward, since memory tends to fragment into page-size areas. Unfortunately, it appears that this part of the logic is implemented in the chip itself instead of on the board. NOW I understand the problem and what must be done to make VMS run on a KA620. I still think I can do it, but it will be a big job. The question is, is it worth it? Alan ================================================================================ Note 17.2 KA620 2 of 16 EISNER::MCCARTHY 14 lines 20-MAY-1987 17:09 -------------------------------------------------------------------------------- > I still think I can do it, but it will be a big job. > The question is, is it worth it? > > Alan We hope not. We spent a good deal of money, obviously, to make sure that it wasn't. Remember, each process, at creation, will require a static allocation of a page table of size VIRTPAGECOUNT, which is going to eat physical memory. -Brian ================================================================================ Note 17.3 KA620 3 of 16 EISNER::FRISBIE "Alan E. Frisbie" 18 lines 21-MAY-1987 00:45 -< Today's special on Page Tables: $235 >- -------------------------------------------------------------------------------- >> Remember, each process, at creation, will require a static >> allocation of a page table of size VIRTPAGECOUNT, which is >> going to eat physical memory. The current SYSGEN parameters on my VAXstation are VIRTPAGECOUNT = 18288 and MAXPROCESSCNT = 18. These look small, but are what Autogen set. Using these figures, we need (18288/128) * 18 pages for our page tables. This is 2572 pages or 1.26 MB. With 3rd party 8 and 16 MB memory boards going for less than $187/MB, I think that paying $235.00 for my page tables is reasonable. This price will only go down. Of course, if you buy DEC memory, your price will be higher. Much higher. :-) Now all I need is a KA620 to borrow for a few weeks. Alan ================================================================================ Note 17.4 KA620 4 of 16 EISNER::PROVOST 21 lines 21-MAY-1987 09:59 -< *128! >- -------------------------------------------------------------------------------- < < Note 17.3 by EISNER::FRISBIE "Alan E. Frisbie" > < -< Today's special on Page Tables: $235 >- < The current SYSGEN parameters on my VAXstation are < VIRTPAGECOUNT = 18288 and MAXPROCESSCNT = 18. These < look small, but are what Autogen set. Using these < figures, we need (18288/128) * 18 pages for our page <ÿ tables. This is 2572 pages or 1.26 MB. With 3rd < party 8 and 16 MB memory boards going for less than < $187/MB, I think that paying $235.00 for my page tables < is reasonable. This price will only go down. < Of course, if you buy DEC memory, your price will be < higher. Much higher. :-) < < Now all I need is a KA620 to borrow for a few weeks. < Alan UNITS FOR VIRTPAGECOUNT ARE PAGES, NOT WORDS! Multiply your neÿûed by 128. tjp ================================================================================ Note 17.5 KA620 5 of 16 EISNER::FRISBIE "Alan E. Frisbie" 16 lines 21-MAY-1987 12:24 -< Memory calculations >- -------------------------------------------------------------------------------- >> UNITS FOR VIRTPAGECOUNT ARE PAGES, NOT WORDS! >> Multiply your need by 128. I think I got it right the first time. One virtual page requires one Page Table Entry (PTE), or 4 bytes. One page (512 bytes) of the Page Table will hold 128 PTE's (512/4). Therefore, 18288 virtual pages of process space requires 18288 PTE's, or 18288/128=143 pages in the Page Table. 143 Pages per process, times 18 processes is 2574 total pages needed for the Page Tables. Did I screw up somewhere? Alan ================================================================================ Note 17.6 KA620 6 of 16 EISNER::COAR "Ken A L Coar - OPUS/OPS" 15 lines 21-MAY-1987 12:30 -< Correct, but.. >- -------------------------------------------------------------------------------- > VIRTPAGECOUNT = 18288 and MAXPROCESSCNT = 18. These > look small, but are what Autogen set. Using these > figures, we need (18288/128) * 18 pages for our page > tables. This is 2572 pages or 1.26 MB. > UNITS FOR VIRTPAGECOUNT ARE PAGES, NOT WORDS! He's dividing by 128 because that's how many virtual pages are mapped by a single page of page-table. So, his result is correct - for the amount of memory needed for his page tables. However, he DOES need to multiply by 128, and add to the previous figure, if he wants the processes to actually have the memory the page tables are so busily talking about.. #k ================================================================================ Note 17.7 KA620 7 of 16 EISNER::MCCARTHY 8 lines 21-MAY-1987 17:33 -------------------------------------------------------------------------------- > However, he DOES > need to multiply by 128, and add to the previous figure, if he wants > the processes to actually have the memory the page tables are so > busily talking about.. Not really. The process space is stil virtual. -Brian ================================================================================ Note 17.8 KA620 8 of 16 EISNER::FRISBIE "Alan E. Frisbie" 9 lines 21-MAY-1987 17:51 -< RSX has the power! >- -------------------------------------------------------------------------------- I still don't understand why some clever VMS hacker hasn't already done this. Are they just waiting for an RSX hacker to show them how? By the way Brian, since when did you get involved with VAXen? Are you trying to show them how to do multiprocessing right? Alan ================================================================================ Note 17.9 KA620 9 of 16 EISNER::CETRON 11 lines 21-MAY-1987 18:06 -< but who would want vms??? >- -------------------------------------------------------------------------------- if allan is correct, and I know that the virtual memory part is correct, and given most of the configurations i set up have 8-9 plus mbytes, sounds like there should be no problem running vms on the 620. if true, dec has a real problem on their hands (assuming that they really do care that the 620 doesn't run vms). but the important problem is will it run rt-32??? :-{) -ed ================================================================================ Note 17.10 KA620 10 of 16 EISNER::FRISBIE "Alan E. Frisbie" 5 lines 21-MAY-1987 19:25 -< RT-32 on the KA620 >- -------------------------------------------------------------------------------- >> but the important problem is will it run rt-32??? The KA620 will run RT-32 just as well as any other VAX. Alan ================================================================================ Note 17.11 KA620 11 of 16 EISNER::HASSINGER "Bob Hassinger" 9 lines 22-MAY-1987 09:40 -< One little question... >- -------------------------------------------------------------------------------- There might be just one tiny little problem with patching up VMS to run on the KA620 if you wanted the effort to be more than an academic exercise. I should think you might need a copy of VMS that was licensed for use on the CPU in question. Just how would people who wanted to put up one of these systems handle this issue? Bob H ================================================================================ Note 17.12 KA620 12 of 16 EISNER::CETRON 5 lines 22-MAY-1987 16:29 -< ..and one little answer >- -------------------------------------------------------------------------------- does the microvms license really say ka630? or does it just day microvax?? this could be the loophole... ================================================================================ Note 17.13 KA620 13 of 16 EISNER::HASSINGER "Bob Hassinger" 3 lines 22-MAY-1987 17:06 -< how much? >- -------------------------------------------------------------------------------- How much does a microVMS license cost? I thought that recent price changes had shifted a good bit more cost into the OS - the license may cost as much as the KA620! ================================================================================ Note 17.14 KA620 14 of 16 EISNER::FRISBIE "Alan E. Frisbie" 15 lines 23-MAY-1987 01:14 -< License to do the impossible? >- -------------------------------------------------------------------------------- >> How much does a microVMS license cost? $2000 for a microVAX-II, $1000 or $1500 (depending on part number, which I don't understand) for a VAXstation-II I looked at the SPD and it only says "MicroVAX-II", not KA630 or anything like that. However, the brochure for the KA620 never calls it a MicroVAX, as such. This could get interesting. I wonder what the local office would say if I called up and asked for the price of microVMS for a KA620? Since "it can't run on a KA620", the price should be real cheap. Alan ================================================================================ Note 17.15 KA620 15 of 16 EISNER::RICE "Gary Rice" 8 lines 23-MAY-1987 10:46 -< Hyroglifics >- -------------------------------------------------------------------------------- > I looked at the SPD and it only says "MicroVAX-II", not > KA630 or anything like that. However, the brochure for > the KA620 never calls it a MicroVAX, as such. I have been following this conversation with keen interest. I have a question, though: Just what IS a KA620? Gary ================================================================================ Note 17.16 KA620 16 of 16 EISNER::FRISBIE "Alan E. Frisbie" 21 lines 23-MAY-1987 23:56 -< Hyroglifics decoded >- -------------------------------------------------------------------------------- >> Just what IS a KA620?? First, it is necessary to know that the CPU board in the microVAX-II and VAXstation-II is called the KA630. Next, DEC will NOT sell you a KA630 board to put in your imbedded application -- they fear that someone will build a "system" around it and undercut DEC's prices. (Who would want to do such a dastardly act?) So..., DEC took the exact same board (I think), modified the CPU chip slightly and said, "Here is the board for your real-time imbedded systems. Oh, by the way, it won't run VMS. That shouldn't bother you, however, since you ARE going to plug it into an imbedded application and not sell it as a general purpose computer, aren't you?" That is a KA620. Alan  ================================================================================ Note 18.0 DOCUMENTATION ERRORS - HARDWARE 4 replies EISNER::KILLEEN 1 line 5-JUN-1987 23:48 -------------------------------------------------------------------------------- This topic will be used to report hardware documentation errors ================================================================================ Note 18.1 DOCUMENTATION ERRORS - HARDWARE 1 of 4 EISNER::KILLEEN 32 lines 6-JUN-1987 00:23 -< RD54 DOC ERRORS >- -------------------------------------------------------------------------------- The RD54-D,-R FIXED DISK DRIVE SUBSYSTEM OWNER'S MANUAL (EK-LEP07-OM-001) as several errors. In refering to the rear panel on the RD54-R,-D drives the manual states on page 3-6 "The J1 connector responds to DRV SEL 3 (L) signal on the drive interface cable from the host. When you use J1, the system software labels the RD54 as DU0 (disk unit 0). Use J1 only when are installing an RD54 subsystem with a PDP-11/23 PLUS systems. Appendix A lists DU (disk unit) designations and possible logical unit numbers (LUNs) on the RQDX3 modules. The J2 connector responds to a DRV SEL 4 (L) signal on the drive interface cable from the host. When you use J2, the system software lables the RD54 as DU1. Appendix A list......" The J1 and J2 connectors DO NOT cause the drive to be designated DU0 or DU1. The designation for the drive will be a function of the jumper settings on the RQDXE module and the type and number of drives you have in the system. The J1 connectors can be used with ALL PDP-11 and uVAX systems and NOT just with a PDP-11/23 PLUS system. The use of the J1 and J2 connectors is determined by the jumper settings on the RQDXE module For more information please refer to the RQDXE EXPANDER MODULE USER'S GUIDE. (EK-LEP07-OM-001) ================================================================================ Note 18.2 DOCUMENTATION ERRORS - HARDWARE 2 of 4 EISNER::KILLEEN 18 lines 6-JUN-1987 00:41 -< RDn ordering info >- -------------------------------------------------------------------------------- The ordering menus for both the MICRO PDP-11 and MICRO VAX in the systems and option catalogs imply that you can only place a limited number of RDn drives on a system. For example the PDP-11 systems and option catalog on page 1.62 list add-on mass storage options for the MicroPDP-11/73 BA23 standard system. Under item 6 it lists the add-on RD54, RD53, and RD52's and has a note that states "choose only one". Two different local DEC salesman rejected orders for dual RD54-DA's on this system because of the menu. They did not understand the menu meant chose only one RQDXE-AA/RD54 combination NOT only one RD54 drive. Refer your saleman to page 5.6 in the PDP-11 catalog and PAGE 5.20 in the VAX catalog. You can have up to 3 RDn disk drives with a BA23 box, one internal, two external, as long as you do not have an RX50 or dual RX33's on the system. It took me a week to get DEC to place the order! ================================================================================ Note 18.3 DOCUMENTATION ERRORS - HARDWARE 3 of 4 EISNER::RICE "Gary Rice" 10 lines 6-JUN-1987 08:19 -< Beware the nasty Electrons >- -------------------------------------------------------------------------------- > the add-on RD54, RD53, and RD52's and has a note that states "choose > only one". Two different local DEC salesman rejected orders for > dual RD54-DA's on this system because of the menu. They did not One word of caution regarding these puppies: The RD54 is a power hog. Our DEC people say that while the BA123 box has room for 5 devices (and policy says 4), putting 4 drives in could each power supplies alive. Gary ================================================================================ Note 18.4 DOCUMENTATION ERRORS - HARDWARE 4 of 4 EISNER::KILLEEN 9 lines 6-JUN-1987 17:37 -< NEVER A POWER PROBLEM WITH -DA and -RA DRIVES >- -------------------------------------------------------------------------------- The RD54 may be a power hog but the RD54-DA and RD54-RA have their own power supplies. All RDxx-DA and RDxx-RA drives do. There will be no power problems with these drive when used externally on a BA23 or BA123 box. Note: you can use external drives with a BA123 as long as you purchase another RQDX3 controller and a RQDXE module.  ================================================================================ Note 19.0 MODEM AND COMMUNICATION EQUIPMENT 60 replies EISNER::KILLEEN 3 lines 10-JUN-1987 10:12 -------------------------------------------------------------------------------- This topic is to be used for a discussion of modems and other related communication equipment. ================================================================================ Note 19.1 MODEM AND COMMUNICATION EQUIPMENT 1 of 60 EISNER::KILLEEN 41 lines 10-JUN-1987 10:38 -< BASIC MODEM STUFF >- -------------------------------------------------------------------------------- Some basic modem information. 1. When installing modems to a DEC computer you should never force any of the standard pins on. Some modems allow you to force such things as CD, DSR, RI, and DTR on. The modem will not sequence properly with these forced on. 2. Only use a 10 wire cable unless you know a reason not to. Connect pin 1-8, 20 & 22. 3. The sequence for answering a call is as follows: 1. The lines rings and the modem ethier flashs or holds the Ring Indicator (RI, pin 22) on. 2. The computer turns on Data Terminal Ready (DTR, pin 20). 3. The modems turns on Data Carrier Detect (DCD, pin 8) and optionally Clear To Send (CTS, pin 5). 4. Depending on what type of interface you have the computer ethier turns on Request To Send (RTS, pin 4) or jumper RTS to CTS. 4. The sequence for hanging up a call is as follows: 1. The computer drops DTR and optionally RTS. 2. The modem hangs up and drops DCD and optionally CTS. OR 1. The modem drops DCD and optionally CTS. 2. The computer drops DTR and optionally RTS. This is the BELL 212A handshake standard. ================================================================================ Note 19.2 MODEM AND COMMUNICATION EQUIPMENT 2 of 60 EISNER::CETRON 11 lines 10-JUN-1987 16:26 -< but beware the state machine bugaboo.... >- -------------------------------------------------------------------------------- note that under vms (which has somewhat brain-damaged modem control) that if you don't raise dsr at all, you can still log in, an do all you want, BUT VMS WILL NOT TREAT THE LINE AS A MODEM, even though it says it will. Once you raise dsr, then the other lines MUST sequence correctly or it will cycle dtr. I have a long involved exposition which I will gladly mail to anyone via arpanet, bitnet or dcs, but I can't upload it here so, sorry. -ed ================================================================================ Note 19.3 MODEM AND COMMUNICATION EQUIPMENT 3 of 60 EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sa" 21 lines 11-JUN-1987 21:43 -< Kill that process ! >- -------------------------------------------------------------------------------- I am having problems getting the process killed when someone (or line noise) drops the phone line. Seems like the process just stays around forever which is a slight security problem. What terminal characteristics (VMS Set terminal command) do I need and what is the minimum number of pins I need connected on my RS232 ? Do any of them need to be crossed or jumpered ? Additionally, I am NOT intertested in setting up virtual term- inal ports and I am connected to a DZ11 port which I assume has modem control. The modem I am using, although it should not matter too much is a VENTEL 1200-31 and I do NOT have the dip switch so that DTR is always high. I have six wire cable so if I can do it with six wires it would be nice. Hopefully this is a common question. - Bert ================================================================================ Note 19.4 MODEM AND COMMUNICATION EQUIPMENT 4 of 60 EISNER::KILLEEN 19 lines 11-JUN-1987 22:52 -< BEWARE OF EARLY VENTEL 1200 MODEMS >- -------------------------------------------------------------------------------- re: note 19.3 Ventel 1200 modems (which we used to use) are known for not following the 212A standard. You may have CDC or RI force high and not know it. Get a Breakout box that has an LED for each pin (at least pins 1-8,20,&22). Connect your modem to the computer through it. Watch the lights and see if they follow the right sequence (see note 19.1). A 6 wire cable cannot work!!!!!! The absolute minimum is 8. Connect pins 1-3, 6-8, 20, & 22. I do not recommend this cable. It is better to go with a 10 wire cable. Connect pins 1-8, 20, & 22. The fact you are using a 6 wire cable proves you have pins forced on!!! ================================================================================ Note 19.5 MODEM AND COMMUNICATION EQUIPMENT 5 of 60 EISNER::TINKELMAN "Bob Tinkelman" 121 lines 12-JUN-1987 11:58 -< Modem leads - a rambling discourse >- -------------------------------------------------------------------------------- re: < Note 19.4 by EISNER::KILLEEN > < < A 6 wire cable cannot work!!!!!! The absolute minimum is 8. < Connect pins 1-3, 6-8, 20, & 22. I do not recommend this cable. < It is better to go with a 10 wire cable. < Connect pins 1-8, 20, & 22. < < The fact you are using a 6 wire cable proves you have pins forced < on!!! I beg to differ - at least with "cannot work". But first a digression: -------------------------- DIGRESSION -------------------------- HOW MANY PINS ARE ASSOCIATED WITH ASYNCH? First a comment on your recommendation for 10 wire pinning. If you have 10 wire cable, then I agree with your recommended pinning. However, do not be deceived into thinking that these are all the pins that might be reasonable used in asynch communications. For example, pin 25 is often used between modems and DTE to mean "busy out the line". AND WHAT ABOUT SYNCH? If your are doing a large installation or if you are concerned on standardizing on one type of cable, you might want to pay a little extra to allow for future synchronous (vs asynch) uses. When we had EIA cables installed during the (re)construction of our offices, we chose to include 15 and 17 also so that we could support most synch applications also. (Not to a DZ of course.) ---------- THE REAL POINT - WHAT CAN I GET AWAY WITH? ---------- PIN 1 - I CAN ELIMINATE FRAME GROUND There are two "grounds" - pins 1 and 7. Pin 1 is designated "frame ground", "chassis ground", or something similar. It's supposed to be tied to the real ground. Pin 7 is called "signal ground" (SG). It is simply a reference voltage against which other signals (eg. Transmit Data on pin 2) is measured. Most modems either totally ignore pin 1 or jumper it to pin 7 right at the EIA connector. Hence it's safe to eliminate it. PINS 4 AND 5 - YOU DON'T NEED RTS/CTS Pins 4 and 5 (request and clear to send, respectively - RTS and CTS) were designed for use in a half duples (or multidrop) environment where the modem was not always putting its own carrier signal on the line. In the normal "point to point" configurations, this functionallity is not needed. If the DTE insists on seeing CTS before sending data, the typical solution is to "jumper" pins 4 and 5 at the DTE end of the cable. However, there is no need to bring these signals from one end to the other. [Don't confuse the above "traditional" rs232 use of pins 4 and 5 with a bastardization which uses them for "hardware flow control".] PIN 22 - YOU DON'T NEED RING INDICATOR. While a full modem-answer interaction goes like this: 0. Modem is powered on and indicates that by holding DSR (data set ready) high (i.e. "on"). 1. Modem sees ring voltage on phone line and asserts RI (ring indicator) to DTE (i.e. the computer) 2. DTE sees RI, decides it's OK to answer the phone, and sets DTR (data terminal ready) to modem. 3. Modem sees DTR and answers the phone. Modem does its magic with the calling modem and, when it sees that modem's carrier signal, raises CD (carrier detect) to the DTE. in a (normal) situation, where the DTE will always decide that it's OK to answer the phone, the DTE (i.e. the VAX) can leave DTR up while the modem is waiting for incoming calls. The modem may assert RI when the call comes in, but the computer doesn't have to see it (as it's already raising DTR). This eliminates the need for pin 22. PINS 6 AND 8 - YOU REALLY NEED ONLY ONE OF THEM. Pin 6 is DSR (data set (i.e. modem) ready). Pin 8 is CD (carrier detect). Both are signals from the DCE to the DTE. In concept they have very different purposes. From the point of view of the DTE, DSR means that the modem is on, functioning, and ready to respond to changes the DTE may make on any of the DTE to DCE modem leads. DCD means that the modem can hear a carrier signal on the telephone wire and that it's valid for the DTE to look at incoming data on pin 3 (receive data). However, in configuration (dial-up to a computer) it works to simply take pin 8 (DCD) from the modem and to jumper 6 to 8 at the DTE end of the cable. Under DEC's (now standard) hadling of modem leads, the VAX, is perfectly happy to respond to DCD and DSR both going high simultaneously, even in the absence of RI. ------------------------- CONCLUSIONS ------------------------- You only need 5 wires, a ground, a data lead in each direction, and a modem lead in each direction: DCE=modem DTE=computer SG 7-------------------------7 FG 1 1 TD 2<------------------------2 RD 3------------------------>3 RTS 4 ,----4 CTS 5 `--->5 DSR 6 _________________,--->6 CD 8--' `--->8 DTR 20<-----------------------20 RI 22 22 AND NOT EVEN 5 REAL WIRES In one instance, where the cables were already in the floor and would have cost thousands of dollars to replace, we used "shielded quads" - using the drain wire for pin 7 and the four real wires as shown above. ================================================================================ Note 19.6 MODEM AND COMMUNICATION EQUIPMENT 6 of 60 EISNER::KILLEEN 55 lines 12-JUN-1987 16:12 -< HE WHO JUMPERS MAY FIND LATER HE WISHED HE DID NOT >- -------------------------------------------------------------------------------- < A 6 wire cable cannot work!!!!!! The absolute minimum is 8. < Connect pins 1-3, 6-8, 20, & 22. I do not recommend this cable. < It is better to go with a 10 wire cable. < Connect pins 1-8, 20, & 22. < < The fact you are using a 6 wire cable proves you have pins forced < on!!! >>>I beg to differ - at least with "cannot work". >>>You only need 5 wires, a ground, a data lead in each direction, >>>and a modem lead in each direction: >>> >>> DCE=modem DTE=computer >>> >>> SG 7-------------------------7 >>> FG 1 1 >>> TD 2<------------------------2 >>> RD 3------------------------>3 >>> RTS 4 ,----4 >>> CTS 5 `--->5 >>> DSR 6 _________________,--->6 >>> CD 8--' `--->8 >>> DTR 20<-----------------------20 >>> RI 22 22 >>>AND NOT EVEN 5 REAL WIRES I BEG TO DIFFER WITH YOU!!! The example you show is not a true 5 wire cable. You have heavily jumpered and crossed this cable. While I sure it will work with a VAX and the right modem it will not allow for any deviations from your assumptions. I know it will not work with RSTS or any other DTE device that does not raise DTR until it see RI. >>>In one instance, where the cables were already in the floor and >>>would have cost thousands of dollars to replace, we used >>>"shielded quads" - using the drain wire for pin 7 and the four >>>real wires as shown above. I assume the real point of your note is to say if you can't rewire then there are ways around the problem. I think you would agree with me that it is a good practice to use 10 wire cables if you can. >>>If your are doing a large installation or if you are concerned >>>on standardizing on one type of cable, you might want to pay a >>>little extra to allow for future synchronous (vs asynch) uses. >>>When we had EIA cables installed during the (re)construction of >>>our offices, we chose to include 15 and 17 also so that we could >>>support most synch applications also. (Not to a DZ of course.) I AGREE ================================================================================ Note 19.7 MODEM AND COMMUNICATION EQUIPMENT 7 of 60 EISNER::WALLIS "Barry Wallis - Fleetwood Enterprise" 21 lines 12-JUN-1987 16:25 -< Re 19.3: Making Processes Go Away >- -------------------------------------------------------------------------------- >>> < Note 19.3 by EISNER::ROSEBERRY "Bert Roseberry-that Oak among Saplings" > >>> -< Kill that process ! >- >>> >>> >>> I am having problems getting the process killed when someone >>> (or line noise) drops the phone line. Seems like the process >>> just stays around forever which is a slight security problem. We have encountered similar problems and have solved it in two ways: 1) We have MICOM modems so this may not help much: We have to set our modems (on both sides) so that they disconnect on recieving space. On MICOMs you do this via software settings rather than dip switches. 2) This may be all it takes: Make sure your terminals are set to be /DISCONNECT and /HANGUP. When we did the above two things the problem went away. ================================================================================ Note 19.8 MODEM AND COMMUNICATION EQUIPMENT 8 of 60 EISNER::HASSINGER "Bob Hassinger" 19 lines 12-JUN-1987 18:01 -< How about the "speed control" pins? >- -------------------------------------------------------------------------------- < Connect pins 1-8, 20, & 22. I have been learning there may be more to the story even for "simple" async. I have been moving our modems over to a DECserver 200. They happen to be AT&T 2224Bs that support 300, 1200 and 2400. Besides the usual control lead issues as noted before I now find I have speed control leads to worry about. The modems have "CI" on 12 and "CI2" on 13 that indicate the speed the modem has set up at as well as "CH2" on 19 that can control speed selection on originate. All these change the way they work depending on various option switch settings. The DECserver supports a primary and a secondary speed that can be controlled via pin 12 ("CI"). When all is said and done it looks like I have to connect pin 12 on the server to pin 13 on the modem because the manual seems to say that is the only way to get one pin from the 2224B that signals 1200/2400. I am beginning to think this multispeed 1200/2400 thing may not be very well standardized yet. What have you all found? ================================================================================ Note 19.9 MODEM AND COMMUNICATION EQUIPMENT 9 of 60 EISNER::KILLEEN 20 lines 12-JUN-1987 19:35 -< AVOIDING SPEED PIN ALL TOGETHER >- -------------------------------------------------------------------------------- >>> I have to connect pin 12 on the server to pin 13 on the modem because >>> the manual seems to say that is the only way to get one pin from the >>> 2224B that signals 1200/2400. I am beginning to think this multispeed >>> 1200/2400 thing may not be very well standardized yet. >>> What have you all found? We went through a long search for a the 300/1200/2400 baud modems. We finally decided on Black Box MD797-786 DIAL MODEM 24+. The reason why was because these modems will auto speed sense yet keep the modem to interface speed at 2400. So if you dial into a 1200 baud modem then Black Box translates the speed back to 2400 baud for the interface. The bigest problem we had with speed pins was their absence from most DEC interfaces. We sent back a bunch of DEC DF124 modems due to this. The DF124 requires you have a speed pin or someone to throw the speed switch. This does not work well with a DZ11 on a unattended dialout modem. The Black Box dials out and will automatically connect the target modem at it's highest speed. ================================================================================ Note 19.10 MODEM AND COMMUNICATION EQUIPMENT 10 of 60 EISNER::TINKELMAN "Bob Tinkelman" 34 lines 12-JUN-1987 19:44 -< Avoiding speed-select issues >- -------------------------------------------------------------------------------- Re: < Note 19.8 by EISNER::HASSINGER "Bob Hassinger" > -< How about the "speed control" pins? >- < I am beginning to think this multispeed < 1200/2400 thing may not be very well standardized yet. < What have you all found? Regarding standardization, I agree. The problem is at least in part due to the fact that when "speed select" was first used, it meant 1200 vs less-that-or-equal-to-300. In addition, it appears that originally, only one pin was designated - to be used either for DTE-to-DCE or vice versa, strap selectable or by choice of modem manufacturer. I've also seen various choices of pins for the "other direction" speed select pin. As a result of the above, we've avoided using modem lead speed- select entirely. At this time, all our modems are capable of "buffering traffic" and speaking different bit rates on their analog and digital sides. (This is actually an easy thing for the manufacturer to implement once doing something like MNP.) The modem connected to our local data switch and used for dial- out from our office is set at 9600 bps on the digital side. It will automatically talk the right analog speed when calling a 300, 1200 or 2400 bps modem at the other side, whether or not in MNP mode. However, the digital side stays 9600. Similarly, the modem connected to the terminal I'm using right now from home is always set to talk to the VT220 at 4800 bps (as I found this less bursty than 9600 in MNP mode). Using these buffered modems allows us to totally sidestep the "speed select" and autobaud issues with only some minor problems related to the flow control. ================================================================================ Note 19.11 MODEM AND COMMUNICATION EQUIPMENT 11 of 60 EISNER::KILLEEN 16 lines 12-JUN-1987 19:57 -< CURIOUS ABOUT MNP >- -------------------------------------------------------------------------------- >>> It will automatically talk the right analog speed when calling a >>> 300, 1200 or 2400 bps modem at the other side, whether or not in >>> MNP mode. However, the digital side stays 9600. Similarly, the >>> modem connected to the terminal I'm using right now from home is >>> always set to talk to the VT220 at 4800 bps (as I found this >>> less bursty than 9600 in MNP mode). >>> Using these buffered modems allows us to totally sidestep the >>> "speed select" and autobaud issues with only some minor >>> problems related to the flow control. Have you tried doing system to system binary file transfers with your modems in MNP mode?. If so what happens with XON/XOFF flow control. You can get char 17. and char 19. in a binary transfer. ================================================================================ Note 19.12 MODEM AND COMMUNICATION EQUIPMENT 12 of 60 EISNER::TINKELMAN "Bob Tinkelman" 19 lines 12-JUN-1987 22:05 -< Kermit, MNP and ^S/^Q >- -------------------------------------------------------------------------------- >>> re: < Note 19.11 by EISNER::KILLEEN > >>> Have you tried doing system to system binary file transfers with >>> your modems in MNP mode?. If so what happens with XON/XOFF flow >>> control. You can get char 17. and char 19. in a binary transfer. If ^S and/or ^Q appear inside a file being transfered, then everyone along the path needs to be set to "no flow control" or (if they support it) to "modem lead flow control". With our modems, we simply turn off all flow contrl. We don't seem to get into trouble with overflowing any buffers along the way, even with speeds as different as 1200 analog vs 9600 digital. I believe that what's saving us is that the higher level file transfer protocols (eg Kermit) have relatively small "window sizes". That is, the total amount of data that they will send before receiving acknowledgment from the other end is small enough so that there is never enough data "in the pipe" to overflow any buffer. ================================================================================ Note 19.13 MODEM AND COMMUNICATION EQUIPMENT 13 of 60 EISNER::DEEM 3 lines 20-AUG-1987 19:46 -< Trailblazer by Telebit >- -------------------------------------------------------------------------------- Does anyone have any experience with the Trailblaizer by Telebit? We are looking for a means of transferring CAD drawing files between installations at some speed greater than 1200/2400 baud. ================================================================================ Note 19.14 MODEM AND COMMUNICATION EQUIPMENT 14 of 60 EISNER::PRIGOT "Jonathan M. Prigot" 9 lines 8-SEP-1987 17:28 -< Dialup DECnet modems >- -------------------------------------------------------------------------------- Our DECnet data traffic to some sites was not enough to justify leased lines, so we configured a pair of MICOM 2400 bps dialup modems to run in Sync mode. Normally things worked fine except that when we tried to tranfer a file, the modems would hang up (as in drop the line). It seems that during the file transfer, some bit combinations would force the line to SPACE for a period of time. One of the options in the modem is to disconnect after receiving about 1.6 seconds of SPACEing, and that's what they were doing. Setting the appropriate DIP switch fixed the problem. ================================================================================ Note 19.15 MODEM AND COMMUNICATION EQUIPMENT 15 of 60 EISNER::OSTROSKY "Jim "Crash" Ostrosky" 65 lines 21-OCT-1987 08:11 -< TRAILBLAZER EXPERIENCE >- -------------------------------------------------------------------------------- > Does anyone have any experience with the Trailblaizer by Telebit? We > are looking for a means of transferring CAD drawing files between > installations at some speed greater than 1200/2400 baud. I have used the Telebit Trailblazer in both interactive mode and in dynamic, dialup, asynchronous DECnet. I have the modem connected to a DMF-32 modem port with a "standard" DEC modem cable (BC22F-25), and the port is set to 9600 baud, DIALUP, HANGUP, MODEM. The Telebit is capable of: o 300 baud (Bell 103 emulation) o 1200 baud (Bell 212A emulation) o 1200 baud (CCITT V.22 emulation) o 2400 baud (CCITT V.22bis emulation) o PEP mode (up to 18,000 BPS); hereafter either PEP or "fast" The modem is Hayes compatible, and the manual for setup is reasonably good; there are some *VERY* confusing parts, but we managed to get them straightened out. Some things we did: a) Used CTS/RTS flow control, not XON-XOFF flow control, between the modem and the DMF-32 port (watch the replies on this one!). Because we use this modem to talk to other modems that are NOT Trailblazers, the modem to modem speed is not always fast, but the DMF-32 port is ALWAYS 9600 baud. This requires some kind of flow control. Since we also use asynch DECnet on the same modem, we opted for hardward flow control; no problems since then. b) We disabled the use of the automatic speed determination (i.e. autobaud on dialing out). We may have done this too soon. When the modem was set to automatically determine the speed of the modem it was calling, it goes through one of the following sequences (depending on certain Hayes compatible registers): o PEP, V.22bis, 212A, 103 o PEP, V.22bis, V.22 with no guard tone o PEP, V.22bis, V.22 with 1800 hz guard tone o PEP, V.22bis, V.22 with 550 hz guard tone What we saw happening was that the modem on the other end hung up before the autobaud sequence got to the right speed! I ended up re-writing my autodial procedure so that the modem used the correct speed automatically. c) When talking to another Trailblazer, the "echoplex" delay (i.e., line turnaround for character echo) can be maddeningly slow, in interactive mode. Granted, the first "fast" modems were designed to go fast, in a packet transmission environment. What this means is, don't try editing with this modem; it can be done, but 1200 baud is less painful. I think covers most of what we did with the thing. I myself am not extremely happy with it (mostly because it was crammed down my throat, and I had no choice in the matter). There are some newer modems coming out designed more for the interactive user in mind. I have not run asynch DECnet over the Trailblazer very much, so I hesitate telling you anything. It did run, but no stats. Hope this helps; any other questions, either here or (412) 281-8988 x339 (except on 10/22 and 10/23; out of town). Jim ================================================================================ Note 19.16 MODEM AND COMMUNICATION EQUIPMENT 16 of 60 EISNER::SALLEE "Gary Sallee" 19 lines 14-NOV-1987 15:49 -< Dial out set speed AND auto baud >- -------------------------------------------------------------------------------- < We disabled the use of the automatic speed determination (i.e. > < autobaud on dialing out). > For a multi speed modem (Ditto 2400XL - Hayes compatible 2400 1200 300 - Rockwell board), I allow the modem to auto baud on dial out, but also allow the user to set the dial out speed, via command files. This avoids the other modem time out problem when the user knows what speed to expect. It still allows the modem to move the speed down when there are several possible modems on the answering line. I have also modified Kermit-11 to receive the Hayes extended response set, then automatically set the system serial port speed. This setting of the port speed is not needed with the Trailblazer. Gary S. ================================================================================ Note 19.17 MODEM AND COMMUNICATION EQUIPMENT 17 of 60 EISNER::HAHN "DECUServe Vice Chair" 6 lines 6-JUL-1988 23:22 -< A good modem >- -------------------------------------------------------------------------------- This is not a request to buy (ala swap_meet). I have just been reduced to 1200 baud; my personal 2400 baud modem just gave up. If you have to pay out of your own pocket which modem would you suggest I obtain. Pros and Cons ? Pierre ================================================================================ Note 19.18 MODEM AND COMMUNICATION EQUIPMENT 18 of 60 EISNER::KENNEDY "Terry Kennedy" 15 lines 7-JUL-1988 03:05 -< Suggestion for Hayes-compatible modem >- -------------------------------------------------------------------------------- Well, I don't buy [m]any stand-alone originate modems - most of my outbound modems are in PC's here on campus. We have standardized on the Everex EV-940 modem based on price/features/reliability. They do have an stand-alone version of the 940 available. It is Hayes 'AT' command set compatible. Note that this is pretty much a vanilla 2400 - Async, Sync, U.S. & Euro standards, but *no* MNP. Those of you who frequent the other conferences may notice my non-MNP stance. This modem is good enough for most cases even without MNP, and if you want error correction there are other ways of doing it. [Fuzzy disclaimer: At our in-house volume purchase level, Everex thinks we're a dealer. However, we do not resell; it is all internal use in the school; and besides, we buy so many because we like them.] ================================================================================ Note 19.19 MODEM AND COMMUNICATION EQUIPMENT 19 of 60 EISNER::TINKELMAN "Bob Tinkelman - CCA" 21 lines 7-JUL-1988 06:30 -< A little agout CDS224SeriesII modems >- -------------------------------------------------------------------------------- We use Concord Data System 224 Series II modems with MNP. I can't compare it in detail with other modems. We started using CDS when they were the only vendor with 2400 and just stayed with them, never feeling any good reason to change. Some miscellaneous positive points: 1. Flexible setup and commands. There is, in addition to the Hayes command mode, a CDS command mode which gives tremendous flexibility in modem setup. 2. Speed conversion. Many modems have this. I keep my home modem and VT220 set to 4800 - independent of what speed the other end answers. 3. Dial-time choice of MNP. In one setup mode you tell the modem to attempt MNP on a call by adding an L to the end of the dial string. This is convenient when calling some systems which do and others which don't support MNP. Negative points: 1. Sensitivity to electrical surges. With a population of about 20 modems, we send 2-3 a year back for repair, always after lightning storms. 2. Price. ================================================================================ Note 19.20 MODEM AND COMMUNICATION EQUIPMENT 20 of 60 EISNER::DOHERTY "Bob Doherty" 13 lines 7-JUL-1988 07:46 -< PP 2400SA's get our vote >- -------------------------------------------------------------------------------- We have started using Personal Peripheral 2400SA modems both for dialin and for autoanswer on our VAX, and have been very satisfied with them. They are sufficiently Hayes compatible that SmartCom thinks they're real Hayes modems. The only problem we've had so far(ca. 3 months), is that occasionally, after a power hit, the modems on our VAX go into an indeterminant state and will neither answer or dialout. Remedy -- power them down and up. For dialin they have been flawless, and seem to be more noise resistant than the modems they have replaced, Micom's, Racals and Rixons. They are available from a number of sources for under $200 and they carry a five year warranty. We got ours from MacConnection in NH. ================================================================================ Note 19.21 MODEM AND COMMUNICATION EQUIPMENT 21 of 60 EISNER::NAGY "Frank J. Nagy, VAX Guru" 14 lines 7-JUL-1988 08:00 -< Microcom AX2400c >- -------------------------------------------------------------------------------- We use Microcom AX2400C modems. These will handle 300, 1200 and pure 2400 baud as well as MNP up through class 5 (or is it 6). Worked very well for me when the DECUServe modems were talking MNP, now I just have to remember to disable auto-reliable mode (turn off MNP and use straight 2400) when talking to DECUServe. The AX2400C uses either a native Microcom command set (which I find myself using) or can be put in a Hayes-compatible mode. On advantage of the high level MNP class is that when talking to another AX2400C (we have a bank feeding into our Micom port selector) data compression is engaged for an effective (and believeable) 4800 baud performance. Normally we setup the modems with a 9600 baud link between the terminal and modem or the computer and the modem (usually this is autobaud connection). ================================================================================ Note 19.22 MODEM AND COMMUNICATION EQUIPMENT 22 of 60 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 10 lines 7-JUL-1988 08:20 -< Ditto .-1 No problems >- -------------------------------------------------------------------------------- > < Note 19.21 by EISNER::NAGY "Frank J. Nagy, VAX Guru" > > -< Microcom AX2400c >- >We use Microcom AX2400C modems. These will handle 300, 1200 and pure >2400 baud as well as MNP up through class 5 (or is it 6). Worked very >well for me when the DECUServe modems were talking MNP, now I just have >to remember to disable auto-reliable mode (turn off MNP and use straight >2400) when talking to DECUServe. AX2400c MNP class 5. Just under $800 from Glasgal. ================================================================================ Note 19.23 MODEM AND COMMUNICATION EQUIPMENT 23 of 60 EISNER::PRIGOT "Jonathan M. Prigot" 2 lines 7-JUL-1988 17:17 -< A.K.A. >- -------------------------------------------------------------------------------- CASE sells the Microcom line of modems with a few improvements at a cost comparable with their Microcom counterparts. ================================================================================ Note 19.24 MODEM AND COMMUNICATION EQUIPMENT 24 of 60 EISNER::KOZAM 16 lines 7-JUL-1988 19:37 -< Another vote for Concor >- -------------------------------------------------------------------------------- I've been using Concord Data Systems 224 modems (without MNP) for about 4 years and have, without question been very pleased. They support 1200 and 2400 baud (I don't care about 300 baud). Most importantly, they automatically switch between 1200 and 2400 both on dial-in AND dial out (earlier firmware revisions didn't autobaud on dial-out, but were inexpensively upgraded). They handle DTR, RI, CTS, RTS, etc. correctly. I used to take this for granted until I got some DEC DF124, which make quite a mess in all these areas. I think they're the best and at between $ 300 - $ 400, they're also a great value. I haven't seen any unusual failures, but then again, they're on a pretty clean power source. Marc Kozam ================================================================================ Note 19.25 MODEM AND COMMUNICATION EQUIPMENT 25 of 60 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 6 lines 7-JUL-1988 20:02 -< Vadic VA2400-VP >- -------------------------------------------------------------------------------- I am very happy with my Vadic VA2400-VP. My early unit has a fairly low level of MNP, but they have probably improved that by now. It is super-reliable and features both Hayes and Vadic (which I prefer) command sets. Alan ================================================================================ Note 19.26 MODEM AND COMMUNICATION EQUIPMENT 26 of 60 EISNER::ERSKINE "Chris Erskine" 4 lines 8-JUL-1988 06:44 -< Vadic >- -------------------------------------------------------------------------------- I second Alan's comment. We also have the rack mount modems which have not had many problems (out of about 20). I like it much better than the DEC Scholar that I am having to suffer with at work on my PC. ================================================================================ Note 19.27 MODEM AND COMMUNICATION EQUIPMENT 27 of 60 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 5 lines 18-JUL-1988 16:56 -< Another vote for Vadic >- -------------------------------------------------------------------------------- Another vote for Vadic. We have both 1200 and 2400 bps rack mount and some standalone 2400 bps units. The price is a bit higher than other but I've never had a problem with any of them (knock wood) ! ================================================================================ Note 19.28 MODEM AND COMMUNICATION EQUIPMENT 28 of 60 EISNER::LEDERMAN "Bart Z. Lederman" 14 lines 18-JUL-1988 19:21 -< ZOOM modem MX sort of o.k. >- -------------------------------------------------------------------------------- I just got a ZOOM Telephonics MX modem (2400 baud). Good points are low cost (under $200, I paid $150 at PC Expo), 30 day trial, 2 year warranty, compact size, and, according to several PC magazines, an instruction set so close to Hayes' that it will run their Smartcom software. Bad points are it's rather stupid about dialing. It sometimes fails to recognize dial tones, thinks ringing is busy, etc. It is also sometimes slow to connect, and on the local Tymnet 2400 baud only number will sometimes connect at 1200 (the result being totally unuseable). On the other hand, once it does connect it does adaptive equalization, and I get NO noise at all (even though it has no MNP). ================================================================================ Note 19.29 MODEM AND COMMUNICATION EQUIPMENT 29 of 60 EISNER::LEDERMAN "Bart Z. Lederman" 5 lines 10-AUG-1988 20:24 -< ZOOM down, Practical Peripherals up. >- -------------------------------------------------------------------------------- After getting fed up with the ZOOM modems' stupidity about dialing (and connecting at the wrong speed) I talked to the manufacture, who sent me a second one. It was a little better, but not much. So it went back for a refund and I bought a Practical Peripherals 2400SA and am very happy with it. ================================================================================ Note 19.30 MODEM AND COMMUNICATION EQUIPMENT 30 of 60 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 18 lines 11-AUG-1988 00:10 -< Not a happy Racal Vadic user >- -------------------------------------------------------------------------------- We're using some Racal-Vadic 2400PA modems. I was just at DEC's Santa Clara training center - we drove around some and when we passed Racal Vadic's office, I lowered my window and, ummm.... abruptly ejected some saliva on their lawn. I'm tempted to return the modems to Racal Vadic in very small pieces (a fine powder?). These modems have been a royal pain. They intermittantly fail to connect, connect and do nothing, die in the middle of dialing, spew garbage, and do stuff like the following: * command to dial a number DIALING... BUSY! (after a suitable delay - no problem here) * command to dial another number BUSY! Immediately! * same command again (dial second number) works fine ARRGGGHHHH! Wish I could be rid of them.... ================================================================================ Note 19.31 MODEM AND COMMUNICATION EQUIPMENT 31 of 60 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 54 lines 23-SEP-1988 17:34 -< Help me understand modem control (once again) ! >- -------------------------------------------------------------------------------- We have just migrated from a 780 to an 8700 (big step, but we needed it) and now I'm having modem problems when I try to dial out ! First, on the 780, I had DZ11's connected to Racal-Vadic 2400 PA's and had TTY_DIALTYPE=1 and every thing was fine. Now, on the 8700, I have DHB32's (they are supposed to be identical to DHU11's) and still have TTY_DIALTYPE=1. When inactive, the Racal-Vadic modem has DSR high, CTS low, CD low, RI low and the DHB32 has DTR high and RTS high, which is what I would expect. On the "A" DHB32 (ie. TXAn) DTR and RTS are dropped every 37 seconds and the modem turns off DSR until DTR comes back !!! If your trying to do a SET HOST/DTE (or KERMIT or HOST32) this causes VMS to send your process a "Data Set Hangup Error" !!!!! On the "VMS side", I have the port set for 2400 bps, noautobaud, modem and dialin. Now if you look at the picture in the I/O User's Guide Part I (page 8-15 for V4 doc, page 8-14 for V5, same picture) which is a flow chart of how modem control is "supposed to work", it actually make sense (but why they would want to do this I don't know). Now what gets crazier is that on the "B" controller (ie. TXBn) which is connect to another brand of modem that has DSR low, CTS high, CD low and RI low, none of this craziness happens !!!!!! Now before you take that flow chart as "gospel" there are at least 4 know errors. 1) The right hand exit from the "delay 2 sec" box should say "TTY_DIALTYPE Bit 0 = 1". 2) The bottom exit from the "start 30 sec timer" box should say "reference count > 0". 3) The path going around the "start 30 sec timer" box should say "TTY_DIALTYPE Bit 2 = 1". 4) The left most exit from the "Transmit1" box should say "TTY_DIALTYPE Bit 1 = 1". Bottom line here: WHAT THE *^&%#*&^%)*&^ DO I DO SO THAT I CAN USE MY MODEMS FOR DIAL OUT ! (P.S. Dial in seems to be working just fine !!!!) (P.P.S. TTY_DIALTYPE Bit 1 = 1, says hangup immediately if I loose carrier instead of waiting for 2 seconds.) ================================================================================ Note 19.32 MODEM AND COMMUNICATION EQUIPMENT 32 of 60 EISNER::KENNEDY "Terry Kennedy" 18 lines 23-SEP-1988 19:43 -< Try these... >- -------------------------------------------------------------------------------- Check the cables - DZ11's did *not* generate/respond to the full set of DEC 'standard' modem control signals. The 30+-second burps generally only hapen if something is cabled or pinned wrong. Most DEC stuff is standard enough that a 25-pin ribbon cable between the modem and the host port should work. Late flash: You can also get this problem from a *timing* problem. For some unknown reason, some DEC devices (like the DS200) actually look at the timing of the various signals. You might try this: o strap RTS to CTS on both sides (hey, we're full duplex, right?) o Ensure DTR is delivered asserted to the modem. o Likewise, DSR back from modem. o Is Carrier Detect (CD) asserted? (probably necessary for dialout) o If not, is DTR cycling? If it is, force CD at the computer side. Does DTR cycling stop after 1 more cycle? If so, try strapping DSR high at the computer side and reconnect CD to modem... ================================================================================ Note 19.33 MODEM AND COMMUNICATION EQUIPMENT 33 of 60 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 18 lines 27-SEP-1988 16:46 -< Field Service to the Rescue >- -------------------------------------------------------------------------------- Well, my Digital Field Service rep, decided to help out with this one (seeing as he likes to use our system for dialing out, also). Look carefully at the flow chart. You will note that if DSR is not asserted, then you will not leave the first "Wait" box. So we set (one of the **MANY**) dip switches on the modem to only assert DSR when there is a answer tone. Bottom line here ... THE 35 SEC HANGUP LOOP IS **NORMAL** !!! (given that DSR is asserted all the time). My guess is that DEC would probably like you to set bit 1 of TTY_DIALTYPE which will hold DTR and RTS low until there is a RING. ================================================================================ Note 19.34 MODEM AND COMMUNICATION EQUIPMENT 34 of 60 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 12 lines 28-SEP-1988 11:06 -< Modems on server ports >- -------------------------------------------------------------------------------- We've recently moved several modems from VAX DMF ports to DS200 modem enabled ports. We've found that two (Racal-Vadic VA4224Es) work just like they did when hooked to the VAX. One can allocate the LT port and SET HOST/DTE to it or use Kermit. Two others, a Microcom RX/1200 and an Avatex 1200, won't respond at all. In fact, while LAT connections get established when using the VAs, no connection appears for the other two. What should I look for? The server ports are configured identically for all four. As an aside, as you all know, modems that echo commands can get stuck in "Huh?-What?" loops with VMS if they allow dial-in. Does the same thing happen on terminal servers? ================================================================================ Note 19.35 MODEM AND COMMUNICATION EQUIPMENT 35 of 60 EISNER::KENNEDY "Terry Kennedy" 15 lines 28-SEP-1988 22:52 -< A hint... >- -------------------------------------------------------------------------------- Here's a hint. [I'd say 'answer', but I did so badly on the last modem ques- tion someone asked. And here I am giving a modem session at DECUS next month!] DS200's have *three* important modem setups - 1) The software setup (DEF PORT X MODEM ENABLED, etc...) 2) The hardware cabling 3) The *timing* on the interface signals It's probably either 2 or 3, which can sometimes be corrected by 1. Do you have a breakout box? If so, compare pins 6, 8, & 20 between the modems that work and those that don't. By the way, the DS200 docset is pretty clear on all this, if you understand that they're doing funny things with the timings. ================================================================================ Note 19.36 MODEM AND COMMUNICATION EQUIPMENT 36 of 60 EISNER::MARTIN_T 37 lines 5-APR-1989 16:20 -< Another Hayes/DS200 Timing Problem >- -------------------------------------------------------------------------------- I've encountered an interesting timing problem trying to connect a Hayes- compatible modem to a DECserver 200/MC. I've setup the server port as follows: Port 1: Modem #1 Character Size: 8 Primary Speed: 2400 Flow Control: XON Alternate Speed: 1200 Parity: None Modem Control: Enabled Access: Dynamic Local Switch: ^A Backwards Switch: None Name: MODEM_1 Break: Local Session Limit: 4 Forwards Switch: None Type: Soft Enabled Characteristics: Autoprompt, Broadcast, Dialup, DTRwait, Inactivity Logout, Input Flow Control, Loss Notification, Message Codes Output Flow Control, Password, Verification The port will not sense the fallback speed unless the modem is strapped to track the incoming carrier (versus setting CD high forever). Yet, this same setting causes the server to disregard (i.e. throw away) the modem's CONNECT message intended for the user. This leads to blind dialing and an inability to programmatically automate dialing procedures. I've talked to DEC and the modem manufacturer and received the same explanation that I had already reasoned: it's a timing problem between the two devices. Apparently the server will not allow any data transfer until all pins have been satisfied and the modem is transmitting its response codes prior to that satisfaction. Other than an inability to receive the CONNECT status message, the connection works fine. Does anyone have any ideas or successful Hayes-compatible/DS200 modem recommendations? The modems I have tried are the EVERCOM EVEREX EV-945 and a LEADING EDGE 24E. Terry J. Martin State of Florida (904) 487-1455 ================================================================================ Note 19.37 MODEM AND COMMUNICATION EQUIPMENT 37 of 60 EISNER::TINKELMAN "Bob Tinkelman - CCA" 38 lines 6-APR-1989 00:34 -< Call progress msgs on DS200 dynamic ports >- -------------------------------------------------------------------------------- > Yet, this same setting causes the server to disregard (i.e. throw > away) the modem's CONNECT message intended for the user. Your analysis is the same as ours, though I don't think I'd call this a "timing problem". We ran into it when configuring bidirectional modem ports. The DS200 manual (which is of course not near me at the moment) describes reasonably precisely how it works wrt the modem lead changes. From memory, the scenario you described goes as follows: Your application, wanting to 'dial out', makes a connection to the LAT port. The DS200 raises DTR, gaining the attention of the modem. Your program (or you via Kermit) interact with the modem, eventually giving it a "Dial" command). The "call progress" reports from the modem reach you (or your program) up to the point that your modem first hears the far end modem. At that point your modem enters its initiation phase where it figures out whether it's to talk to the far end at 300/1200/2400/... and whether to use MNP, etc. Some modems raise DSR at this point, and then raise DCD only later when the modem-modem negotiation is complete. The problem is that once the DS200 sess DSR it decides it's talking to a modem and then, at that point, it insists on seeing DCD. Until it does, it ignores all characters coming in from the modem. Thus you do not see those nice messages such as "INITIATING" or "CONNECTED 1200" or ... . If you are lucky, the far end blasts a banner message or a Username prompt at you. More likely, it is waiting for you to send a so it can "autobaud". So you have to sit and wait and guess when to press return. Some modems will take a premature as a signal to abort the call. Now, what to do? For some modems, I don't know of anything to really "fix" this. You just wait and guess (from Kermit) or program your script to wait for some reasonable number of seconds before sending the . Some modems have mode under which they hold DCD high all the time, except that they drop it for a brief period (1 second) at the end of a call. This seems like the best solution I've seen. The CDS DialAccess 296 modem is one such. [It supports V.32 as well as the lower speeds and MNP.] I'm sure that there are other modems that do this too. ================================================================================ Note 19.38 MODEM AND COMMUNICATION EQUIPMENT 38 of 60 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 13 lines 9-APR-1989 00:25 -< DSR should follow CD in DECland. >- -------------------------------------------------------------------------------- >At that point your modem enters its initiation phase where it figures out >whether it's to talk to the far end at 300/1200/2400/... and whether to use >MNP, etc. Some modems raise DSR at this point, and then raise DCD only later >when the modem-modem negotiation is complete. The problem is that once the >DS200 sess DSR it decides it's talking to a modem and then, at that point, it >insists on seeing DCD. Until it does, it ignores all characters coming in >from the modem. Some modems can be optioned so that DSR follows CD. If yours can't, it can be fixed within the cable: Break the line that connects DSR from the modem to the port, and jumper CD to DSR at the port end of the cable. Now the DS200 won't see DSR "early". ================================================================================ Note 19.39 MODEM AND COMMUNICATION EQUIPMENT 39 of 60 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 22 lines 9-APR-1989 00:27 -< ..and for those of you still using traditional muxes... >- -------------------------------------------------------------------------------- >Some modems can be optioned so that DSR follows CD. If yours can't, it >can be fixed within the cable: Break the line that connects DSR from the >modem to the port, and jumper CD to DSR at the port end of the cable. >Now the DS200 won't see DSR "early". Incidently, modems connected directly to VAX terminal muxes MUST be set up this way (either via configuration or cabling), unless you can guarantee that CD will be raised within thirty seconds after raising DSR. (Maybe okay for incoming calls, almost certainly not for outgoing, especially since many of the offending modems leave DSR up all the time, leading to the "DTR cycling" that is so often reported.) Refer to Figure 8-1 in the terminal driver section of the I/O user's guide. When the port is idle, with DSR low, we're sitting in the "Wait" state. When DSR is raised a 30-second timer is started; if CD is not seen before the timer expires, the "timeout" branch is taken and we go to the "Shutdown" state and drop DTR,forcing the modem to hang up. This behavior cannot be circumvented via the TTY_DIALTYPE sysgen parameter. ================================================================================ Note 19.40 MODEM AND COMMUNICATION EQUIPMENT 40 of 60 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 45 lines 9-APR-1989 00:43 -< Typos in VMS Modem Control state diagram >- -------------------------------------------------------------------------------- > Refer to Figure 8-1 in the terminal driver section of the I/O user's guide. Someone else pointed out -- don't know if it was here or on some other system -- that there are a few typoes in that diagram (though none that affect the discussion of the last few notes). Here is a recap. (I wish I could give credit to the person who originally posted this info, since it's been a real lifesaver!) Get out your I/O User's Guide, Part I, find Figure 8-1, and make a copy of it. Mark up the copy as follows: Reading from the top down.... The second box from the top, the one that says "delay 2 sec" with a branch to the right that says "TTY_DIALTYPE=1", leading to a box labelled "RING Wait": That branch should actually be labelled, "TTY_DIALTYPE bit 0 set". Almost in the center of the page, there is a box that says "start 30 sec timer" (NOT the "Init" box, which says "DTR->ON, RTS->ON, start 30 sec timer", but the unnamed box just below "Init", the one that ONLY says "start 30 sec timer"): The left-hand branch, which currently says "reference count=0", should be labelled "timeout" (ie "reference count is still zero after 30 seconds"). The bottom exit path, currently labelled "reference count=0", should be labelled "reference count > 0". The path that jumps around this box, currently labelled "TTY_DIALTYPE=4", should be labelled "TTY_DIALTYPE bit 2 set". Finally, about a third of the way up from the bottom, there is a box labelled "Transmit1" (God only knows why, since there is no Transmit0, Transmit2, or anything else that I can find, but never mind). There are three paths leading out of the bottom of this box to the next one, labelled "Shutdown". The leftmost of these three paths, currently labelled "TTY_DIALTYPE=1", should be labelled "TTY_DIALTYPE bit 1 set". Fun stuff, eh? Please let me know if this description is not clear. ================================================================================ Note 19.41 MODEM AND COMMUNICATION EQUIPMENT 41 of 60 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 8 lines 10-APR-1989 10:25 -------------------------------------------------------------------------------- I think I was one of the few who had spotted the errors in the previously mention diagram. I even SPR'd it (about two years ago) along with corrections. But did DEC change the picture in the V5 doc set ? N-o-o-o !!!! It would be nice to see exactly how the /MODEM and /HANGUP bits modified the flow through that diagram ================================================================================ Note 19.42 MODEM AND COMMUNICATION EQUIPMENT 42 of 60 EISNER::BYRNE_C "Charlie Byrne" 12 lines 16-APR-1989 19:42 -< On/Off switch too much to ask? >- -------------------------------------------------------------------------------- Speaking of modem hardware, Does any one know why my Microcom AX9624 modem, among others, does not have a simple little on/off toggle switch in back? It seems the only way to turn the darn thing on/off is to unplug either the power supply from the AC outlet, or unplug the 6 pin DC connector in the back. My connection to AC is not in a convenient location, and I think it is only a matter of time before one of the male pins for the DC connection breaks off. And I'd prefer not to leave the modem on all the time. Comments? ================================================================================ Note 19.43 MODEM AND COMMUNICATION EQUIPMENT 43 of 60 EISNER::GROSSER "GIGO/WYSIWYG" 8 lines 16-APR-1989 22:01 -< How do you turn this thing off? >- -------------------------------------------------------------------------------- I was one of the lucky people to see the Microcom AX line of modems before they hit the street. I made the same comment on the beta-test report. The product manager said that nobody asked for it when they did their product research. He also said that it would add to the cost of the product! I replied that I didn't think that a $1.00 switch would add that much to a $1000 modem (price before it was released). He said that they were trying to compete in a $500 modem market, regardless of functionality. Marketing strikes again! ================================================================================ Note 19.44 MODEM AND COMMUNICATION EQUIPMENT 44 of 60 EISNER::NAGY "Frank J. Nagy, VAX Guru" 2 lines 17-APR-1989 07:55 -< Eternally ON, but not eternal power >- -------------------------------------------------------------------------------- I leave my AX-2400c powered all the time. Except when the city turns it off for me ;-) ================================================================================ Note 19.45 MODEM AND COMMUNICATION EQUIPMENT 45 of 60 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 2 lines 17-APR-1989 12:23 -< Leave 'em on >- -------------------------------------------------------------------------------- Our 9624s are also on all the time. If I have to turn them off, I pull the transformer cord from the back of the AX. ================================================================================ Note 19.46 MODEM AND COMMUNICATION EQUIPMENT 46 of 60 EISNER::MAYHEW "Bill Mayhew" 11 lines 17-APR-1989 16:26 -< Digital Has It Now, Too >- -------------------------------------------------------------------------------- As noted, this problem is not unique to 9624s. I have here a DEC DF224 which is guilty of the same sin, and a Supra 2400 (cheapie, but it works... more reliably than the DF) which *does* have an on-off switch -- but the switch is on the modem box, not on the plug-in power supply... so power gets consumed, no matter what. I really dislike those plug-in power supplies, as they generate heat but are often in relatively inaccessible (and poorly ventilated) places. Who (if anyone) makes a standalone 2400-baud modem, these days, with a power switch that *works* (i.e. including an in-the-modem power supply and a plain, vanilla (chocolate?) line cord)? ================================================================================ Note 19.47 MODEM AND COMMUNICATION EQUIPMENT 47 of 60 EISNER::BYRNE_C "Charlie Byrne" 8 lines 17-APR-1989 21:00 -< Surprised at heat dissapation from simple device >- -------------------------------------------------------------------------------- Normally I wouldn't give a second thought to leaving a sensitive, low power consumption device turned on all the time. The reason I'm concerned about the Microcom (the AX2400 has same problem) is that even after just a few minutes plugged in the top of the chassis becomes fairly warm to the touch (quite warm, actually). No, I don't have the top vents covered with manuals. ================================================================================ Note 19.48 MODEM AND COMMUNICATION EQUIPMENT 48 of 60 EISNER::HARVEY "Jack Harvey" 5 lines 18-APR-1989 00:46 -< High Temperatures Improve Cooling >- -------------------------------------------------------------------------------- > No, I don't have the top vents covered with manuals. That's the problem, Charlie. No, I'm serious. Use the AX to keep your coffee warm. If it likes that, fine. If it blows up, tell Glen I told you to do it. ================================================================================ Note 19.49 MODEM AND COMMUNICATION EQUIPMENT 49 of 60 EISNER::ERSKINE "Chris Erskine" 4 lines 18-APR-1989 07:09 -< Vendor says leave it on >- -------------------------------------------------------------------------------- Some modems (Racal-Vadic for one) recomend that the modem be left on all of the time. This is used to keep the setup parameters saved. When you turn off the modem, the battery will save them but the battery only will last for so long without power. ================================================================================ Note 19.50 MODEM AND COMMUNICATION EQUIPMENT 50 of 60 EISNER::NAGY "Frank J. Nagy, VAX Guru" 8 lines 18-APR-1989 07:21 -< Microcom modem cases supposed to be warm >- -------------------------------------------------------------------------------- >> The reason I'm concerned about the Microcom (the AX2400 has same >> problem) is that even after just a few minutes plugged in the top >> of the chassis becomes fairly warm to the touch (quite warm, actually). Our AX2400s are the same way; in fact they don't have vents at all. We asked about this and were told that the chassis is used directly for heat dissipation and should be warm to the touch. Of course, the first few units we had would have made nice coffee warmers. :-) ================================================================================ Note 19.51 MODEM AND COMMUNICATION EQUIPMENT 51 of 60 EISNER::BYRNE_C "Charlie Byrne" 4 lines 20-APR-1989 20:45 -< DECUServe does it again! >- -------------------------------------------------------------------------------- Thanks to Grosser (ok I give up what is GIGO?), Nagy, Tillman Mayhew, Jack "Coffee Cup Heat Sink" Harvey, and Chris Erskine. I'll sleep better knowing my modem is not about to catch fire. ================================================================================ Note 19.52 MODEM AND COMMUNICATION EQUIPMENT 52 of 60 EISNER::HAHN "DECUServe Vice Chair" 1 line 20-APR-1989 21:08 -< GIGO = garbage in garbage out >- -------------------------------------------------------------------------------- ================================================================================ Note 19.53 MODEM AND COMMUNICATION EQUIPMENT 53 of 60 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 0 lines 21-APR-1989 11:06 -< GIGO == Garbage in, GOSPEL out! >- -------------------------------------------------------------------------------- ================================================================================ Note 19.54 MODEM AND COMMUNICATION EQUIPMENT 54 of 60 EISNER::ROBERTS_D "don roberts #241" 2 lines 21-APR-1989 19:14 -< What is that supposed to mean? >- -------------------------------------------------------------------------------- > -< GIGO == Garbage in, GOSPEL out! >- ================================================================================ Note 19.55 MODEM AND COMMUNICATION EQUIPMENT 55 of 60 EISNER::MERRIMAN "George Merriman - CCA/NY" 3 lines 21-APR-1989 21:27 -------------------------------------------------------------------------------- Please, folks. Let's get back to talking about MODEMs. Thanks. George Merriman, H-H Co-moderator ================================================================================ Note 19.56 MODEM AND COMMUNICATION EQUIPMENT 56 of 60 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 8 lines 30-APR-1989 03:17 -------------------------------------------------------------------------------- >< Note 19.54 by EISNER::ROBERTS_D "don roberts #241" > > -< What is that supposed to mean? >- > > -< GIGO == Garbage in, GOSPEL out! >- This refers to the tendency of people to regard anything that comes out of a computer as gospel -- just because it came out of a computer. "... and none dare question it." -- R. A. Heinlein ================================================================================ Note 19.57 MODEM AND COMMUNICATION EQUIPMENT 57 of 60 EISNER::PABON 10 lines 19-JUL-1989 20:54 -< Turn Off Call Waiting?? >- -------------------------------------------------------------------------------- Is there a way to turn off "Call Waiting"? Everytime I use my modem at home a call comes in and screws up the communication. I was told by someone that it could be turn off but they didn't know how to do it. I have a USRobotics Courier 2400e Modem (Hayes Compatible). I assume that it would be an * or # then some number (to turn off Call Waiting) before dialing. Thanks. ================================================================================ Note 19.58 MODEM AND COMMUNICATION EQUIPMENT 58 of 60 EISNER::COY "Dale E. Coy" 9 lines 19-JUL-1989 22:46 -< Yep >- -------------------------------------------------------------------------------- > Is there a way to turn off "Call Waiting"? Yes. Every system I'm aware of has some way to turn it off. They sometimes vary from area to area. Somebody is sure to post the most common set(s) of codes. However, a call to your provider's "customer service rep" is (virtually) certain to get the information for your area. ================================================================================ Note 19.59 MODEM AND COMMUNICATION EQUIPMENT 59 of 60 EISNER::FULLER_S "Funky cold medina" 5 lines 19-JUL-1989 22:47 -------------------------------------------------------------------------------- re: .57 Call your local Telco to be sure, but in most locations, dialing '*70' before the phone number will disable call waiting for that specific phone call. ================================================================================ Note 19.60 MODEM AND COMMUNICATION EQUIPMENT 60 of 60 EISNER::MAYHEW "Bill Mayhew" 6 lines 19-JUL-1989 23:22 -< It's in there... >- -------------------------------------------------------------------------------- The command: $ SEARCH DECUSERVE_ALL_TITLES WAITING reveals that DECUSERVE_FORUM topic 32 discusses this (with several notes).  ================================================================================ Note 20.0 RX50 drive differences 2 replies EISNER::GARDNER "Tim Gardner" 11 lines 16-JUN-1987 07:52 -------------------------------------------------------------------------------- Can anyone tell me the differences in RX50 drives marked with bar code label that read SPxnnnn and ZGxnnnn (where x is a letter and n is a digit)? It seems we have problems with the drives under RSX-11M V4.2B with the SP drives, but not with the ZG drives. Thanks. T. ================================================================================ Note 20.1 RX50 drive differences 1 of 2 EISNER::FRISBIE "Alan E. Frisbie" 9 lines 16-JUN-1987 14:51 -< Plant location codes (?) >- -------------------------------------------------------------------------------- >> Can anyone tell me the differences in RX50 drives marked with bar >>> code label that read SPxnnnn and ZGxnnnn (where x is a letter and >> n is a digit)? If they are the standard DEC Location Codes, ZG is the Singapore plant and SP is the Springfield, MA plant. I can't guarantee that this assumption is correct, however. Alan ================================================================================ Note 20.2 RX50 drive differences 2 of 2 EISNER::MCCARTHY 3 lines 17-JUN-1987 15:33 -< Alphabet soup >- -------------------------------------------------------------------------------- Alan is correct. -Brian  ================================================================================ Note 22.0 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 15 replies EISNER::KILLEEN 5 lines 17-JUN-1987 19:19 -------------------------------------------------------------------------------- This note will be used to ask other users if they have already implemented a particular hardware configuration. ================================================================================ Note 22.1 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 1 of 15 EISNER::KILLEEN 13 lines 17-JUN-1987 19:20 -< BA23 BOX EXPANSION >- -------------------------------------------------------------------------------- We a PDP-11/73 in a BA23-A box that has run out power. We also have some BA11-S boxes in storage. 1. Can a BA11-S box be used as an expansion box for a BA23-A box. Have you done it? 2. Do you know what BCVxx-xx cable should be used? 3. What jumpers do you set on the backplane and bezel in the BA11-S box? ================================================================================ Note 22.2 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 2 of 15 EISNER::KILLEEN 17 lines 17-JUN-1987 19:21 -< ba123 expansion >- -------------------------------------------------------------------------------- <<< EISNER::DUA0:[NOTES$LIBRARY]HARDWARE_HELP.NOTE;1 >>> -< HARDWARE_HELP >- ================================================================================ Note 21.2 HAVE YOU THIS HARDWARE CONFIGURATION? 2 of 2 EISNER::CETRON 8 lines 17-JUN-1987 18:53 -< ba123 expansion/ >- -------------------------------------------------------------------------------- along this line, we are configuring a vaxlab in a ba123 with the gpx option. 12 slots is just no enough for all of the a/d and d/a devices that we need. is there an expansion pathway?? is it the rumoured ba223??? -ed ================================================================================ Note 22.3 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 3 of 15 EISNER::KILLEEN "Jeff Killeen" 4 lines 13-JUL-1987 18:59 -< DHQ11 CABINET KITS >- -------------------------------------------------------------------------------- Does anyone know if you can use DHV11 cabinet kits on DHQ11's? I know you can use DZV11 cabinet kits on DZQ11's ================================================================================ Note 22.4 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 4 of 15 EISNER::KILLEEN "Jeff Killeen" 5 lines 13-JUL-1987 19:06 -< MULTI USER VS2000'S >- -------------------------------------------------------------------------------- Does anyone know if you can legally convert a VAX Station 2000 into a multi-user VAX? If the VS2000 is attached to a E-NET wire there is no phyiscal reason why another user could not login the system. Obviously if this could be done it make more sense to buy VS2000's to use as compute servers rather than MV2000's. ================================================================================ Note 22.5 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 5 of 15 EISNER::KENNEDY "Terry Kennedy" 14 lines 14-JUL-1987 02:12 -< Converting LSI to 22-bit addressing >- -------------------------------------------------------------------------------- Here's a 2-parter: I've gotten enough scavenged parts around to have a go at building another system. I've got an old 4-slot 11/03 box (18-bit), a quad 11/23+ board, a MSV11-PL 512Kb memory, and an original RX01 subsystem. 1) Can I just run the 4 new wires on the backplane to bring it up to 22 bits (yes, I know its 4 wires x 4 slots, and be careful with the routing)? 2) If I do so, will RT-11 5.1 work correctly (XM monitor) with the RX01 controller which doesn't know what 22-bit addressing is? Currently I've got the MSV11 strapped in 18-bit mode, running the SJ mon- itor, and all seems well. ================================================================================ Note 22.6 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 6 of 15 EISNER::RICE "Gary Rice" 13 lines 14-JUL-1987 10:05 -< Maybe YES, Maybe NO >- -------------------------------------------------------------------------------- > Does anyone know if you can legally convert a VAX Station 2000 into a > multi-user VAX? If the VS2000 is attached to a E-NET wire there is no > phyiscal reason why another user could not login the system. I attended a big blow-out where DEC "introduced" the new VAX 2000 line. I too asked that question of the tech rep present. The asnwer is "YES via a terminal server", "NO NOT via direct connect". He did however state that the O/S (what later turnbed out to be VMS 4.5c) wasn't final yet so he couldn't predict whether there would be any traps built in to catch such "unauthorized" activity. Gary ================================================================================ Note 22.7 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 7 of 15 EISNER::HASSINGER "Bob Hassinger" 31 lines 14-JUL-1987 10:31 -< Input on multiuser VAXstations... >- -------------------------------------------------------------------------------- > Does anyone know if you can legally convert a VAX Station 2000 into a > multi-user VAX? If the VS2000 is attached to a E-NET wire there is no > phyiscal reason why another user could not login the system. I spent some time at the Spring Symposium trying to get an answer to the same question about multiusers on a VS, particularly the 2000 because of the obvious cost advantages over the multiuser uVAX configurations. I had a lot of trouble getting answers that would help me. 1) yes, hardware-wise a terminal server and Ethernet connection should do it for VMS. However, 2) the software issue is a problem. No one is quite clear on the current situation but more to the point, they indicate that whatever the current situation is it "may change" in the future. I think the real store for my purposes is this. DEC has got to get the price of single user VSs down to compete in that market, but they want to protect their multiuser prices that are a good bit higher. Right now it looks like they do not have this base covered very well. One way or another, particularly if you have a full VMS system available, you can find a way to get the VS to accept multiple users (something like making a copy of the right module - LOGINOUT for example?). This is not to say it is legal under your licenses however. The thing to be aware of is that protection of their pricing levels is very important to DEC so it would be prudent to anticipate they may do things in future releases to control this sort of thing. In the mean time, everyone seemed to think it was possible and currently legal to submit batch jobs to the VS over the Ethernet. Bob H ================================================================================ Note 22.8 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 8 of 15 EISNER::FRISBIE "Alan E. Frisbie" 33 lines 14-JUL-1987 12:55 -< Q18 to Q22 conversion works >- -------------------------------------------------------------------------------- >> -< Converting LSI to 22-bit addressing >- >> 1) Can I just run the 4 new wires on the backplane to bring it up to 22 >> bits (yes, I know its 4 wires x 4 slots, and be careful with >> the routing)? Yes, I did the same thing with my DDV11-B and it works just fine with an 11/73 (dual width version) and a 2MB Andromeda memory. At one time, I also used an 11/23 in it so I know that it will work also. As far as the routing goes, do NOT follow the serpentine pattern. Just run them straight down one side and up the other. You should probably get a copy of DEC's Micronote #005 (new series) (23-Apr-84) which discusses Q18 to Q22 compatibility issues, including problems with certain DEC modules. >> 2) If I do so, will RT-11 5.1 work correctly (XM monitor) with the RX01 >> controller which doesn't know what 22-bit addressing is? I have run RT-11 v5.1 on my converted system with RX01 (non-DMA) and RX02 (handler does the moves above 18-bit space). RSX-11M/M-Plus also work just fine. The only problem I ran into is that the RSX RK05 driver ALWAYS assumes that a 22-bit system has UMR's. Thus, it will crash on a Q-bus system. If you have any problems, I will try to help you. However, this is a very common hack and I would not expect you to run into any difficulty. Alan ================================================================================ Note 22.9 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 9 of 15 EISNER::HANRAHAN "Jamie Hanrahan" 15 lines 14-JUL-1987 14:37 -< Multiuser VS2000 worked in N'ville >- -------------------------------------------------------------------------------- > > Does anyone know if you can legally convert a VAX Station 2000 into a > > multi-user VAX? If the VS2000 is attached to a E-NET wire there is no > > phyiscal reason why another user could not login the system. > I spent some time at the Spring Symposium trying to get an answer to > the same question about multiusers on a VS, particularly the 2000 > because of the obvious cost advantages over the multiuser uVAX > configurations. I had a lot of trouble getting answers that would help > me... In the VAX Campground in Nashville they had a VS2000 and a VS-something (one built around a BA123; anyway, not a VS2000), connected by Ethernet ThinWire. There were certainly no problems doing SET HOSTs from each to the other. ================================================================================ Note 22.10 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 10 of 15 EISNER::KILLEEN "Jeff Killeen" 9 lines 14-JUL-1987 15:16 -< I WANT TO DO IT LEGALLY! >- -------------------------------------------------------------------------------- >>> In the VAX Campground in Nashville they had a VS2000 and a VS-something >>> (one built around a BA123; anyway, not a VS2000), connected by Ethernet >>> ThinWire. There were certainly no problems doing SET HOSTs from each to >>> the other. The VS2000 is shipped with a one user VMS license. Find me an order number to *LEGALLY* covert it to an 8 user system! I know the E-NET links work. The question is are there VMS licenses to do it? ================================================================================ Note 22.11 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 11 of 15 EISNER::KENNEDY "Terry Kennedy" 9 lines 15-JUL-1987 01:50 -< But of course... >- -------------------------------------------------------------------------------- [banging head on wall] Of course, I {know} that the RX01 is not a DMA device. Too bad I can't remember all of the things I know. By the way, I've heard of these 'micronotes', but I could never find out where to get them. Can you or anyone else point me in the right direction? Thanks, Terry ================================================================================ Note 22.12 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 12 of 15 EISNER::FRISBIE "Alan E. Frisbie" 23 lines 15-JUL-1987 02:28 -< MicroNotes and how to get them >- -------------------------------------------------------------------------------- The MicroNotes are a set of application notes related to Q-Bus systems. They include short technical articles, hints & kinks, product changes and other information not found in the regular documentation. I find them to be an invaluable source of information. They are published by DEC's OEM Technical Support Group and are free if you ask nicely. The original set consisted of 111 notes on such vital topics as "Core memory in 11/03 backplanes". The current set covers more recent topics such as I mentioned in my previous note. To get the current collection, write to: OEM Micros Technical Support Group Attn: Cindy Dorval Digital Equipment Corporation 2 Iron Way MRO3-3/G20 Marlboro, MA 01752 Ask for the Collected MicroNotes and to be placed on the MicroNotes distribution list. Expect it to take a month or two before you receive them. Alan Frisbie ================================================================================ Note 22.13 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 13 of 15 EISNER::CETRON 8 lines 15-JUL-1987 03:05 -< ok, lets be legal >- -------------------------------------------------------------------------------- try the same part number for uVMS upgrades of licenses, then substitute q4 for qz and you are set, or even order the qz number. I have several machines that started as q4's (vs's) and are now qz's (mv's) by bying license upgrades. and have 1 qz which is now a q4.... -ed ================================================================================ Note 22.14 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 14 of 15 EISNER::KILLEEN "Jeff Killeen" 15 lines 17-JUL-1987 16:05 -< *** NUTS *** >- -------------------------------------------------------------------------------- >>> Does anyone know if you can legally convert a VAX Station 2000 into a >>> multi-user VAX? If the VS2000 is attached to a E-NET wire there is no >>> phyiscal reason why another user could not login the system. Obviously >>> if this could be done it make more sense to buy VS2000's to use as >>> compute servers rather than MV2000's. Just got the official word from a trusted DEC source. 1. There is no legal way to do this - nor will there be a legal way. 2. There will be *PRO-ACTIVE* steps taken in future releases of VMS to keep this from happening. *NUTS* ================================================================================ Note 22.15 HAVE YOU TRIED THIS HARDWARE CONFIGURATION 15 of 15 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 9 lines 5-JUL-1988 12:42 -< upgrade jumper setting, too? >- -------------------------------------------------------------------------------- >>>I have several machines that started as q4's (vs's) and are >>>now qz's (mv's) by bying license upgrades. and have 1 qz >>>which is now a q4.... Did the upgrade lic. come with instructions for moving the jumper that will tell possibly more restrictive future versions of VMS that you are now nolonger just a VaxStation? Does anyone know where this jumper is?  ================================================================================ Note 23.0 MMJ to DB25 help needed 15 replies EISNER::FRISBIE "Alan E. Frisbie" 33 lines 22-JUN-1987 12:16 -------------------------------------------------------------------------------- A friend of mine is supposed to pre-wire an office for a MicroVAX 2000 with several VT-220 terminals. All the wire must be run through conduit, direct from the CPU to the terminal locations. With knowledge (?) that the terminal connections on the back of the MicroVAX 2000 are Modified Modular Jacks (MMJ), I proposed the following: 1) Do all of the terminal wiring with telco-type wire with MMJ's on both ends. This allows easy installation using a single crimping tool (less expense). 2) Use MMJ-to-DB25 adaptors at the VT-220 ends. The problems are: 1) What are the MMJ to DB25 pin-to-pin connections? 2) Where can the MMJ crimping tool be purchased, how much does it cost and what is the part number? Neither my latest DECdirect, Inmac, Black Box, nor Mod Tap catalogs have it. 3) Is there such a thing as a MMJ-to-DB25 adaptor? Same questions and comments as (2). Or, is my solution not the best one? What else would you suggest? Thanking you in advance, Alan ================================================================================ Note 23.1 MMJ to DB25 help needed 1 of 15 EISNER::GARDNER "Tim Gardner" 19 lines 23-JUN-1987 07:52 -< Conduit? >- -------------------------------------------------------------------------------- > A friend of mine is supposed to pre-wire an office > for a MicroVAX 2000 with several VT-220 terminals. > All the wire must be run through conduit, direct > from the CPU to the terminal locations. I don't have any information on MMJ wiring, but I can make one suggestion that might help. If you are using conduit because the wires must run through walls, ceilings and "environmental airspaces," you might consider using Teflon coated telco cables instead. You will save an enormous amount of time (and expense) over conduit. Our building is wired exclusively with such cables (for both voice and data). Check the local Belden supplier for cable specs. T. ================================================================================ Note 23.2 MMJ to DB25 help needed 2 of 15 EISNER::HASSINGER "Bob Hassinger" 19 lines 23-JUN-1987 09:37 -< Answers in Networks Buyer's Guide >- -------------------------------------------------------------------------------- > Neither my latest DECdirect, Inmac, Black Box, > nor Mod Tap catalogs have it. Did you try the DECconnect section in DEC's Networks and Communications Buyer's Guide? Our copy comes from our DEC rep or "local sales office". > 3) Is there such a thing as a MMJ-to-DB25 adapter? From above book, 1987 April-June edition, page 5.13 - passive adapters: H8571-A for terminals (25-pin D subminiature) and H8571-B for printers (9-pin D subminiature). Also, H3105 Active Converter to convert to full RS423 if the other end is an RS423 device. > 2) Where can the MMJ crimping tool be purchased, > how much does it cost and what is the part number? On pages 5.14 & 5.15: H8241 6-conductor crimp tool for use with H8220 loose MMPs on H8240 unterminated 6-conductor cable. ================================================================================ Note 23.3 MMJ to DB25 help needed 3 of 15 EISNER::TINKELMAN "Bob Tinkelman - CCA" 27 lines 23-JUN-1987 09:53 -< Sources for modular wiring stuff >- -------------------------------------------------------------------------------- First, DB25-to-MMJ adapters are made my both DEC, ModTap and (probably) others. There are different versions (at least from ModTap) dependent on whether you are going to DTE/DCE and whether (in ModTap's terminology) you are a SYS or DROP. If you are (or might ever be) installing more than two terminals in a location, I suggest considering running 25-pair Telco cable with the appropriate 25-pair to Modified Modular adapters. These are also available from Digital and Modtap. I suggest looking in the DECdirect catalog around page 90 in their "Office Communications" section (for the DECserver 200 DL). Also on page 136 in the Video Terminals section, there are some pictures. If you do not have a ModTap catalog, you can also look in the back of a Glasgal catalog. In terms of crimping tools, AMP probably makes the best one. Lots of people resell it at a large markup. (AMP is also a good source for the Ethernet boring tool, as they make the tap assembly most people now use.) The MMF dies should now be available from them. Telephone as opposed to Data oriented dealers tend to be cheaper. We've used TWComm Corp on Long Island (Call 1-800-555-1212 for their number). The DECdirect listed "cable concentrator" goes from 8 MMJs to a 26 wire connector. DEC sells (or will soon) an adapter to go from 36 wire to standard 25-pair phone cable. Modtap might make (at least they should) an adapter to go directly. ================================================================================ Note 23.4 MMJ to DB25 help needed 4 of 15 EISNER::KILLEEN 3 lines 23-JUN-1987 10:25 -< ANOTHER OPTION >- -------------------------------------------------------------------------------- You may want to think about wiring the computer end with the DEC plugs with the offset clip and the terminal end with the standard telco plugs. This will give you more options. ================================================================================ Note 23.5 MMJ to DB25 help needed 5 of 15 EISNER::FRISBIE "Alan E. Frisbie" 50 lines 23-JUN-1987 16:04 -< Comments and THANKS >- -------------------------------------------------------------------------------- >> If you are using conduit because the wires must run >> through walls, ceilings and "environmental airspaces," you might consider >> using Teflon coated telco cables instead. A good idea, but this installation is at a county fairgrounds where the wires must be physically protected. >> Check the local Belden supplier for cable specs. Another critical piece of data besides the above is DELIVERY. Usually you can get it right away, but sometimes the delay can cost more than you save over conduit! Thanks anyway. ------------- >> From above book, 1987 April-June edition, page 5.13 - passive adapters:... Now THAT is what I was looking for! Thank you. ------------- >> First, DB25-to-MMJ adapters are made my both DEC, ModTap and (probably) >> others. I am on ModTap's mailing list, but the items do not appear to be in their current catalog. >> you can also look in the back of a Glasgal catalog. That's one that I never heard of. Where are they and what is their phone number? >> In terms of crimping tools, AMP probably makes the best one. Any idea what the part number is? And the die for MMJ? >> You may want to think about wiring the computer end with the DEC >> plugs with the offset clip and the terminal end with the standard >> telco plugs. This will give you more options. Since there will be some REAL dummies using this installation, we want to forever avoid the possibility of someone plugging an RS-232 telco plug into a phone line and getting ringing voltage zapped into the MicroVAX! That could get expensive. I thank all of you for your very helpful information and comments. Alan ================================================================================ Note 23.6 MMJ to DB25 help needed 6 of 15 EISNER::KASPER "Beverly T. Kasper" 0 lines 30-JUN-1987 11:20 -< Glasgal, 151 Veterans Dr, Northvale, NJ 07647 >- -------------------------------------------------------------------------------- ================================================================================ Note 23.7 MMJ to DB25 help needed 7 of 15 EISNER::TINKELMAN "Bob Tinkelman - CCA" 5 lines 30-JUN-1987 14:09 -< Glasgal address and phone number >- -------------------------------------------------------------------------------- Glasgal's phone number is 201-768-8082 and their full address is: Glasgal Communications, Inc. 151 Veterans Drive Northvale, NJ 07647 ================================================================================ Note 23.8 MMJ to DB25 help needed 8 of 15 EISNER::NORTON 12 lines 23-SEP-1987 17:27 -< AMP MMJ crimp tool >- -------------------------------------------------------------------------------- > In terms of crimping tools, AMP probably makes the best one. >Any idea what the part number is? And the die for MMJ? DEC's LP of $175 seemed too high for my budget, so decided to see what AMP charges. Supposedly the commercial tool is part number 231652-6 at $108.75, from AMP distributors. They have other options available through AMP Special Industries if you're DEC or a DEC subcontractor. ================================================================================ Note 23.9 MMJ to DB25 help needed 9 of 15 EISNER::HAHN 7 lines 23-SEP-1987 19:08 -< Another Vendor >- -------------------------------------------------------------------------------- I do not have my catalog handy but Black Box in Pittsburgh also carries the AMP tool, I use it often to crimp RJ11 and RJ45, it works wonders. I would get at least 2 sets of dies, we have had one die failure. Fortunately I followed my advice and was able to continue. Pierre ================================================================================ Note 23.10 MMJ to DB25 help needed 10 of 15 EISNER::NORTON 7 lines 24-SEP-1987 09:50 -< Black Box doesn't have MMJ's >- -------------------------------------------------------------------------------- > I do not have my catalog handy but Black Box in Pittsburgh also > carries the AMP tool, ... I called Black Box Technical Support today, and the man said they don't have the tool yet, nor a source for the connectors. Everything they currently sell is for standard modular connectors, *not* DEC's MMJ. ================================================================================ Note 23.11 MMJ to DB25 help needed 11 of 15 EISNER::BYRNE_C "Charlie Byrne - Reuters Info Svcs" 12 lines 24-SEP-1987 10:09 -< Black Box Details >- -------------------------------------------------------------------------------- Black Box (412)-746-5565 May 1987 catalog Page 52 I assume these are not the AMP tools, but B.B. has the following: Modular Crimping Tool (GE-FT040) $130 RJ-45 Adapter (GE-FT045) $ 46 RJ Crimp Tool (GE-FT030 $ 32 Replacement RJ-11 Die (GE-FT039) $ 46 ================================================================================ Note 23.12 MMJ to DB25 help needed 12 of 15 EISNER::HAHN 2 lines 25-SEP-1987 06:47 -< It is AMP >- -------------------------------------------------------------------------------- The modular Crimping Tool is AMP Pierre ================================================================================ Note 23.13 MMJ to DB25 help needed 13 of 15 EISNER::HAHN 20 lines 25-SEP-1987 12:10 -< MMJ --> vendor >- -------------------------------------------------------------------------------- Definitive information (as of yesterday) "only DEC and MOD-TAP have die and tool for sale, it is some kind of DEC restriction" Black Box AMP tool will NOT fit the new MMJ die MOD-TAP has the tool and die for MMJ MOD-TAP # 04301 (~$113) AMP modified tool for MMJ die MOD-TAP # 04305 (~$ 90) MMJ die for above MOD-TAP # 02-555 (~$66/100) MMJ plugs YANKEE is a distributor for MOD-TAP and has (as of today) the above items in stock. They have 800 # but it varies from where you are. They are located in MANCHESTER, NH. Best of luck Pierre ================================================================================ Note 23.14 MMJ to DB25 help needed 14 of 15 EISNER::FRISBIE "Alan E. Frisbie" 10 lines 25-SEP-1987 12:15 -< Tool and die compatibility >- -------------------------------------------------------------------------------- >> Black Box AMP tool will NOT fit the new MMJ die >> MOD-TAP has the tool and die for MMJ Will the new (MMJ compatible) AMP tool work with the old dies for RJ11 and RJ45, or do we have to buy two different crimping tools? Alan ================================================================================ Note 23.15 MMJ to DB25 help needed 15 of 15 EISNER::HAHN 4 lines 25-SEP-1987 16:25 -< OLD and NEW both should work. >- -------------------------------------------------------------------------------- The new MOD-TAP is an updated/reworked tool it "downward" compatible with RJ11 and RJ45 existing AMP dies (this is a statement heard from YANKEE).  ================================================================================ Note 24.0 VT240,2planes,sixels HELP 2 replies EISNER::GOLDSTEIN 31 lines 30-JUN-1987 09:51 -------------------------------------------------------------------------------- I have been trying to use the 2 bitplanes of the vt240 to display a grey-scale image that has been thresholded to contain only 4 levels (00, 01, 10, 11). According to the VT240 programmers reference manual, the beast has 2 bitplanes. It is obvious that there actually is 2 bitplanes since I have observed blank, low intensity, medium intensity and high intenstiy graphics on the screen. I am downloading the image to the vt240 using sixels. I have verified however, that when I download the sixel data, it is going into both bitplanes, and therefore I am only getting 00 and 11 as my levels. Therefore, I thought that if I go into ReGIS mode and mask the appropriate planes (using the ReGIS W(F0-3) commands), I can allow writes to plane 0 only send the 1st bitplane allow writes to plane 1 only send the second bitplane. Unfortunately, every time I exit ReGIS mode and go back into sixel mode, the screen is erased !! I actually have done a 4 level display using ReGIS, however this method is SIXTY times slower than sixels. I am hoping that someone can tell me how to do it using sixels. ---------------------------------------------------- Bob Goldstein goldstein%eri.decnet@mghccc.harvard.edu Eye Research Institute of Retina Foundation 20 Staniford St Boston, MA 02114 (617) 742-3140 x404 ================================================================================ Note 24.1 VT240,2planes,sixels HELP 1 of 2 EISNER::HASSINGER "Bob Hassinger" 15 lines 30-JUN-1987 17:59 -< Check "PRINT COLOR" capability in V2.1 firmware >- -------------------------------------------------------------------------------- The "classic" sixel transfer back and forth to the 240/241 was a single bit per pixel. When we got the V2.1 firmware update for our 240s, additional documentation and a second edition of the programmers manual, I noticed there had been a provision added to support "color" printing. I have not tried it and I do not think I have not heard of anyone else doing it but I believe it gives a way to send more than one bit plane of pixels. Reference VT240 Additional Features Notice EK-VT240-AN-001, page 21. Mentions among other things Graphics Print Color Syntax (DECGPCS) in VT240 Programmer Reference Manual EK-VT240-RM-002. (see page 76 and 77) - Since color uses the same bit planes as shades of gray this may (should?) give a way to do what you want. Bob H ================================================================================ Note 24.2 VT240,2planes,sixels HELP 2 of 2 EISNER::GOLDSTEIN 12 lines 3-JUL-1987 10:51 -< Thanks for the reply >- -------------------------------------------------------------------------------- < Note 24.1 by EISNER::HASSINGER "Bob Hassinger" > -< Check "PRINT COLOR" capability in V2.1 firmware >- Reference VT240 Additional Features Notice EK-VT240-AN-001, page 21. Mentions among other things Graphics Print Color Syntax (DECGPCS) in VT240 Programmer Reference Manual EK-VT240-RM-002. (see page 76 and 77) Thanks for the thoughtful reply Bob. I'll try to get ahold of the appropriate document and give it a try. I'll post a reply detailing my success or failure...  ================================================================================ Note 25.0 PDP-11 CPU PERFORMANCE 20 replies EISNER::KILLEEN 1 line 1-JUL-1987 23:41 -------------------------------------------------------------------------------- This topic will be used to discuss PDP-11 cpu performance. ================================================================================ Note 25.1 PDP-11 CPU PERFORMANCE 1 of 20 EISNER::KILLEEN 16 lines 1-JUL-1987 23:50 -< J-11 CPU PERFORMANCE >- -------------------------------------------------------------------------------- I have found the following about J-11 CPU performance 11/53 vs. 11/73-A - The 11/73 is 1.8 times faster 11/73-A vs 11/73-B with standard memory - Same speed 11/73-A vs 11/73-B with 11/83 type PMI memory (Clearpoint QED-1) - The 11/73-B 1.18 times faster 11/73-A vs 11/83 with non PMI memory - the 11/83 is 1.2 times faster 11/73-A vs 11/83 with PMI memory - the 11/83 is 1.4 times faster I am curious about the 11/44 vs 11/73-A and 11/44 vs 11/83 ================================================================================ Note 25.2 PDP-11 CPU PERFORMANCE 2 of 20 EISNER::PROVOST 13 lines 2-JUL-1987 14:50 -< Useful, but... >- -------------------------------------------------------------------------------- >>> < Note 25.1 by EISNER::KILLEEN > >>> -< J-11 CPU PERFORMANCE >- >>> I have found the following about J-11 CPU performance I assume we're talking about integer compute speed. I have seen different results with benchmarks on floating point and interactive performance of systems. Such studies involve isolating performance of systems with diffeng disk drives, operating systems, workloads, etc. Should we try to attack this more complex problem? tjp ================================================================================ Note 25.3 PDP-11 CPU PERFORMANCE 3 of 20 EISNER::KILLEEN 18 lines 2-JUL-1987 15:50 -< MORE INFO >- -------------------------------------------------------------------------------- All the benchmarks I used were Basic-Plus-2 compiles. I initially did a bunch of benchmarks that simulated the commercial environment e.g. sorting, data lookup, transaction processing, and data entry. I found that BP2 compiles seem hit the average of what the other tasks did to the J-11. The 11/53 vs 11/73 benchmark was done with an RQDX3 and RD53 All the 11/73-A vs 11/73-B vs 11/83 was done with a Spectra-25 and a CDC 9715-500 We have done benchmarks with the Spectra 501 with 1MB of on board cache. This allowed us to shut down the 1MB of RSTS memory cache. The 11/53 was 15% faster with it. The 11/73 was 10% faster with it. And the 11/83 was 8% faster with it. This is the only time we have seen a device change the timing factors between two CPU's when everything else was the same. ================================================================================ Note 25.4 PDP-11 CPU PERFORMANCE 4 of 20 EISNER::KILLEEN 9 lines 2-JUL-1987 15:55 -< YES! >- -------------------------------------------------------------------------------- >>> Such studies involve isolating performance of >>> systems with diffeng disk drives, operating systems, >>> workloads, etc. Should we try to attack this more >>> complex problem? YES! - I think it is interesting to see what device mixes do. Such as adding a Spectra-501 to a RSTS 11/53 and shuting down the RSTS memory cache is also a CPU speed improvement. ================================================================================ Note 25.5 PDP-11 CPU PERFORMANCE 5 of 20 EISNER::CETRON 9 lines 2-JUL-1987 16:30 -< 11/4 numbers >- -------------------------------------------------------------------------------- 11/44 vs 11/84 in numerical fortran crunchies 1/3.5 11/44 vs 11/73 in numerical fortran ccrunchies 1/2. all of the above are timed examples of one small type of function. -ed ================================================================================ Note 25.6 PDP-11 CPU PERFORMANCE 6 of 20 EISNER::KENNEDY "Terry Kennedy" 15 lines 2-JUL-1987 17:34 -< Compile speed, 44 vs. J-11 >- -------------------------------------------------------------------------------- Regarding comparing such things as BP2 compiles, the CPU speed is dwarfed by such things as the disk speed. Putting the compiler itself ($BP2IC2.TSK for RSTS) on the RAM disk (DV0:) cuts a 30-minute (wall time) compile down to two minutes (also wall). The above relies on the RSTS 9.x virtual disk driver. This driver uses the CIS option, if present, for all of its string moves. That is rather undocumented, though. Therefore, in the above case, the biggest speed variation is whether or not you use the virtual disk, and whether or not you have CIS. Since CIS is not available (to the best of my knowledge) on any J-11 based system, the best performance for compiles seems to come from the 11/44. ================================================================================ Note 25.7 PDP-11 CPU PERFORMANCE 7 of 20 EISNER::KILLEEN 13 lines 2-JUL-1987 19:04 -< HMMMM.... >- -------------------------------------------------------------------------------- The timing tests were of course based on the CPU times and not Wall times. We find a 3/2 relationship between Wall to CPU time with BP2 on a unloaded system. It is interesting to note that on the two J-11's we have tried that DV0: trick the WALL and CPU times went UP not down. 1. 11/53, RQDX3, RD53 2. 11/73-B, PMI memory, Spectra-25, CDC 9715-500 It must be the CIS block moves that bought you the improvements. ================================================================================ Note 25.8 PDP-11 CPU PERFORMANCE 8 of 20 EISNER::KENNEDY "Terry Kennedy" 14 lines 2-JUL-1987 20:29 -< Addt'l info on timings >- -------------------------------------------------------------------------------- I should clarify my hardware info for timings: CPU: 11/44 OPT: FPU, CIS MEM: 1280Kw DSK: Fujitsu Eagle / Emulex SC31 All compiles were done with all files on the Eagle, including the compiler work files. The BP2 seconday cache was installed, size=16. Yes, the CIS makes a *BIG* difference. Without it, DV0: I/O is word at a time, with a remap between source and destination. What a CPU hog! tmk ================================================================================ Note 25.9 PDP-11 CPU PERFORMANCE 9 of 20 EISNER::KASPER "Beverly T. Kasper" 3 lines 6-JUL-1987 15:01 -< How about using the same benchmark? >- -------------------------------------------------------------------------------- We have an 11/84 with RA80's. I'd like to run the same benchmark you've been running (we have BASIC+2) -- can I get the code? ================================================================================ Note 25.10 PDP-11 CPU PERFORMANCE 10 of 20 EISNER::KILLEEN "Jeff Killeen" 8 lines 6-JUL-1987 15:18 -< GOOD IDEA - WRONG OS >- -------------------------------------------------------------------------------- >>> We have an 11/84 with RA80's. I'd like to run the same benchmark >>> you've been running (we have BASIC+2) -- can I get the code? I would be happy to let you have the code if you get rid of it after you run the benchmark. One small problem.... You have the *WRONG OS* (RSTS people loved to refer to RSX that way 8-)}} TONE=JOKE). Seriously, the timings would be off because of the different operating systems. ================================================================================ Note 25.11 PDP-11 CPU PERFORMANCE 11 of 20 EISNER::LEDERMAN "Bart Z. Lederman" 18 lines 6-JUL-1987 19:46 -< Some other measurement techniques. >- -------------------------------------------------------------------------------- I had to evaluate computer performance once, and did it a few ways. One way was to write a program that actually measured how long it took the CPU to execute various instructions in various addressing modes. It works on the right operating system (RSX), and I ran it on the 11/70, 11/84P and PRO, and wrote up the results and published them on at least one RSX SIG tape. (I seem to recall submitting them to the RSX Newsletter, but don't remember if they were published). All you have to do is run the program and it tells you how fast the computer is. I'd really like to see it run on a number of other computers so we can see just how fast they are. The instructions also tell how to take into account the effect of cache memory, which can make a difference. I also wrote a program in Fortran that did semi-random I/O on disk, but found that the I/O Exerciser (IOX) also did a fairly good job of comparing disks doing random I/O. That was also writen up in the article. ================================================================================ Note 25.12 PDP-11 CPU PERFORMANCE 12 of 20 EISNER::KASPER "Beverly T. Kasper" 11 lines 7-JUL-1987 10:05 -< Oh, is this just a RSTS topic? >- -------------------------------------------------------------------------------- Oh, yuck, yuck. If you were only interested in RSTS, then why did you title the topic PDP-11 performance?!! I *know* there are other RSX users on this system, we're interested in performance, too. {;-) Yes, I'd like to see the Basic code. We have an application we OEM on 11/44's, and there's been some interest expressed in upgrading some sites to 11/84's. Bart, any ideas on approximately when the tape submission and/or article were? I have Multi-Taskers back to March 1980, with some holes. ================================================================================ Note 25.13 PDP-11 CPU PERFORMANCE 13 of 20 EISNER::RICE "Gary Rice" 14 lines 7-JUL-1987 10:39 -< YES it's on the SIG tape >- -------------------------------------------------------------------------------- > It works on the right operating system (RSX), and I ran it on > the 11/70, 11/84P and PRO, and wrote up the results and published them > on at least one RSX SIG tape. (I seem to recall submitting them to the > RSX Newsletter, but don't remember if they were published). All you I collected this stuff for my RX50 media conversion to PRO project. It is located on the Spring '85 tape in [350,50]. I was thinking of excerpting the .DOC file and publishing it as an article in the PRO section of the PC SIG Newsletter. Bart would you object? Gary ================================================================================ Note 25.14 PDP-11 CPU PERFORMANCE 14 of 20 EISNER::KENNEDY "Terry Kennedy" 12 lines 7-JUL-1987 18:15 -< Who, me? >- -------------------------------------------------------------------------------- I'd have no problems getting you the code I used for the compile time tests. It is, however, quite large and extremely badly written. It is an IBM VM internal tape format conversion utility for RSTS. It requires the MAGTAPE() function to run, so I don't know if the RSX compiler will flag that as an error. My (limited) RSX understanding is that you embed QIO directives in the source program to do that sort of thing. If anybody wants the code for timing purposes, let me know and I'll arrange something to get it to you. If you want the code for the function it performs, sorry. It was written under contract for a company which has exclusive rights. tmk ================================================================================ Note 25.15 PDP-11 CPU PERFORMANCE 15 of 20 EISNER::LEDERMAN "Bart Z. Lederman" 23 lines 7-JUL-1987 19:38 -< I think there is a newer version >- -------------------------------------------------------------------------------- > It is located on the Spring '85 tape in [350,50]. I could have sworn there was also a newer version, but that is the right UIC. (Somewhere in [350,5x]) The RSX SIG had a lot of trouble with their newsletters for a while, and at one point I sent the same article in twice, and sent magnetic media, and it never made it into print even though they were short of material. If someone want's to publish it, I'd prefer to send in the latest version of the article. I'd also like it to reach the people who are most interested, and without meaning to denigrate the PC newsletter, I really don't think many people who are interested in CPU performance are going to look at the PC section, since there was only the PRO there, and it's now officially dead. (The program could be adapted to 8086 family processors, though, and I even have a VAX version). Since this is a public forum, I'm going to let the "vox populi" decide where they want to see the article, if at all. ================================================================================ Note 25.16 PDP-11 CPU PERFORMANCE 16 of 20 EISNER::RICE "Gary Rice" 17 lines 8-JUL-1987 10:12 -< Try us on for size >- -------------------------------------------------------------------------------- > The RSX SIG had a lot of trouble with their newsletters for a > while, and at one point I sent the same article in twice, and sent > magnetic media, and it never made it into print even though they were > short of material. I can't help very much with the RSX problem, but I CAN tell you that you will NOT have the same trouble using the PC SIG as your vehicle. > ...performance are going to look at the PC section, since there was only > the PRO there, and it's now officially dead. I haven't seen any formal announcement. Do you know something that the rest of the world should know? Gary ================================================================================ Note 25.17 PDP-11 CPU PERFORMANCE 17 of 20 EISNER::KILLEEN "Jeff Killeen" 4 lines 8-JUL-1987 11:16 -< PRO TALK >- -------------------------------------------------------------------------------- >>> I haven't seen any formal announcement. Do you know something that the >>> rest of the world should know? PRO life and death talk should go in the PRO conference! ================================================================================ Note 25.18 PDP-11 CPU PERFORMANCE 18 of 20 EISNER::KASPER "Beverly T. Kasper" 5 lines 8-JUL-1987 11:41 -< The Multi-Tasker is okay now >- -------------------------------------------------------------------------------- The RSX Newsletter is alive and well under the able editorship of Bruce Mitchell. If you send him a technical article, it Will get published! A note to the editors of the other PDP-11 newsletter editors might be appropriate, so they could point there readers to it. ================================================================================ Note 25.19 PDP-11 CPU PERFORMANCE 19 of 20 EISNER::BYRNE_C "Charlie Byrne" 28 lines 9-SEP-1987 12:11 -< 11/70 versus 11/84 on M+ >- -------------------------------------------------------------------------------- We have run CPU benchmarks on 11/70 vs 11/84 using rsx-11m+ V2.1. The code was loops of executive directives. No other activity on systems. Here are some results. Times are in seconds to perform 1 directive. (the loop was 10000). Directive 11/70 11/84 % Improvement --------- ------ ------ ------------- CLEF local .00025 .00022 12 CLEF group .00030 .00028 7 RDEF local .00025 .00022 12 RDEF group .00039 .00027 10 WTLO local .00039 .00031 21 WTLO group .00042 .00035 17 WTSE local .00039 .00031 21 WTSE group .00048 .00039 19 QIOW simple .00133 .00085 33 (all QIOs 256 bytes to nl:) QIOW/IOS .00137 .00089 35 QIOW/AST .00203 .00138 32 MAP same .00172 .00133 23 MAP move .00171 .00133 22 I was delighted by these results as my 70 is saturated. Unfortunatly, when we introduced a large amount of Unibus DMA (appx 1Mbps) on the systems, the overall improvement dropped to around 5%. It appears the the 84 starts dying with a lot of traffic on the Unibus. Any comments/advice? inprovement of 84 vs 70 dropped to about 5% ================================================================================ Note 25.20 PDP-11 CPU PERFORMANCE 20 of 20 EISNER::KASPER "Beverly T. Kasper" 9 lines 9-SEP-1987 16:48 -< We like our 11/84 >- -------------------------------------------------------------------------------- Thanks for the performance figures! Our upgrade from 11/70 to 11/84, in February 1986, was precipitated by the unreliability of the 11/70, so I was unable to do any benchmarking before wheeling out the old. While it's hard to give a comparison of the 2, we've been pretty happy with the 11/84. It's failures have been few and far between, and those we have had have been **MUCH** easier to troubleshoot than the 11/70's.  ================================================================================ Note 26.0 DEBET/DEREP fiber question 2 replies EISNER::HORN "Larry Horn" 9 lines 14-JUL-1987 20:04 -------------------------------------------------------------------------------- We are planning to install a fiber optic link between two campus locations about 1000 feet apart with a DEBET-R bridge at one end and a DEREP-R at the other. Can anyone provide us with the following information? Thanks. 1. What is the transmitter power level (dbm) of these devices 2. What is the transceiver sensitivity (dbm) 3. The connector types 4. the wavelength and size of fiber used ================================================================================ Note 26.1 DEBET/DEREP fiber question 1 of 2 EISNER::SILL "Jerry Sill" 14 lines 9-JUN-1988 18:30 -< FO Specifications >- -------------------------------------------------------------------------------- We just installed the same thing only using the bridge rather than the repeaters. We used SEICOR Multi-mode 12 Fiber Di-Electric, 62.5 core/125 clad with 3.75 db loss-160 MHZ P/Meter @ 850 NM and 1.75 db loss-500 MHZ P/Meter @ 1300 NM. The cost was $7.00 P/Meter. We used Zero db attenuators for terminators which were hooked into an AT&T fiber optic Patch panel. We then had fiber jumpers ran from the patch panel the the unit. ================================================================================ Note 26.2 DEBET/DEREP fiber question 2 of 2 EISNER::MERRIMAN "George Merriman - CCA/NY" 4 lines 9-JUN-1988 19:42 -< Zero db attenuation? >- -------------------------------------------------------------------------------- > We used Zero db attenuators for terminators. . . What's a zero db attenuator supposed to do?  ================================================================================ Note 27.0 POWER TOPICS 40 replies EISNER::PROVOST 2 lines 21-JUL-1987 14:56 -------------------------------------------------------------------------------- This topic is for discussion of POWER problems, POWER monitoring, POWER protection, etc. ================================================================================ Note 27.1 POWER TOPICS 1 of 40 EISNER::PROVOST 7 lines 21-JUL-1987 14:59 -< LINE MONITOR >- -------------------------------------------------------------------------------- We have some interesting power problems here at the Bates Linear Accelerator Center. Many years ago we had DIGITAL Field Service monitor our power. They used a DRANETZ line monitor, the output of which was quite useful. Unfortunately it suicided overnight. DIGITAL declined to loan us another at the time. We now have need for a similar device for similar reasons. Does anybody have experience or suggestions in this area? ================================================================================ Note 27.2 POWER TOPICS 2 of 40 EISNER::PROVOST 10 lines 21-JUL-1987 15:04 -< POWER GATE >- -------------------------------------------------------------------------------- We have found that preceding and following many power failures the power often fluctuates violently. It has been our practice to shut down all computer systems at the first hint of impending power fail, and to start them only after the power is smooth for some period of time. I remember hearing of a device built by a Florida company which would do this automatically. It provided the ability to set the time interval over which power must be bad before shutdown, the time interval over which power must be good before turn-on, and the definitions of good and bad power. Can anyone remind me who made the device and what it was called? Has anyone had experience with similar devices? ================================================================================ Note 27.3 POWER TOPICS 3 of 40 EISNER::PROVOST 17 lines 21-JUL-1987 15:12 -< POWER CONDITIONING DEVICES >- -------------------------------------------------------------------------------- Knowing we were going to have power problems in a newly constructed computer area, we purchased and implemented a TOPAZ 35KVA power controller. After a period of lightning storms several devices would fail inexplicably and repeatedly in the area. When replaced, they would fail again. Eventually the TOPAZ would fail. The TOPAZ would then be bypassed until repair could be arranged (usually several months.) Once by-passed, everything would work. This process repeated every year around lightning season. Eventually we put in a switch enabling us to by-pass the power controller easily. Still later we stopped repairing it. I heard later that high voltage spikes had been detected in the output of such devices in their failure mode. This would explain our experience. Now the point: Has anyone successful experience with later models of this device, or with some other similar device? tjp ================================================================================ Note 27.4 POWER TOPICS 4 of 40 EISNER::KASPER "Beverly T. Kasper" 20 lines 21-JUL-1987 16:50 -< Liebert products have served us well >- -------------------------------------------------------------------------------- When we built a new facility about 1-1/2 years ago, we installed a pair of Liebert Datawave Power Conditioners. They've been quite reliable, and aside from the fact that we need a UPS (Uninterrupted Power Supply), they meet our needs admirably. THey're controlled from a common console. The Liebert air conditioners in the room hook up to the same console, which has a piercing alarm horn. The power conditioners can be configured to restart if power comes up, or to stay down until manually reset (the option we chose, since we often get fluctuations here). The air units don't go through the power conditioners, so they come right back on when power comes up. This is useful, since it sometimes warms up a lot during the outage. This whole mess is also hooked up to the Halon system, which shuts off the air and power if Halon is discharged. What I think would be ideal for us is exactly this configuration plus a 5-10 minute UPS. This would allow us to ride brownouts, and to bring the systems down gracefully if it looked like we were in for a wait. ================================================================================ Note 27.5 POWER TOPICS 5 of 40 EISNER::KILLEEN "Jeff Killeen" 7 lines 21-JUL-1987 19:50 -< REAL POWER >- -------------------------------------------------------------------------------- Tom... Have you considered building your own N-plant and cutting the lines to Middleton Municipal Light? If you can run a Linear Accelerater you should be able to build your own N-plant. Or maybe you could buy that unused plant up the road in N.H. and take it off the hands off those "lucky" people who live in that state and will have to pay for it one day. 8-}}} ================================================================================ Note 27.6 POWER TOPICS 6 of 40 EISNER::TABOR "Bill Tabor" 9 lines 21-JUL-1987 21:17 -< Motor generator >- -------------------------------------------------------------------------------- Living in thunderstorm alley, getting power from Florida Plunder and loot, and having been the manager of a site where my DEC field service people wanted to know if Planet Earth was grounded here, the solution to our power problem was a motor generator and power distribution system from a company called EPE in calf. After we installed this unit we went for 6 months with out a system crash. We used to average 2 to 3 weeks between major hardware failures. ================================================================================ Note 27.7 POWER TOPICS 7 of 40 EISNER::GARDNER "Tim Gardner" 23 lines 22-JUL-1987 08:06 -< Funny you should ask... >- -------------------------------------------------------------------------------- > After a period of lightning storms several > devices would fail inexplicably and repeatedly in the area. When > replaced, they would fail again. Eventually the TOPAZ would fail. Gee, there are times when some corroboration is nice. Unfortunately, this isn't one of them. We have had numerous problems with our 50KVA Topaz unit, several of which sound similar to yours. Most recently, the RA81 failed twice and the RA60 shortly after that, all with failures in the drive power supplies. Typical you say. Random failures you say. Infant mortality you say. Well just for fun, consider that each of the failures were detected when we powered up the systems after the Topaz shut down. Coincedence? I don't think so. We have had other failures, mostly involving Topaz shutdowns for no apparent reason (no lightning in the area, other power conditioners in the building don't have any trouble.) I had previously believed that we simply had a lemon flavored Topaz. After reading .3 I'm not so sure anymore. T. ================================================================================ Note 27.8 POWER TOPICS 8 of 40 EISNER::CETRON 22 lines 30-JUL-1987 02:05 -< dec cvc's are wonderfull... >- -------------------------------------------------------------------------------- We have been using dec's CVC's for 4-6 years now and have yet to blow anything due to passed thru power problems (and next door in CS, they have gone through 2 eagles and an ra80...) we did have some problems with very low voltages upping the currents due to the 'regulation' nature of the cvc and having nominal 20amp machines draw 30+ amps....this has been fixed (don't electricians have any pride in their work??)) also, the new dranetz's that digital uses are fully digital and do everything but monitor the atmospheric charge density of mars ...the last time we had one on, we had a state wide power failure and I have a beautiful log of the power slowly sagging to 30 volts and then surging to 700 volts on a 120 volt line....nothing was seen on the other side of the cvc.....I laugh everytime we have a less than 1 sec dropout and my computers are the only ones which stay up..... (out of almost 100 around campus...) -ed ================================================================================ Note 27.9 POWER TOPICS 9 of 40 EISNER::LEDERMAN "Bart Z. Lederman" 31 lines 30-JUL-1987 19:13 -< If you really want reliable power.. >- -------------------------------------------------------------------------------- > (don't electricians have any pride in their work??)) No, they don't. Whenever we had new wiring done we had to check it ourselves to see that they didn't put the circuit breaker in the neutral line, or wire the hot lead to the ground pin of the power plug. (I think the electricians we have now are better). We have to operate 24 hours a day, every day, no matter what. Con Ed power goes through transformers and is rectified to DC, where it "floats" over banks of 48 volt lead-acid batteries. It then goes through inverters which convert it back to 120/208 Volt AC power for the computers, modems, fire alarms, and other essential equipment. Even if Con Ed dies completely, everything keeps running until the gas turbine generators come on-line. (In less critical applications, it might be enough to just have 10 minutes of battery capacity without the generators, if you want to ride out short outages, or just have enough time to shut down cleanly after saving to disk/tape, etc.) No matter what comes in on the power line this system can handle it. If there is a long outage we can switch in some lights, and even feed power to the elevators if we have to (no joke in a 33 story building). Equally important: we run practice drills periodically (on weekends, when there is relatively little demand on the systems) where power is deliberately cut and we see if the system really works. One of our competitors had a nearly identical power system but had never tested it "all the way" as we had, so when a real power failure hit... We also shut down and disconnect 1/2 of each redundant system before these tests just in case there is a transient that damages something. ================================================================================ Note 27.10 POWER TOPICS 10 of 40 EISNER::DROPPERS "Seton R Droppers" 26 lines 5-AUG-1987 14:40 -< UPS, New Buildings, elbows, etc. >- -------------------------------------------------------------------------------- Here at PBS we feel it would be highly impolite (embarrassing?) to all of a sudden drop the feed of your favorite PBS show. In response we have installed an UPS (15 minutes) and backup generator (no not gas turbine, simply diesel). All of this worked fine during our "live" tests when we cut power to the technical center. Only problem was when the commercal power went out during one of the areas more impressive thunder boombers the A/C units dropped out. Before we got power back it approached 85 in the computer room. The problem turned out to be that the generator kicked in so quickly that the A/C could not recycle and restart. Repairs are underway to add a delay to the A/C unit startup. We have never had any down time due to comercal power loss, only when a contracters go-fer hit the emergancy power button for the computer room... (At least they didn't try to find out what happened when they pulled the halon release...) Remember: . Cover you switch . Attempt to test not only your technical center/computer room, but, if you can, try shutting the whole building off . Your UPS/generator is only as good as your A/C, (at least in long run) ================================================================================ Note 27.11 POWER TOPICS 11 of 40 EISNER::ERNST_R "Bob Ernst" 5 lines 10-AUG-1987 18:31 -< Terminals & Power >- -------------------------------------------------------------------------------- There seems to be a controversy at most sites about shutting off terminals at night. Some people point to the savings in power and to the avoided damage during lighting storms. Other people point to the increased wear and tear on the machines (and people's patience). Does anyone know of any studies on this topic? ================================================================================ Note 27.12 POWER TOPICS 12 of 40 EISNER::KILLEEN "Jeff Killeen" 19 lines 10-AUG-1987 20:03 -< NOT SCIENTIFIC >- -------------------------------------------------------------------------------- >>> There seems to be a controversy at most sites about shutting off terminals >>> at night. Some people point to the savings in power and to the avoided >>> damage during lighting storms. Other people point to the increased >>> wear and tear on the machines (and people's patience). Does anyone >>> know of any studies on this topic? We manage 78 PDP-11 sites with over 2000 terminals. Those who shut off their terminals at night have had more problems than those who do not - with two execeptions. 1. Terminals without screen savers 2. Wyse 85's - they have heat problem and their cases will turn yellow after a year if left on all the time. There is no question that turning the systems on and off causes many problems. The two sites we had the most problems with were turning the systems on and off. ================================================================================ Note 27.13 POWER TOPICS 13 of 40 EISNER::DELARISCH 12 lines 11-AUG-1987 16:53 -< Leave Them On >- -------------------------------------------------------------------------------- I agree with Jeff. We have found that leaving our terminals have greatly decreased our problems (especially with the VT240 series). As for the lighting, we are in South Florida (Boca Raton) and we have yet (In our Department) had a verifiable lighting strike on our power (unconditioned & w/o UPS). I don't know if we are just lucky or what. Just think of the shock the terminal (or any IC based equipment) receives when it is turned on and off. If the terminal has a screen saver function ... leave it ON! -Arnold ================================================================================ Note 27.14 POWER TOPICS 14 of 40 EISNER::FRISBIE "Alan E. Frisbie" 9 lines 11-AUG-1987 22:12 -< What about color monitors? >- -------------------------------------------------------------------------------- What about the color VAXstation monitors (VR290)? Do the color tubes deteriorate over time with the power on, or should we follow the same rules as for terminals? Personally, I leave my (monochrome) VAXstation and monitor on 24 hours a day unless I will be away for more than a day or two. So far no problems. Alan ================================================================================ Note 27.15 POWER TOPICS 15 of 40 EISNER::TABOR "Bill Tabor" 12 lines 11-AUG-1987 22:42 -< lucky >- -------------------------------------------------------------------------------- >>>> As for the lighting, we are in South Florida (Boca Raton) and we >>>> have yet (In our Department) had a verifiable lighting strike on >>>> our power (unconditioned & w/o UPS). I don't know if we are just >>>> lucky or what. You are close enough to Computer Products (Pompano Beach) that you may be protect by the grounding of my old Computer Room. (or just plain lucky) see note 27.6 ================================================================================ Note 27.16 POWER TOPICS 16 of 40 EISNER::NAGY "Frank J. Nagy, VAX Guru" 9 lines 24-AUG-1987 07:55 -< Poof go the VT100s >- -------------------------------------------------------------------------------- When Fermilab installed the Accelerator Control System in 1981, several VT100s were acquired with the initial equipment. Many of these VT100s were in public areas and left on continuously. Luckily, these VT100s were on Digital maintenance, because after about 5 years we discovered that the video boards would die violents (screen would brighten, dim and then shink to vanishing dot - this was accompanied by a "poofing" noise and a fair amount of smoke out the top of the case). Of course, this did not happen to all the VT100s but we did experience 1-2 failures per week for about 2 months. :-) ================================================================================ Note 27.17 POWER TOPICS 17 of 40 EISNER::KASPER "Beverly T. Kasper" 6 lines 2-SEP-1987 15:07 -< Leave 'em on >- -------------------------------------------------------------------------------- I agree with leaving them on -- we've had a lot more failures on terminals which get power cycled. I leave VT1xx's on, too, usually -- just use reset or setup/9/9 to clar the screen. Is there some reason that this is a bad idea? ================================================================================ Note 27.18 POWER TOPICS 18 of 40 EISNER::TILLMAN "Brian Tillman, Lear Siegler, Inc." 10 lines 3-SEP-1987 09:37 -< Turn 'em off. >- -------------------------------------------------------------------------------- RE: < Note 27.17 by EISNER::KASPER "Beverly T. Kasper" > -< Leave 'em on >- | I leave VT1xx's on, too, usually -- just use reset or setup/9/9 to | clear the screen. Is there some reason that this is a bad idea? Other than the fact that it's environmentally unsound, no. Personally, I'd rather save the electricity for my children and their contemporaries than squander our world's resources on keeping a terminal running because I MIGHT experience a little inconvenience. ================================================================================ Note 27.19 POWER TOPICS 19 of 40 EISNER::KASPER "Beverly T. Kasper" 7 lines 3-SEP-1987 11:22 -< Question is: does it hurt the VT1xx to be left on? >- -------------------------------------------------------------------------------- My reason for leaving them on is *Not* to avoid "a little inconvenience." I was treated to a lecture back in college (we had VT52's; no screen save), from a EE friend, on how bad it was for electronic equipment to have power cycled repeatedly. I therefore leave the terminals on because I believe that I am saving my company (and my department, I might add) from the expense of having my terminal repaired. ================================================================================ Note 27.20 POWER TOPICS 20 of 40 EISNER::TILLMAN "Brian Tillman, Lear Siegler, Inc." 2 lines 3-SEP-1987 11:54 -< I don't think the facts support terminal damage >- -------------------------------------------------------------------------------- I've always turned by terminals off every night when I go home and in 12 years I've never had one fail. ================================================================================ Note 27.21 POWER TOPICS 21 of 40 EISNER::LEDERMAN "Bart Z. Lederman" 21 lines 6-SEP-1987 20:29 -< Turn them off. Period. >- -------------------------------------------------------------------------------- As someone who used to make a living fixing TV and other electronic equipment, and who has a Masters Degree in E.E., etc.... If the equipment comes anywhere close to being properly designed, it will not be hurt by being turned on. People who think so are probably being influenced by 30 or more year old thinking having to do with vacuume tubes. Second, there is very definately a problem leaving any CRT equipment on in that the deflection and high voltage circuits have to be running at full power. Screen saving circuitry (which VT100s don't have anyway) don't help. These are the circuits which are most likely to fail, and the longer they are on the more likely they are to fail. Next comes the CRT itself, and it's life is based almost entirely on how long it's powered up. Quite apart from conserving electrical energy (a sentiment I agree with), your CRT's are much more likely to last a long time if they are turned off when they are not to be used for any length of time (nights and weekends). You may not save much if you turn it of for 1/2 hour during lunch. ================================================================================ Note 27.22 POWER TOPICS 22 of 40 EISNER::KENNEDY "Terry Kennedy" 23 lines 6-SEP-1987 22:56 -< A responsible opposing viewpoint... >- -------------------------------------------------------------------------------- > ... thirty or more year old thinking having to do with vacuum tubes But, the CRT display is a vacuum tube - granted, the only one, but... Also, there are two other issues to take into account: First, it is much harder to build/buy a properly designed switch-mode power supply than ti build/buy a badly designed one which works, most of the time. Switchers are very itchy at turn-on time, and if you have the slightest bounce in the power switch, poof goes the supply! Since DEC exhibis the poor tast to buy ASTEC (pronounced spazz-tech) supplies, I wouldn't turn anything on or off more than I have to. For an example of ASTEC in inaction, find a Radio Shack Model 2 (you'll irritate the owner of it less than the owner of a Micro-11 when you blow up the supply). Now, assuming the unit is on, flip the power switch off and then RIGHT back on. Observe the whine and the pop. This is the same thing that will happen if the switch bounces when you power your DEC terminal on/off. Secondly, much RS-232 equipment (including some of DEC's), will inter- pret 0 volts on the signal line as a BREAK keypress. It isn't supposed to happen that way, but it does. If the computer you are connected to stays on, so should the terminal. ================================================================================ Note 27.23 POWER TOPICS 23 of 40 EISNER::ROCHELEAU "Rod Rocheleau" 14 lines 8-SEP-1987 11:21 -< Leave 220's on ... power switch will fail >- -------------------------------------------------------------------------------- >> >> Secondly, much RS-232 equipment (including some of DEC's), will inter- >> pret 0 volts on the signal line as a BREAK keypress. It isn't supposed >> to happen that way, but it does. If the computer you are connected to >> stays on, so should the terminal. >> Hear hear ... This is most true. I have a data switch that see's a dead terminal as a break. It put their drop out of service. This cure's the problem of people turning off their terminal. One other comment. If you have VT220 you would be advised to leave the power switch alone. It is very prone to failure ... in the off state. ================================================================================ Note 27.24 POWER TOPICS 24 of 40 EISNER::MERRIMAN "George Merriman" 19 lines 8-OCT-1987 20:10 -------------------------------------------------------------------------------- > After a period of lightning storms several devices would fail > inexplicably and repeatedly in the area. When replaced, they > would fail again. Eventually the TOPAZ would fail. > I had previously believed that we simply had a lemon flavored > Topaz. After reading .3 I'm not so sure anymore. I once had a Topaz power unit that went unstable during a brown-out and faded the output line voltage between 0 and 90 volts at a rate of about one second. I'm sure that did the disk drive a lot of good. Speaking of power conditioners, can anyone tell me why it is that I can buy a box to set next to my computer to filter out all the glitches that my computer maker claims will ruin his fancy computer but that same computer maker can't build the same functionality into his equipment's power supplies? ================================================================================ Note 27.25 POWER TOPICS 25 of 40 EISNER::KENNEDY "Terry Kennedy" 34 lines 10-OCT-1987 00:00 -< Cost is the main (only) factor >- -------------------------------------------------------------------------------- > Speaking of power conditioners, can anyone tell me why it is that > I can buy a box to set next to my computer to filter out all the > glitches that my computer maker claims will ruin his fancy > computer but that same computer maker can't build the same > functionality into his equipment's power supplies? Well, I used to be one of those computer manufacturers (president of one, even - who let me in?) and this was one I was asked many, many times. So, here is the recorded explanation: Back when computers were very expensive, it made good sense to include a many-thousand dollar piece of power conditioning gear with the system. Many systems of that period required odd power, anyway, so the power con- ditioner could double as the power supply. For example, my old IBM 370/138 had a motor-generator set which eliminated sags/spikes as well as providing the 400 Hz power the system required. Yet, today, computers have reduced in size and cost far more rapidly than useful power protection gear, which is still largely mechanical (witness motor/generators, batteries, etc.) When the vendor is selling a $10,000 computer the $2,500 for a good, solid UPS is 1/4 the price, or maybe all of the profit. People also ask - why $2,500 for a UPS - I can go to Radio Shack or some place like that and get the same protection for $29.95. The answer is, it is nowhere near the same protection. The 30-dollar box will filter out maybe 1/2 to 2/3 of your minor power line spikes, but does nothing for dips or surges/sags. The $2,500 unit will keep your MicroVAX running during a complete power outage, as well! [Note - I no longer sell systems or UPS gear - in fact I now buy them! I do have definite experience with many popular units - if anyone wants to sit through a sermon on what works and what doesn't (and why), MAIL me a note and I'll post another reply here] ================================================================================ Note 27.26 POWER TOPICS 26 of 40 EISNER::KENNEDY "Terry Kennedy" 5 lines 10-OCT-1987 19:54 -< Coming soon... >- -------------------------------------------------------------------------------- > If anyone wants to sit through a sermon on what works and what doesn't... Ok, many of you have asked for such an item. I am working on preparing it now, and if it's not too big I'll post it here. Otherwise, I'll make it a Newsletter article. ================================================================================ Note 27.27 POWER TOPICS 27 of 40 EISNER::CONROY 38 lines 1-JUL-1988 17:07 -< My 2 cents >- -------------------------------------------------------------------------------- Here's my 2 cents on turning electrical equipment on and off. My company repairs monitors (only when asked, but we've been asked a lot) and what we have discovered is that certain discrete components have a problem with power cycling (turning on and off and ...). Basically, they eventually wear out. Obviously, mechanical parts (like switches) get wear and tear, but it has been our experience that they last longer than many other things. Also, the longer a semiconductor is running, the sooner it will die. But, again, this can take longer than the life of the equipment. In fact, a well designed and constructed terminal will last a LONG time. So much for well designed and constructed devices. Now to DEC terminals. The VT100s had an apparent manufacturing flaw which caused a cold-solder joint. After enough cycles this would break down and cause a resistor to smoke. Hence the problem with "puffs of smoke" and other problems mentioned in a previous message. The fix is simple: resolder the connection. This has fixed more VT100s than anything else I've seen. Of course, if you don't know how, you can take it to DEC for a "nominal" fee. My opinion would be, find some third-party company or buy a new unit. Some of the early VR201 monitors (used on Rainbows, PROs, VT220s, etc) were built with a coil that was marginal enough that it would break after about two-three years of normal use ("normal" being one cycle per day). This would cause a resistor to burn and the screen to go out. The replacement part is, I believe, an improvement. However, none of our repaired units has been in the field long enough to know for a fact. The parts cost about $5 and can be replaced in about 1/2 hour. OR you can have DEC fix it for $120.00 - or the price of a new monitor. I have mentioned some specifics because that is the only safe thing to do. Based on my experiences I personally prefer to leave equipment on constantly (less cycles=less problems). But that can lead to problems (witness power spikes), although turning terminals off my not prevent such things. There are a lot of variables invovled, such as the type of equipment, the rev level of the equipment, how clean your local power is, how clean your building power is, the weather, and the phase of the moon :-} - (just kidding about that!) The temptation is, of course, to generalize from one's own experience, which may or may not be valid for others. So, I'll leave my conclusions at this: For people with VR201 monitors and/or VT100s who are in the Seattle metro area - you'd probably come out ahead if you leave your equipment on over-night (assuming you reduce the brightness on the VT100s as much as possible when not in use). ================================================================================ Note 27.28 POWER TOPICS 28 of 40 EISNER::MERRIMAN "George Merriman - CCA/NY" 5 lines 1-JUL-1988 17:23 -< Semiconductors don't like power cycles either >- -------------------------------------------------------------------------------- > Also, the longer a semiconductor is running, the sooner it will die. Some semiconductor devices, particularly "power" devices, are very susceptible to temperature cycling problems and this is often what sets their lifetime. ================================================================================ Note 27.29 POWER TOPICS 29 of 40 EISNER::SHUMAKER 9 lines 7-JUL-1988 16:42 -< Equipment on and unattended m be a problem... >- -------------------------------------------------------------------------------- Re - Leaving equipment on. If you do leave equipment on and unattended, do go through your fire insurance policy carefully for possible requirements for smoke and/or temperature detectors which ring alarms at a guard station, or fire station, or otherwise attract the attention of humans. Mark Shumaker ================================================================================ Note 27.30 POWER TOPICS 30 of 40 EISNER::SIMONS "Paul, Not that CONVEX!" 11 lines 20-DEC-1988 10:47 -< PC power controllers >- -------------------------------------------------------------------------------- We are in the process of installing a number of PC's. It was decided that they should be accompanied by one of those slim boxes that fit between the system unit and the monitor. The ones that have separate switches for the monitor, modem, hard drive, system unit, etc. These boxes have easily generated the most controversy associated with the procurement of these systems. Can anyone provide me with a reasonable specification for these devices? For instance, is UL approval necessary? What is a reasonable figure for "surge suppression"? Models from Radio Shack and Tripp Lite cost about $80.00, whereas Black Box is selling one for $170.00. What's the difference? ================================================================================ Note 27.31 POWER TOPICS 31 of 40 EISNER::CONROY "Alan Conroy" 28 lines 20-DEC-1988 12:57 -< Some answers >- -------------------------------------------------------------------------------- > is UL > approval necessary? What is a reasonable figure for "surge suppression"? > Models from Radio Shack and Tripp Lite cost about $80.00, whereas > Black Box is selling one for $170.00. What's the difference? Whether or not UL approval is "necessary" depends upon who okays the purchase. There are no laws that I know of that require UL approval on anything, but I'd check with your local Fire Department/Building Inspectors/etc to be sure. The difference between different power conditioning equipment (besides price) includes such things as noise filtering, power draw through the equipment, speed of surge supressing, etc. What you NEED depends entirely upon your situation. Do you have very clean power coming into your building (some buildings have their own power conditioning)? If so, you need less protection that others might. Of course, will you always have clean power? Do you want something that will keep your PCs up long enough to save your work after a power failure (a la UPS's)? The amount of power that your PCs need will have to be supplied by the equipment you choose. For example, if your CPU takes 2 watts, your disk takes another 2 watts, and your monitor takes another 2 watts, then you need equipment that can handle at least 6 watts continuous and probably twice that for short periods of time. Do you want fuses on the equipment? Breakers? On/off switches? What about other considerations - like handling line voltage drops (which are more common{than spikes)? So, as you can see, there is no simple answer to your question about "reasonable figures". Someone else can probably give more details about things like noise ratios, cut-off time, etc. But to make things simple: assuming all other things are equal, the more expensive boxes generally offer you more protection. Also, having some protection is probably better than having none. Personally, I'd never buy ANYTHING from Radio Shack unless I was desperate. ================================================================================ Note 27.32 POWER TOPICS 32 of 40 EISNER::MERRIMAN "George Merriman - CCA/NY" 9 lines 20-DEC-1988 13:56 -< UL approval sometimes needed >- -------------------------------------------------------------------------------- > Whether or not UL approval is "necessary" depends upon who okays the > purchase. There are no laws that I know of that require UL approval on > anything, but I'd check with your local Fire Department/Building > Inspectors/etc to be sure. I think some jurisdictions require UL approval in their fire and building codes -- Los Angelese comes to mind. Your insurance carrier might be interested as well -- the UL approval is granted by an insurance industry body. ================================================================================ Note 27.33 POWER TOPICS 33 of 40 EISNER::WICHMAN "Jon Wichman" 4 lines 20-DEC-1988 17:22 -< What's the difference! >- -------------------------------------------------------------------------------- > Models from Radio Shack and Tripp Lite cost about $80.00, whereas > Black Box is selling one for $170.00. What's the difference? $90.00? ================================================================================ Note 27.34 POWER TOPICS 34 of 40 EISNER::HARVEY "Jack Harvey" 9 lines 20-DEC-1988 18:42 -< Asking us will only get opinions >- -------------------------------------------------------------------------------- >< Note 27.30 by EISNER::SIMONS "Paul, Not that CONVEX!" > > -< PC power controllers >- > > We are in the process of installing a number of PC's. Why not test? Put "power controllers" on ten percent of the PC's. Then see if you can tell the difference. P.S. If you do, you must report the results back here. ================================================================================ Note 27.35 POWER TOPICS 35 of 40 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 12 lines 21-DEC-1988 13:23 -< Get UL approved equipment! >- -------------------------------------------------------------------------------- >> I think some jurisdictions require UL approval in their fire and >> building codes -- Los Angelese comes to mind. My former employer (in Los Angeles) got a visit from the Fire/Building/???? Inspector who found several items of non-UL approved equipment (printers, etc.). It cost a bunch of bucks to have the equipment inspected by the city and get "LA City Approved" stickers for all of them. It isn't worth the problems to get non-UL approved equipment. If we hadn't been super cooperative with the inspector, we could have had our power shut off that day. ================================================================================ Note 27.36 POWER TOPICS 36 of 40 EISNER::LEDERMAN "Bart Z. Lederman" 3 lines 21-DEC-1988 20:35 -< Non UL implies small, not very good company >- -------------------------------------------------------------------------------- I can't remember the last time I saw a piece of equipment that wasn't UL approved (and most also had Canadian and VDE approval as well). Where is this stuff comming from, anyway? ================================================================================ Note 27.37 POWER TOPICS 37 of 40 EISNER::SIMONS "Paul, Not that CONVEX!" 6 lines 27-DEC-1988 17:27 -< Nobody knows? >- -------------------------------------------------------------------------------- > Where is this stuff comming from, anyway? Indeed, nobody wants to admit who *really* makes it. I know it's not made by Radio Shark. However, I am trying to justify buying the TrippLite or Black Box models simply on hearsay - and I'm losing. Nobody out there uses these things? ================================================================================ Note 27.38 POWER TOPICS 38 of 40 EISNER::KENNEDY "Terry Kennedy" 36 lines 28-DEC-1988 02:17 -< I am not impressed by them... >- -------------------------------------------------------------------------------- > Nobody out there uses these things? We had to set up several dozen PC's for people who never used them before (administrative secretaries, etc.). The base configuration was a turbo XT (10 Mhz V20), color display/adapter, 20Mb hard disk, 80-column printer, power controller, and 2 pieces of security hardware. So far, most of the power controllers have failed, the users have demanded mono monitors instead of CGA (which *they* wanted in the first place), and the security hardware has been yanked. So, feel free to use this as a case history. Also, ask them why they want a power controller - the three plugs you have (system unit, monitor, and printer) could just as well be plugged into a switched outlet strip... About the quality (or lack thereof) of these things - the majority of them are made in Taiwan, etc. and don't receive the care of assembly that motherboards, for example, get. Many of the power-related items coming from there have UL stickers - but that's all - they've never been tested and the stickers are fake. 'Buying American' in this case will get you a higher-priced piece of gear, but it's still a glorified on-off switch. I don't think much of the protection afforded when the suppression parts have one leg tied to the ground leg of the cord, but the cord is plugged into a 2-prong wall wsocket with an ungrounded 'cheater'... Most of your machine problems relating to power will be from things like a photocopier on the same circuit, etc. which *cannot* be helped by such devices. Proper wiring is a must, and any attempt to correct for bad wiring after the wall socket will be less than optimum. For those really concerned with power surges, there are large (20- to 500-amp) surge suppressors designed to be mounted next to the power dis- tribution panel, and can be used to protect one circuit or the whole panel, depending on rating... ================================================================================ Note 27.39 POWER TOPICS 39 of 40 EISNER::MCCORMICK_B "Bob McCormick" 35 lines 1-JAN-1989 12:54 -< Two Professional Solutions >- -------------------------------------------------------------------------------- Most of the $20-$50 surge suppressors that you plug into the duplex outlet and then plug your equipment into I stay away from. I've got a couple of them that have failed the test of time ... especially after a lightning strike nearby! I have invested quite a bit of money in protection devices for RF + computer equipment (CMOS) that sits on top of a mountain (a hi-tec amateur radio repeater). Protection for antenna coax, connections to the telephone network, external devices hooked up to a computer controller, and (of course) the power source have all been implemented. For protection of things like RS232 data lines you should look for specs like: 500 joule surge handling capability with a maximum surge current of 25,000 amps 20 volt peak clamping voltage and 100 ma load current ... with a response time of less than 5 ns. For AC power lines look for 5,000 amps maximum surge current, and around 500 joule energy dissipation (for L5-15R 15 amp outlets). Surge eliminators should provide protection for things like HV spikes, over voltages and lightning surges. Many of the generic surge suppressors only use one type of suppression technology: varistor, MOV, gas tubes. Some features that you may want to look for in a surge suppressor: High and low frequency filters, voltage control, and high energy dissipation. A fuse in series with the power which is caused to blow when high energy dissipation is shorted to ground also is worthwhile. Two companies that I have obtained protection devices from, that have passed the test of time, are: PolyPhaser Corporation LEA Dynatech Inc. P O Box 1237 12516 Lakeland Road Gardnerville NV 89410 Santa Fe Springs CA 90670 702.782.2511 213.944.0916 Both orgainzations deal direct as well as through distributors. I don't claim to be an expert in any of these areas ... ================================================================================ Note 27.40 POWER TOPICS 40 of 40 EISNER::MERRIMAN "George Merriman - CCA/NY" 6 lines 3-JAN-1989 20:22 -< continued elsewhere >- -------------------------------------------------------------------------------- I have broken the various discussions in this topic into separate topics so that they will all have separate entries so that they will all appear in DECUSERVE_TITLES. I have also set this topic NOWRITE. Please continue the discussions in the separate topics. George Merriman -- HARDWARE_HELP co-moderator  ================================================================================ Note 28.0 VT240 Problems 13 replies EISNER::DELARISCH 8 lines 5-AUG-1987 15:18 -------------------------------------------------------------------------------- Has anyone out there had problems (hardware) with their VT240 series terminals. At our place, we have had nothing but problems with them! Our new service manager wants to know what problems others have had. Anyone else out there had problems? -Arnold ================================================================================ Note 28.1 VT240 Problems 1 of 13 EISNER::ERSKINE "Chris Erskine" 8 lines 6-AUG-1987 08:15 -< One site with many tubes >- -------------------------------------------------------------------------------- The only real problem that we have seen with a VT24x has been when data is being displayed, a ^S might hang the keyboard. After this happens, the only way to clear it is to power off the base and power it back on. We have had the normal case of different types of failures with them which requires service. There does not seem to be any major difference between the VT220's and VT24x's for quantity of failure. ================================================================================ Note 28.2 VT240 Problems 2 of 13 EISNER::PROVOST 4 lines 6-AUG-1987 09:43 -< FCO? >- -------------------------------------------------------------------------------- How old are your VT240's? There were a number of FCO's which corrected many problems. See back issues of HMS Newsletter. Tom ================================================================================ Note 28.3 VT240 Problems 3 of 13 EISNER::ERSKINE "Chris Erskine" 2 lines 6-AUG-1987 12:32 -< With lockups >- -------------------------------------------------------------------------------- The one that I am currently typing on which has locked up on me is V2.2. ================================================================================ Note 28.4 VT240 Problems 4 of 13 EISNER::WALLIS "Barry Wallis - Fleetwood Enterprises" 6 lines 6-AUG-1987 14:49 -< No Problems Here >- -------------------------------------------------------------------------------- >>> Has anyone out there had problems (hardware) with their VT240 series >>> terminals. I have had my VT241 running in excess of two years without any problems. I generally leave it on 24 hours/day 7 days/week (running SPMs video display). ================================================================================ Note 28.5 VT240 Problems 5 of 13 EISNER::PROVOST 2 lines 6-AUG-1987 16:23 -< nor here >- -------------------------------------------------------------------------------- We're only on v2.1, and we have no problems with 6 VT240's and a VT241. ================================================================================ Note 28.6 VT240 Problems 6 of 13 EISNER::KASPER "Beverly T. Kasper" 8 lines 7-AUG-1987 15:53 -< We've seen problems on VT2xx's >- -------------------------------------------------------------------------------- We have a number of VT2xx terminals. They seem to have more problems in the winter, when the relative humidity here drops to single digits. You collect static, you touch the keyboard, and Blamo! I think one of those FCO's Tom mentioned addressed this. They're especially prone to problems if the users cycle power on them -- try to convince your users to leave them on unless they hang (or power goes out in your building). ================================================================================ Note 28.7 VT240 Problems 7 of 13 EISNER::HANRAHAN "Jamie Hanrahan" 3 lines 7-AUG-1987 17:04 -< a few problems here >- -------------------------------------------------------------------------------- My VT241 hangs sometimes. The only way out is to power-cycle it. Also, it often drops into graphics mode when I invoke EDT or Mail (no others). ================================================================================ Note 28.8 VT240 Problems 8 of 13 EISNER::HAHN 7 lines 19-AUG-1987 09:56 -< Turn it on - Turn it off >- -------------------------------------------------------------------------------- We have 3 VT240 just upgraded to 2.2 from 2.1, they are turned off at nite most every nite. We have had 2 repairs in about 2.5 years, and three upgrades from 1.? to 2.2. If your terminal hangs which mine does infrequently, I just go to set up and CLEAR COMM. This has gotten me out of trouble every time, but once. Pierre ================================================================================ Note 28.9 VT240 Problems 9 of 13 EISNER::ERSKINE "Chris Erskine" 4 lines 20-AUG-1987 09:08 -< Hard Hang >- -------------------------------------------------------------------------------- The hang which I have seen is using the HOLD key (or ^S) while I am receiving data. When it hangs in this mode, the keyboard is DEAD. I have not found any of the keys on it working including the SETUP key. ================================================================================ Note 28.10 VT240 Problems 10 of 13 EISNER::HAHN 4 lines 20-AUG-1987 12:23 -< V1.? yes V2.- never >- -------------------------------------------------------------------------------- I vaguely remember the same problem with the version 1.?. When we updated to 2.- I never saw it again. Pierre ================================================================================ Note 28.11 VT240 Problems 11 of 13 EISNER::LEFEBVRE "Kenneth LeFebvre" 7 lines 15-NOV-1987 16:56 -< No Blue in my Tubes >- -------------------------------------------------------------------------------- I have had quite a few problems with the VR241 monitors. We use them on PCs as well as VTs. The most common problem is with one color, usually BLUE for us, going out. Once the cable was bad, but all of the other three or four time, the tube itself stopped displaying Blue. Have any of you experienced this? Did we just receive the bad lot? ================================================================================ Note 28.12 VT240 Problems 12 of 13 EISNER::ERSKINE "Chris Erskine" 2 lines 16-NOV-1987 06:47 -< Not Tubes >- -------------------------------------------------------------------------------- Of the ones I have worked with, I have never seen a problem with the tube. What we have seen is replace the logic board in the base. ================================================================================ Note 28.13 VT240 Problems 13 of 13 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 18 lines 21-APR-1988 09:24 -< VT24x S Problems >- -------------------------------------------------------------------------------- -< Hard Hang >- > The hang which I have seen is using the HOLD key (or ^S) while I > am receiving data. When it hangs in this mode, the keyboard is DEAD. > I have not found any of the keys on it working including the SETUP > key. The problem occurs at a much greater frequency when you operate the vt24x at 19200 baud. After many arguments with my local FS office they bumped it up to engineering. I received a specially modified system board with "NO GUARENTEES". I have been using this modified board with no problems for about 2 months now. You local FS should know about this problem if they read their Tech Tips.  ================================================================================ Note 29.0 VT340 video output? 4 replies EISNER::HORN "Larry Horn" 2 lines 20-AUG-1987 11:05 -------------------------------------------------------------------------------- We'd like to drive a monitor from a VT340 - is there a variation or modification of the '340 that provides video out? ================================================================================ Note 29.1 VT340 video output? 1 of 4 EISNER::HASSINGER "Bob Hassinger" 9 lines 20-AUG-1987 13:12 -< Some information... >- -------------------------------------------------------------------------------- I had several discussions on this point at the Spring Symposium. The spokesman for the VT3xx said they had heard that request several times and that he would "have information on how a user could bring the video out by the end of the (Symposium) week". So far I have not seen it. Supposedly it is not to hard to get at inside. Be aware however that it is not standard rate video. I think the horizontal sweep rate is 31.5 Khz. That greatly limits what you can use it with. ================================================================================ Note 29.2 VT340 video output? 2 of 4 EISNER::HORN "Larry Horn" 7 lines 15-OCT-1987 04:39 -< update? >- -------------------------------------------------------------------------------- >> and that he would "have information on how a user could bring the vide >> out by the end of the (Symposium) week". So far I have not seen it. >> Supposedly it is not to hard to get at inside. Have you heard anything more? Can you provide a contact or suggest where to get started on tracking this down? Thanks. ================================================================================ Note 29.3 VT340 video output? 3 of 4 EISNER::HASSINGER "Bob Hassinger" 7 lines 16-OCT-1987 00:54 -< Nothing yet - will try another route... >- -------------------------------------------------------------------------------- I think that was my note - no, I have heard nothing more. I will try to contact someone else on the system who might have a better chance of getting an answer. If that fails, there should be a chance to follow up at the Fall Symposium (I will not be there however so others will have to do it). Bob H ================================================================================ Note 29.4 VT340 video output? 4 of 4 EISNER::PROVOST 5 lines 20-OCT-1987 08:59 -< I'll ask the experts >- -------------------------------------------------------------------------------- I will pass the question on to the HMS Steering Committee on DCS. The @HMS distribution list includes the DECUS counterpart in the area video terminals, and a contact in what used to be Tech OEM Support. An answer should be forthcoming. Tom  ================================================================================ Note 30.0 finding DEQNA rev. level 2 replies EISNER::HANRAHAN "Jamie Hanrahan" 2 lines 24-AUG-1987 13:51 -------------------------------------------------------------------------------- How do I determine the rev. level of a DEQNA? Specifically, I want to know if this DEQNA is at level E or greater. ================================================================================ Note 30.1 finding DEQNA rev. level 1 of 2 EISNER::KENNEDY "Terry Kennedy" 15 lines 24-AUG-1987 19:32 -< Yours isn't - nobody's is (yet) >- -------------------------------------------------------------------------------- > How do I determine the rev. level of a DEQNA? Specifically, I want > to know if this DEQNA is at level E or greater. If you're curious because of RSTS/E, you need at least a J4 for it to turn off that blabbering non-error non-message. There are currently *NO* J4 boards in existance outside of the RSTS/E engineering group. Anyway, on the magenta handle [trivia digression - DEC modules have the first letter of the module number indicating the color of the module. That is why M's are purple and G's are green], you should find stamped a letter/number pair. That is the revision level the board had when it left the factory. If you have had FCO's added to the board, all bets are off - most people don't mark them. Terry ================================================================================ Note 30.2 finding DEQNA rev. level 2 of 2 EISNER::CETRON 6 lines 30-AUG-1987 23:39 -< if less then 1 year old, likely to be e >- -------------------------------------------------------------------------------- actually, if yours works it is likely to be rev e, the pre-'e' boards run very poorly and constantly caused rsx and vms to report short circuits and line sync errors. -edc  ================================================================================ Note 31.0 VTXXX TERMINALS 16 replies EISNER::KILLEEN "Jeff Killeen" 3 lines 25-AUG-1987 12:27 -------------------------------------------------------------------------------- We are working on where to put hardware announcements - however I could not resist putting the next two up. ================================================================================ Note 31.3 VTXXX TERMINALS 3 of 16 EISNER::FRISBIE "Alan E. Frisbie" 15 lines 25-AUG-1987 16:01 -< It looks good! >- -------------------------------------------------------------------------------- >> Improved keyboard features include the following: >> >> o Compose key disable >> o Selectable backspace/delete key >> o User-configurable escape key option >> o Keyboard selectable brackets < > for data processing or typist mode I see that DEC finally listened to us! I will have to actually see the new keyboard to be sure, but it looks like it takes care of my major complaints. The price is also very attractive. I'm glad I held out this long. Now I can do directly from VT100's to VT300's and avoid the VT200's. :-) Alan ================================================================================ Note 31.4 VTXXX TERMINALS 4 of 16 EISNER::DELARISCH 40 lines 25-AUG-1987 18:43 -< Still Missed the Boat! >- -------------------------------------------------------------------------------- >> < Note 31.3 by EISNER::FRISBIE "Alan E. Frisbie" > >> -< It looks good! >- >> >> >> Improved keyboard features include the following: >> >> >> >> o Compose key disable >> >> o Selectable backspace/delete key >> >> o User-configurable escape key option >> >> o Keyboard selectable brackets < > for data processing or typist mode >> >> I see that DEC finally listened to us! I will have to >> actually see the new keyboard to be sure, but it looks like >> it takes care of my major complaints. The price is also >> very attractive. I'm glad I held out this long. Now >> I can do directly from VT100's to VT300's and avoid the >> VT200's. :-) Yes, DEC did make some minor improvements over the VT200 series, however they still missed the boat in others! o Text and Graphics STILL on the same logical Bit Plane (Shame on you DEC ... Memory isn't that expensive!) o Only TEK 4010/4014 compatability mode on the VT340! (Its a Color Terminal for cryan' out loud ... why not bring the TEK emulation into the 1980's ... TEK 4205 or TEK 4207 would have been in order) o DEC Connect connectors on the Back for Communications (Very cute ... How much does the DB-25 adaptor sell for?) o Where's MY ESCAPE KEY???? These are just a few of my Gripes with the VT340 (VT300 series)! I'm placing my money into Graph On Terminals ... They have complete VT340 emulation and then Some (Tek Graphics)! -Arnold ================================================================================ Note 31.5 VTXXX TERMINALS 5 of 16 EISNER::KENNEDY "Terry Kennedy" 9 lines 25-AUG-1987 20:41 -< None for me, thanks... >- -------------------------------------------------------------------------------- > o DEC Connect connectors on the Back for Communications > (Very cute ... How much does the DB-25 adaptor sell for?) This is one of those silly DEC things again - the announcement reads 'for those who have not yet made the move to DECConnect', or some such. They're pretty vain to think that I am going to rewire 3 PDP-11's with 32 ports on each, a VAX with 64 ports, and a 256 x 128 data switch just to use unreliable little widgets! Terry ================================================================================ Note 31.6 VTXXX TERMINALS 6 of 16 EISNER::HASSINGER "Bob Hassinger" 18 lines 26-AUG-1987 15:41 -< RS232 adaptor cable included? >- -------------------------------------------------------------------------------- | > o DEC Connect connectors on the Back for Communications | > (Very cute ... How much does the DB-25 adaptor sell for?) | | This is one of those silly DEC things again - the announcement reads | 'for those who have not yet made the move to DECConnect', or some such. | They're pretty vain to think that I am going to rewire 3 PDP-11's with | 32 ports on each, a VAX with 64 ports, and a 256 x 128 data switch just | to use unreliable little widgets! It looks like the announcement says you get an adapter cable included in the price. True or not? > The North American VT320 (VT320-xx) provides a single DEC423 MMJ > communication interface port for connection to a host system or terminal > server. Also provided is a DEC423 to RS232 adapter with a 10-foot DEC423 > DECconnect office cable (BC16E-10). This adapter cable allows easy > connection to 25- pin serial cables, for those users who have not yet > made the move to DECconnect. ================================================================================ Note 31.7 VTXXX TERMINALS 7 of 16 EISNER::DELARISCH 17 lines 26-AUG-1987 17:33 -< Correct ... But! >- -------------------------------------------------------------------------------- >> < Note 31.6 by EISNER::HASSINGER "Bob Hassinger" > >> -< RS232 adaptor cable included? >- >> >> It looks like the announcement says you get an adapter cable included >> in the price. True or not? >> >> > The North American VT320 (VT320-xx) provides a single DEC423 MMJ >> > communication interface port for connection to a host system or terminal >> > server. Also provided is a DEC423 to RS232 adapter with a 10-foot DEC423 >> > DECconnect office cable (BC16E-10). This adapter cable allows easy >> > connection to 25- pin serial cables, for those users who have not yet >> > made the move to DECconnect. You are right ... but then again do we REALLY want to make the MOVE to rather unrelaibale MMJ jacks??? -Arnold ================================================================================ Note 31.8 VTXXX TERMINALS 8 of 16 EISNER::ERSKINE "Chris Erskine" 7 lines 27-AUG-1987 08:26 -< But DEC says it is GOOD >- -------------------------------------------------------------------------------- Another problem with the DECconnect (disconnect) system is that it locks you into DEC way of doing things. The DECconnect system only supports the data leads (Then why do I have to run 8 conductor wire?). Also the RS232 adaptors make the cable into null modems. It is left to the USER to make use of this connector for other things such as connecting it to a modem and using the modem leads. ================================================================================ Note 31.9 VTXXX TERMINALS 9 of 16 EISNER::HORN "Larry Horn" 7 lines 27-AUG-1987 09:36 -< VT340 comm >- -------------------------------------------------------------------------------- The VT340 has two comm ports: #1 - switchable (in setup firmware) between a MMJ and a DB25 (physical connectors on the back); the DB25 has full modem capability #2 - MMJ only ================================================================================ Note 31.10 VTXXX TERMINALS 10 of 16 EISNER::ROCHELEAU "Rod Rocheleau" 18 lines 27-AUG-1987 09:45 -< MMJ connector not un-reliable at my site >- -------------------------------------------------------------------------------- >> You are right ... but then again do we REALLY want to make the >> MOVE to rather unrelaibale MMJ jacks??? I would like to disagree with a few folks. MMJ jacks are not un-reliable. I have 1200 VTxxxx terminals that are using RJ45 modular jacks, We have them installed for over two years, and the only problem we had was a batch of parts that were made in south east Asia. They fell apart when they were installed. MMJ jack can become un-reliable if they are abused or used in a patch panel that is heavly used. My $.02 worth. .r ================================================================================ Note 31.11 VTXXX TERMINALS 11 of 16 EISNER::HANRAHAN "Jamie Hanrahan" 9 lines 27-AUG-1987 14:04 -< no modem ctl on DECconnect? >- -------------------------------------------------------------------------------- I find it curious that DECconnect only supports the data leads. Sometime back, DEC told us that the only way to secure terminals against "login simulator" programs was to wire the terminal's DTR to the mux's Carrier Detect, set the port to /MODEM, and teach people to use shift-break (or was it control-break?) on the VT100 to make it drop DTR. (Or something like that; it's been a while.) What can you do if you're using DECconnect? ================================================================================ Note 31.12 VTXXX TERMINALS 12 of 16 EISNER::ERSKINE "Chris Erskine" 17 lines 31-AUG-1987 08:40 -< How DEC does it (to you) >- -------------------------------------------------------------------------------- After having to build a special set of cables to connect to the DECconnect system, I took a look at what DEC is doing in some of their wiring. The jacks which are in the patch panel and in each office uses 4 pair (8 conductor) wire by DEC standards. The patch cables which DEC uses contain only 6 conductors. The printed circuit board which connects from the punch down posts for the twisted pair wire to the jack itself connects only 2 pair to the jack itself. Of these 4 wires, the RS-232 to MMJ adaptor connects the center 2 wires together to pin 7 of the RS-232 connecter. After all of this, you have used 8 conductors to produce a 3 conductor cable from the patch panel to the terminal in the users office. Makes cents to me. :-] Chris ================================================================================ Note 31.13 VTXXX TERMINALS 13 of 16 EISNER::PRIGOT "Jonathan M. Prigot" 11 lines 15-OCT-1987 11:33 -< MMJ signals >- -------------------------------------------------------------------------------- I am writing this reply on my new VT320 terminal. Yes, the adapter cable was included with the terminal. The six pins in the MMJ are (according to Appendix C in the documentation): 1 DTR 2 TXD+ (Like RS232 TXD) 3 TXD- (Like RS232 GND for TXD+ and DTR) 4 RXD- (Like RS232 GND for RXD+ and DSR) 5 RXD+ (Like RS232 RXD) 6 DSR ================================================================================ Note 31.14 VTXXX TERMINALS 14 of 16 EISNER::KILLEEN "Jeff Killeen" 3 lines 17-DEC-1987 10:42 -< VT330 KEYBOARD >- -------------------------------------------------------------------------------- I am right in assuming there is *NO* way to correct the angle brackets and escape key on the VT330 like you can on the VT320? ================================================================================ Note 31.15 VTXXX TERMINALS 15 of 16 EISNER::KILLEEN "Jeff Killeen" 12 lines 17-DEC-1987 10:51 -< VT330 SSU ESCAPE SEQUENCES >- -------------------------------------------------------------------------------- I cannot find the Escape Sequences for doing session switching in the VT330 manual. What I want to do is write a SSU for RSTS. The only ESC SEQ that is documented is the one you send the VT330 to switch sessions. There has to be more - for example... Session enable The Sequence the VT330 sends when you press switch session Session disable Flow control Did I miss something in the manuals? ================================================================================ Note 31.16 VTXXX TERMINALS 16 of 16 EISNER::HASSINGER "Bob Hassinger" 21 lines 12-JAN-1988 13:28 -< Use of both VT340 graphics memories in one session? >- -------------------------------------------------------------------------------- Apparently the VT340 (and, I presume, the 330) have two complete graphics memories. When running two sessions I take it you can have separate graphics displays for each session, one in each memory and switch between them. In the case where you are only running one session you usually only see the one memory that is currently being displayed but as I understand it, it is possible to separately address the two memories and make the terminal switch the display back and forth between them. In fact you are supposed to be able to display one while writing in the other. In some cases this would allow much nicer applications because the next display could be pre-drawn in the other memory so it could be displayed instantly when the time comes. Does anyone have any experience with or background on these capabilities? (see also related question in DEC_SOFTWARE about VAX GKS access to this capability) Bob H  ================================================================================ Note 32.0 Bar Codes 21 replies EISNER::PINSLEY "Howard W. Pinsley" 30 lines 25-AUG-1987 22:02 -------------------------------------------------------------------------------- I originally placed this note in the office automation conference. Someone suggested that I might get a better response in the hardware conference. I am developing an application that, in many respects, resembles a library system (in the sense that items are checked in and out). The system is designed to produce labels to facilitate the filing process. I would like to print a bar codes on the labels to expedite the sign in/out processes. My preliminary inquiries have yielded the following: A VT2XB is a small box with a light pen that sits near your keyboard. It performs the read function by feeding translated ascii characters to your terminal just as your keyboard normally would. An LG02 printer is a high speed/high quality printer for bar coding. It will intermix ascii and bar codes. It is pretty expensive and not practical for my environment. I would like several desk top printers for several sites. Someone THINKS they heard that either the LA210 or LA75 do bar codes. I have not been able to confirm this. Can anyone supplement or correct this information. In addition, any general advice on implementing this functionality would be welcomed. ================================================================================ Note 32.1 Bar Codes 1 of 21 EISNER::KENNEDY "Terry Kennedy" 18 lines 26-AUG-1987 02:48 -< Some info >- -------------------------------------------------------------------------------- > Can anyone supplement or correct this information. In addition, > any general advice on implementing this functionality would be > welcomed. 1) Most any graphics printer can be coerced into printing bar codes, but... 2) It is very hard to get any printer to achieve a sufficient contrast ratio between black and white. You need carbon (one-shot) ribbons and glossy paper, but the carbon tends to easily come off of said paper. 3) My specific experience has been with the DataSouth DS220 printer, which 'knows' how to print bar codes without any special programming. However, the printer in general looks like it was built in a garage, and I went through 15 in 2 months. By the way, even DEC resorts to specially printing bar-code labels for such things as serial numbers on boards. They are printed the same way that check serial numbers are printed - special offset equipment. The DEC inventory tags are dot-matrixed, but they easily become unreadable because the paper surface is not glossy and picks up dirt. ================================================================================ Note 32.2 Bar Codes 2 of 21 EISNER::TILLMAN "Brian Tillman, Lear Siegler, Inc." 2 lines 26-AUG-1987 09:48 -< Talaris does bar codes >- -------------------------------------------------------------------------------- The Talaris Systems, Inc. line of QUIC-based printers do bar codes. Several styles are supported. ================================================================================ Note 32.3 Bar Codes 3 of 21 EISNER::TABOR "Bill Tabor" 9 lines 29-AUG-1987 14:11 -< DataProducts M100L >- -------------------------------------------------------------------------------- Dataproducts M100L printer has several bar codes inbedded in its smart graphics card. I used several of these printers in a shop floor application where I was tracking parts through a plant. To get it to print the bar code you sent an escape sequence followed by the text to be bar coded followed by a terminating escape sequence this allowed us to print a routing sheet with bar codes on it. It also saved the computer from having to do the graphics code for the bar code. ================================================================================ Note 32.4 Bar Codes 4 of 21 EISNER::ROECKEL 53 lines 15-SEP-1987 13:49 -< Some Experience >- -------------------------------------------------------------------------------- A VT2XB is a small box with a light pen that sits near your keyboard. It performs the read function by feeding translated ascii characters to your terminal just as your keyboard normally would. >> An LG02 printer is a high speed/high quality printer for bar coding. >> It will intermix ascii and bar codes. It is pretty expensive and >> not practical for my environment. I would like several desk top >> printers for several sites. >> Someone THINKS they heard that either the LA210 or LA75 do bar codes. >> I have not been able to confirm this. Howard, I have had some experience in the Bar Code arena. I was using some Radio Shack :-( hardware to create and read bar codes. With the help of others in the office, I created a BASIC program that generated bar codes from a database. The bar codes printed with both ASCII and bars. The printer was Dot Matrix. Even though the project was canceled (and I therefore have no extensive experience with the system I designed), I found out the following: a. You need a Dot Matrix printer that can print in the exact same spot -- then use double strike to get nice dark bar codes. (Just like printing BOLD lettering !!!) b. The 'light pen' used to read them must be very forgiving, because bar codes generated by Dot Matrix are not good by any means. c. I sent a copy of my bar codes to an outfit that tests them for Government Standards -- to say the least they FAILED --- BUT I was at least 99.9% succesful in reading them with the light pen!! d. If you can get one of those Laser Guns, they work GREAT and all brands were 100% :-} succesful in reading bar codes generated by my dot matrix printer. Equipment used: Radio Shack Model 100 portable computer. Bar Code Wand for the 100. (By the way, this wand looks identical to the HP wand.) Radio Shack Dot Matrix Printer DMP-430. Please feel free to contact me by phone if you want to talk. Bruce W. Roeckel Florida Power Corporation Project Engineer (813) 384-7851 ================================================================================ Note 32.5 Bar Codes 5 of 21 EISNER::NELSON 25 lines 15-OCT-1987 08:49 -< Source for Bar code Printers >- -------------------------------------------------------------------------------- > I am developing an application that, in many respects, resembles > a library system (in the sense that items are checked in and out). > The system is designed to produce labels to facilitate the filing > process. I would like to print a bar codes on the labels to expedite > the sign in/out processes. I hope this information isn't too late. I am currently involved in a manufacturing automation project involving tracking and shipping. Each piece, as it comes off the assembly line, is labeled with bar code and human readable information. This label is used to track it through the rest of the plant all the way to loading on the truck. We had some specialized reader requirements (portable radio communications equipment) but the printer we used didn't need to be anything special except FAST. After an industry search for printers, we selected Intermec printers. They have a wide range of printers that, at least for the one we got, work great. They use thermal technology and can produce almost any label. Call them at (206) 348-2667 for more info. I think they also have readers too. I have also heard that you can print bar code on a laser printer if you print the bar code lines perpendictular to the paper path. I don't know if I'd really want to feed sticky label stock through a printer though. :) ================================================================================ Note 32.6 Bar Codes 6 of 21 EISNER::FRISBIE "Alan E. Frisbie" 15 lines 15-OCT-1987 20:24 -< Check serial numbers are NOT offset printed >- -------------------------------------------------------------------------------- >> By the way, even DEC resorts to specially printing bar-code labels for >> such things as serial numbers on boards. They are printed the same way >> that check serial numbers are printed - special offset equipment. I just visited a site that develops systems for check printing and wish to differ. True, the printers ARE special, but they were not offset printers. There were two types, both made by a division of Data Products (or is that Dataproducts?) One was a drum printer and the other was a daisy wheel type. They have to meet VERY strict specifications for character placement and formation, but are otherwise just like ordinary printers. Alan ================================================================================ Note 32.7 Bar Codes 7 of 21 EISNER::HASSINGER "Bob Hassinger" 5 lines 16-OCT-1987 01:03 -< Magnetic ink? >- -------------------------------------------------------------------------------- I think at one time check numbers were done in a special, magnetic ink. Is that still true and does it have anything to do with special equipment requirements? Bob H ================================================================================ Note 32.8 Bar Codes 8 of 21 EISNER::FRISBIE "Alan E. Frisbie" 20 lines 19-OCT-1987 22:30 -< Magnetic ink and high precision >- -------------------------------------------------------------------------------- >> I think at one time check numbers were done in a special, magnetic ink. >> Is that still true and does it have anything to do with special >> equipment requirements? The printers used special film ribbons with magnetic "ink". I have one of them and it is just a Diablo-compatible ribbon cartridge with the special film. What makes the printers special is the mechanical precision and hammer force. The Federal Reserve Bank (which processes almost all inter-bank check transfers) charges $0.25 for checks that cannot be read by their machines. The charge is levied against the bank that printed the field in question. This gives the banks a BIG incentive to Get It Right. From what I hear, meeting MICR specifications is as hard (or harder) than meeting bar code specs. Alan ================================================================================ Note 32.9 Bar Codes 9 of 21 EISNER::LEDERMAN "Bart Z. Lederman" 17 lines 6-FEB-1989 14:24 -< Bar codes on U.S. Mail? >- -------------------------------------------------------------------------------- I have seen what looks like a bar-code type strip of information at the bottom of many pieces of mail. It is apparently some sort of zip-code or other routing information. I have seen this printed with what was obviously a dot-matrix printer, so the system apparently isn't too picky about what it gets. (It was a government [IRS] letter, by the way, so it "must" be approved). If you haven't noticed it, it's a strip of tall and short bars along the lower edge of the letter. The tall bars are about 1/8 inch high. Does anyone know what this encoding is called? Better yet, does anyone know the "code" for writing this stuff? It appears to be fairly simple, but I don't have the time to crack it. (I know this isn't hardware in the strictest sense, but the thread was already started here.) ================================================================================ Note 32.10 Bar Codes 10 of 21 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 9 lines 6-FEB-1989 14:42 -< ZIP bar code >- -------------------------------------------------------------------------------- >> I have seen what looks like a bar-code type strip of information >> at the bottom of many pieces of mail. It is apparently some sort >> of zip-code or other routing information. It is just the 5 or 9 digit ZIP code. I decoded it a few years ago and probably have the code around here someplace. I'll look for it. Alan ================================================================================ Note 32.11 Bar Codes 11 of 21 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 54 lines 6-FEB-1989 19:16 -< Postal Barcodes revealed >- -------------------------------------------------------------------------------- The PO has a special rate for Zip+4 Barcoded mail which is a $ carrot dangled in front of large mailers. Their spec is readily available and most companies already have it, but you have to know where to look. Down near the mail machine with all the other junk is a pile of postal and UPS and Fed-X price books. Find the one labeled POSTAL BULLETIN (in fine print it says 'Published Since March 4, 1880', but get a newer one). In the middle of this one (I can see the clinched staples, so it HAS to be the middle) on page 35 amidst all the other fine print sections there is one labeled "622.165 Barcode Requirements" Anyone with a postage meter or any sort of mailing permit gets mailed these automatically, and I am sure you can get a copy just by asking. There are specs I doubt can be followed such as 'must not exceed .003 inch...". There is a nice picture carefully defining the "Barcode Area" and the surrounding "Clear Zone". The one spec that strikes me as odd is that they want 21 +/- 1 bars per inch. 20 - 1 = 20, and with 4 slugs in my chain or band or drum (low-low, low-high, high-low, and high-high) I can do this on any vanilla 10 cpi hard type printer, but why didn't they spec it as 20 +/- 1, given the wide availability of 10 cpi hardware? There are 52 bars in the field. The first and last are are full height and are FRAME bars, the remaining 50 are 10 digits of 5 bars each. The 10 digits are ZIP + 4 + "correction digit". Their text references an EXHIBIT not in this leaflet, that apparently shows the codes themselves, and most likely how the check digit is computed. Looking at random mail it would seem that the digits all have 2 full and 3 half high bars, between our ZIP + 4 code and a DEC prepaid mailer going to Spit Brook Rd. in Nashua, I have decoded the following: 0 = HHLLL 1 = LLLHH 2 = LLHLH 3 = LLHHL 4 = LHLLH 5 = 6 = LHHLL 7 = HLLLH 8 = HLLHL 9 = HLHLL Someone else can attack the check digit, and # 5. ================================================================================ Note 32.12 Bar Codes 12 of 21 EISNER::PINSLEY "Howard W. Pinsley" 7 lines 6-FEB-1989 21:01 -< General Bar Code Info >- -------------------------------------------------------------------------------- I know this is a little off the subject of zip codes, but I thought I'd mention this here in case anyone needs this info. A company called Intermec publishes a free booklet entitled Bar Code Symbology. It is an excellent reference for those who need to produce bar codes; there are several popular codes (e.g. UPC, code 39, etc.) I used this information to implement bar codes for one of my systems -- it came in very handy. ================================================================================ Note 32.13 Bar Codes 13 of 21 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 10 lines 6-FEB-1989 21:34 -< sound good (scans well?), tell us where >- -------------------------------------------------------------------------------- > A company called Intermec publishes a free booklet entitled > Bar Code Symbology. Help! you have the info, pass it on. And their 1-800 for ordering it is? : Or their vanilla phone number? : Or their (yuk) mail address? : And is there some PUB # or is that all they publish? ================================================================================ Note 32.14 Bar Codes 14 of 21 EISNER::KILGALLEN "Larry Kilgallen" 23 lines 7-FEB-1989 04:42 -< Publication 25 (August 1988) says 5 = LLHLHL >- -------------------------------------------------------------------------------- That's from Appendix G, "Reading the ZIP+4 Postnet Bar Code" (I presume they came up with the name "Postnet" sometime after the publication of Barton's manual) of Publication 25, A Guide to Business Mail Preparation. Publication 25 is quite free to those applying for a business reply permit, and is probably also free to anyone who claims to be considering applying for such a permit. If you need it immediately, the trick is finding out where to get it. In Cambridge, they send you to Boston for anything related to Business Reply, and Boston assigns you a Cambridge number. The appropriate counter in Boston's Central Mail Facility is *not* the one where they will take your payments for meter permits, but is sort of "inside" in a location which *looks* like it ought to be marked Employees Only. It is more "inside" than the Credit Union, for example. So find the corresponding location in a *major* Post Office near you, and receive a very nice 42 page publication which goes into such esoteric subjects as spectral response curves of inks and papers. In contrast to the complaints about dot matrix (and Publication 25 does outline their scanning algorithm), I find that the specifications seem *ideal* for a PostScript implementation. (Perhaps that is why they named it POSTnet :-). ================================================================================ Note 32.15 Bar Codes 15 of 21 EISNER::PINSLEY "Howard W. Pinsley" 24 lines 7-FEB-1989 12:00 -< More Info on Intermec >- -------------------------------------------------------------------------------- > > A company called Intermec publishes a free booklet entitled > > Bar Code Symbology. >Help! you have the info, pass it on. >And their 1-800 for ordering it is? : >Or their vanilla phone number? : >Or their (yuk) mail address? : >And is there some PUB # or is that all they publish? Here's what I have: Bar Code Symbology Some Observations on Theory and Practice by David C. Allais Published by: Intermec 4405 Russell Road, P.O. Box 360602 Lynnwood, WA 98046-9702 (206) 348-2600 Telex: 3794107 ================================================================================ Note 32.16 Bar Codes 16 of 21 EISNER::LEDERMAN "Bart Z. Lederman" 26 lines 8-FEB-1989 08:09 -< Zip codes now clear. >- -------------------------------------------------------------------------------- I have examined some more samples of mail. I have a number of letters with the bar code at the bottom which were not printed by the sender. It appears that the post office is printing the bar coded zip on as part of the processing scheme. All of these samples were printed with dot matrix printers and not particularly black ribbons, and you can see the normal variations in dot placement one sees on an old (or cheap or worn) dot matrix printer. Their readers either aren't as fussy as the spec. calls for (.003"?) or else a lot of bar codes just aren't read. It apparently doesn't have to be a 9 digit zip, as I have samples with just the 5 digit zip (plus correction digit). On all of my examples: if you take all of the digits, including the correction digit, and add them up, you get zero (Modulo 10). The correction digit appears to be a simple parity type check. The code must be capable of doing other stuff as well because a number of government envelopes have a second strip of bar coding to the left of the zip code, and it doesn't decode as digits. It would be interesting to know what it is, but not essential. Thanks to the information here I now have a chance of bar coding my mail if I feel like taking the trouble to do so. It's not so much the cost (for my volume of home mail) as the principle that the post office needs all the help it can get. ================================================================================ Note 32.17 Bar Codes 17 of 21 EISNER::MCCORMICK_B "Bob McCormick" 8 lines 8-FEB-1989 19:03 -< Canada does it better >- -------------------------------------------------------------------------------- The canadians use a fluorescent yellow bar codes on their mail, much better than our scheme! The spacing is greater... Canadian postal codes are: XNX NXN where X is a letter and N is a number ... I've got their spec in the office ... I'll see if it tells their encoding scheme! ================================================================================ Note 32.18 Bar Codes 18 of 21 EISNER::OSUDAR "John Osudar" 13 lines 23-FEB-1989 22:45 -< Postal Service bar code encoding -- from a colleague >- -------------------------------------------------------------------------------- > 0 = HHLLL > 1 = LLLHH > 2 = LLHLH > 3 = LLHHL > 4 = LHLLH > 5 = LHLHL > 6 = LHHLL > 7 = HLLLH > 8 = HLLHL > 9 = HLHLL 74210 -- "weights" for "H" in that position Note that a zero is represented by the number 11! ================================================================================ Note 32.19 Bar Codes 19 of 21 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 4 lines 24-FEB-1989 02:33 -< Hmm. This could explain a lot! :-) >- -------------------------------------------------------------------------------- > < Note 32.18 by EISNER::OSUDAR "John Osudar" > > -< Postal Service bar code encoding -- from a colleague >- > Note that a zero is represented by the number 11! ================================================================================ Note 32.20 Bar Codes 20 of 21 EISNER::KOZAM 12 lines 25-FEB-1989 01:10 -< Public Domain Code is Around >- -------------------------------------------------------------------------------- I got some code from usenet quite a while ago that does produces mail bar codes. It's public domain and in C. For example, if you enter the ZIP code 07631, it displays |||...|...|.||....||....||..||.| It converts in the other direction as well. The code provides a model for anyone interested in writing this sort of thing. If someone is interested, I could see if I still have it somewhere. By the way, the code provides for parity, so that postal machinery knows if it was read reliably. Also, you will notice two lengths of code: one for regular ZIP, one for ZIP+4. ================================================================================ Note 32.21 Bar Codes 21 of 21 EISNER::LEDERMAN "Bart Z. Lederman" 6 lines 31-JUL-1989 13:37 -< DEC working for Post Office >- -------------------------------------------------------------------------------- According to the current issue of "InformationWEEK", DEC has won a $20,000,000 portion of a contract with the U.S.P.S., working with ElectroCom Automation to supply an optical character reader and bar-code system that will read and print ZIP codes. So some of that might filter through to us, eventually. The blurb doesn't give any details except that DEC workstations will be used.  ================================================================================ Note 34.0 Ethernet transceivers and cables 22 replies EISNER::FRISBIE "Alan E. Frisbie" 32 lines 27-AUG-1987 18:19 -------------------------------------------------------------------------------- An observation and a few questions: I just experimented and discovered that a DESTA (Ethernet Thin-wire transceiver) can (barely) be plugged directly onto the BA23 or BA123 cabinet kit connector. On a BA123, there is sufficient clearance that the back door can be closed with no problem. On a BA23, however, the DESTA sticks out too far for the back cover to be put on. Obviously, this saves the cost of a transceiver cable (minimum of $58). Does anyone know if there are any problems with this? A note that comes with the DESTA says that it will not be FCC compliant unless at least a 20 meter cable ($135) is used. Just after trying this, I noticed in the latest (24-Aug) Digital Review that Micro Technology Inc. has an adapter that is *designed* to mount directly to the cabinet kit. It is small enough to fit on a BA23 box with the rear cover installed. It is called the MESTA, and costs $250. Questions: 1) What is the actual, physical difference between the *Ethernet* transceiver cables (BNE3A-xx) and the *IEEE 802.3* transceiver cables (BNE3H-xx)? Are there extra wires or what? 2) Is there anything special about DEC's thin-wire Ethernet coax, or is it just good old RG58A/U? Alan ================================================================================ Note 34.1 Ethernet transceivers and cables 1 of 22 EISNER::LEDERMAN "Bart Z. Lederman" 21 lines 28-AUG-1987 11:23 -< Yes there is a difference >- -------------------------------------------------------------------------------- > 2) Is there anything special about DEC's thin-wire Ethernet > coax, or is it just good old RG58A/U? DEC's "experts" who came to talk to the local Cluster LUG meeting say yes, there is a difference between thin-wire and RG58. They say if your runs are short you can get away with it but DEC may not support it. If you want to go close to the maximum specified distances and number of connections you should use the special wire. From my own experience: "normal" RG58 and similar coax is specified for RF work, not for baseband use. With Ethernet, having a difference in time delay for a signal propagated at 1 MHz and one at 10 MHz would be a serious problem, but probably would not affect most RF and Cable TV work. Ethernet is apparently also rather sensitive to noise, hence the double and triple shielding on "true" thick Ethernet wire. The "true" wire is also supposed to be less subceptable to propagation changes and noise when bent. There are many potential problems with a baseband system: but if your runs are short you are not likely to run into most of them (hence the ability to use thin wire in the first place). ================================================================================ Note 34.2 Ethernet transceivers and cables 2 of 22 EISNER::GARDNER "Tim Gardner" 15 lines 28-AUG-1987 13:40 -< 15 wires in IEEE >- -------------------------------------------------------------------------------- >Questions: > >1) What is the actual, physical difference between the > *Ethernet* transceiver cables (BNE3A-xx) and the > *IEEE 802.3* transceiver cables (BNE3H-xx)? Are there > extra wires or what? > In the IEEE versions, all fifteen conductors are present. In the Ethernet versions, only nine of the fifteen are present. Unfortunately, I don't have any information on how the extra conductiors are used. T. ================================================================================ Note 34.4 Ethernet transceivers and cables 4 of 22 EISNER::DELARISCH 18 lines 28-AUG-1987 15:06 -< Yes, Cable Differences! >- -------------------------------------------------------------------------------- >> < Note 34.2 by EISNER::GARDNER "Tim Gardner" > >> -< 15 wires in IEEE >- >> >> In the IEEE versions, all fifteen conductors are present. In the Ethernet >> versions, only nine of the fifteen are present. Unfortunately, I don't >> have any information on how the extra conductiors are used. According to uninformed sources (ME) ... the IEEE 802.3 spec for the tranceiver cable dictates that all unused pins (i.e. those which were not connected in Ethernet V1.0 spec MUST be grounded I'm not sure if there were any other significant changes other than the Heartbeat signal which goes between the tranciever and the E-net board. -Arnold p.s. If you REALLY need the info, I can dig out the IEEE 802.3 spec for the tranciever cable and post it (Maybe in the Standards Conference?) ================================================================================ Note 34.5 Ethernet transceivers and cables 5 of 22 EISNER::MACNEIL "Laird of the Highlands" 5 lines 29-AUG-1987 01:06 -< Plugging in DESTA's >- -------------------------------------------------------------------------------- I don't have any answers to your questions, but I would just like to comment that the MTI's Ethernet Thin-Wire adapter. Sure looks sexy! I have heard, and read that people have had problems with the DESTA's, but can not speak from live, personal experience. I think that I will look at the MESTA instead of the DESTA. ================================================================================ Note 34.6 Ethernet transceivers and cables 6 of 22 EISNER::ROCHELEAU "Rod Rocheleau" 18 lines 10-SEP-1987 11:01 -< Under carpet Thinwire >- -------------------------------------------------------------------------------- Does anyone have any experience with under carpet thin-wire ethernet. I have a pair of VAXMATE's that I have to install in a cubicle area. These cubicles are completely standalone with only under carpet access. My company does not allow power poles or anything that would allow the standard DEC thinwire to be run under the carpet. I have one solution that would require installing a DEMPR with each VAXMATE...what a waste...that is running the transceiver cable over flat cable. It sure would be nice if DEC made a DESTA that converted the other way. Rod ================================================================================ Note 34.7 Ethernet transceivers and cables 7 of 22 EISNER::NORTON 5 lines 11-SEP-1987 11:57 -< Ethernet on twisted pair >- -------------------------------------------------------------------------------- > Does anyone have any experience with under carpet thin-wire > ethernet. One of the DECworld announcements is supposed to be Ethernet on twisted pair wiring. Maybe this would work. ================================================================================ Note 34.8 Ethernet transceivers and cables 8 of 22 EISNER::ROCHELEAU "Rod Rocheleau" 14 lines 11-SEP-1987 14:32 -< DEC twisted pair ... too little, too late >- -------------------------------------------------------------------------------- >> < Note 34.7 by EISNER::NORTON > >> >> One of the DECworld announcements is supposed to be Ethernet on >> twisted pair wiring. Maybe this would work. It would be nice but there are two problems. 1. It is not available until Jan 88 2. The DEMPR will only support eight devices i.e. You can only hang one device off each thin-wire port rod ================================================================================ Note 34.9 Ethernet transceivers and cables 9 of 22 EISNER::NAGY "Frank J. Nagy, VAX Guru" 9 lines 12-SEP-1987 10:38 -< RE: ThinWire under carpet >- -------------------------------------------------------------------------------- In reply to 34.6 about running ThinWire Ethernet under a carpet... The wire used in ThinWire Ethernet is just RG-58 coaxial cable (we are about to install ThinWire here at Fermilab - not using cable bought from Digital). Unless you put the cable inside some sort of protective device, people will walk on it. This will destroy your cable either by crushing the insulator and changing the impedance or by causing the central conductor to short to the outer shield. So, if you do this, make sure your cables are adequately protected. ================================================================================ Note 34.10 Ethernet transceivers and cables 10 of 22 EISNER::NAGY "Frank J. Nagy, VAX Guru" 17 lines 12-SEP-1987 10:45 -< >8 twisted-pair stations on DEMPR >- -------------------------------------------------------------------------------- >> Note 34.8 by EISNER::ROCHELEAU "Rod Rocheleau" >> 2. The DEMPR will only support eight devices >> i.e. You can only hang one device off each thin-wire port According to some paper sitting in front of me: The Unshielded Twisted-Pair Ethernet Wiring Closet/SER Adapter (H3330-AA) is a package that contains four dual adapter boards to support up to 8 users in the SER or wiring closet. Four H3330s will fill a single Rack Installation Kit (H3340-AA) for 32 users. There is also a little picture showing what seems to be a single port of DEMPR connected to do this. It is explicitly mentioned that an Unshielded Twisted-Pair Ethernet line will support on a SINGLE station, but it looks like the ThinWire port on a DEMPR will support a fanout to between 8 (minimum) and 32 twisted-pair stations. ================================================================================ Note 34.11 Ethernet transceivers and cables 11 of 22 EISNER::PRIGOT "Jonathan M. Prigot" 10 lines 13-SEP-1987 22:25 -< My $0.02 >- -------------------------------------------------------------------------------- Two thoughts occur: 1. Are some of the stations close enough so that you can daisy chain them together using Thinwire Ethernet cable before hitting the twisted-pair adapter? (I believe that one of the technical types at DECWORLD said that is permissible.) 2. I believe that I saw a blurb from AMP saying that they had an undercarpet Thinwire Ethernet ribbon cable. I would derate the heck out of it, but you may want to check with them. ================================================================================ Note 34.12 Ethernet transceivers and cables 12 of 22 EISNER::LEDERMAN "Bart Z. Lederman" 14 lines 8-SEP-1989 11:00 -< Does the DESTA work the other way? >- -------------------------------------------------------------------------------- We need to hook up another machine (a MAC with an Ethernet board) to our existing equipment. We have a DELNI so we have some extra transceiver ports available. We would like to use thin wire, but DEC's DEMPR 8 port device would break the budget, and gives us more ports than we need. The DESTA single port repeater looks like the right sort of thing to use, but all illustrations of it show it being used to go from thin wire to a machine (like a Micro-VAX). Has anyone ever hooked this up the other way (from a DELNI to thin wire)? Is it supposed to be able to work this way? If not, is there an inexpensive device that would do this? Our alternative is to run a tranceiver cable from the DELNI to the MAC, which can accept the cable, but we would not be able to daisy chain to other MACs in the future if we wanted to. ================================================================================ Note 34.13 Ethernet transceivers and cables 13 of 22 EISNER::KILGALLEN "Larry Kilgallen" 6 lines 8-SEP-1989 12:13 -< Is your transceiver port free on the DELNI? >- -------------------------------------------------------------------------------- The device provides the power to the tranceiver via the tranceiver cable. The DESTA (a tranceiver-replacement) expects that power. The DELNI provides that power only out the connector which is supposed to go to the transceiver. If that is where you are going to connect the DESTA there should be no problem. If you were going to connect it to one of the eight normal ports, it will not work. ================================================================================ Note 34.14 Ethernet transceivers and cables 14 of 22 EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 3 lines 8-SEP-1989 13:37 -< DESTA is not a DESPR >- -------------------------------------------------------------------------------- There are many third party DEMPR and DESPR (the single station repeater) equivalents on the market that are cheaper than the DEC versions. A DESTA is *not* a repeater. ================================================================================ Note 34.15 Ethernet transceivers and cables 15 of 22 EISNER::MCMICHAEL "Chuck McMichael" 11 lines 8-SEP-1989 13:43 -< DESPR to the Rescue! >- -------------------------------------------------------------------------------- > We need to hook up another machine (a MAC with an Ethernet board) > to our existing equipment. We have a DELNI so we have some extra > transceiver ports available. We would like to use thin wire, but > DEC's DEMPR 8 port device would break the budget, and gives us more > ports than we need. The DESTA single port repeater looks like the > right sort of thing to use, but all illustrations of it show it What you're looking for is not a DESTA, but a DESPR, which is the single-port version of the DEMP If you want to hook it to a DELNI, you need to have the DELNI connected to the standard Ethernet using an H4000-BA transceiver. (I think that's the one where field service cuts out the heartbeat.) ================================================================================ Note 34.16 Ethernet transceivers and cables 16 of 22 EISNER::LEDERMAN "Bart Z. Lederman" 10 lines 8-SEP-1989 15:12 -< I get the picture now. >- -------------------------------------------------------------------------------- Now that I've read the answers, the meaning of a 'repeater' has finally sunk in. What I'm looking for is a single port repeater. I can now go and look up the DESPR. The DELNI is connected to a thick-wire Ethernet, though not through any model of H4000. There is a lot of non-DEC equipment on the cable, and the department responsible for it uses some other companies taps. (I don't know the brand). Thanks. ================================================================================ Note 34.17 Ethernet transceivers and cables 17 of 22 EISNER::LEDERMAN "Bart Z. Lederman" 7 lines 11-SEP-1989 08:22 -< Priced out of our range. >- -------------------------------------------------------------------------------- Well, I looked up the DESPR on the DEC Electronic Store, and it's about $1,000. If we could afford that, we'd probably go ahead and buy the 8 port version, so we probably won't buy either. We'll either look for a less expensive (third party) repeater, or go with a regular tranceiver cable from the DELNI to the MAC. Thanks for the information. ================================================================================ Note 34.18 Ethernet transceivers and cables 18 of 22 EISNER::RICE "Gary Rice" 17 lines 12-SEP-1989 10:46 -------------------------------------------------------------------------------- > Well, I looked up the DESPR on the DEC Electronic Store, and it's > about $1,000. If we could afford that, we'd probably go ahead and... Well, how about a $4.00 solution? At your favorite Electronic supply house you can obtain a "Thick-wire to BNC" connector. DEC refuses to believe that these things exist (and even if they did, they are CERTAINLY unsupported). We have one and it works just fine. Only drwaback is it confuses the rules about cable distances. But if you remember to multiply the length of the thin-wire segments by 4 and use the results in your length calculations, you can get by. Also, you need to have access to the END of the thick-wire. It is NOT a Tee. Gary ================================================================================ Note 34.19 Ethernet transceivers and cables 19 of 22 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 23 lines 12-SEP-1989 22:16 -< N female to BNC female connector needed >- -------------------------------------------------------------------------------- >you can obtain a "Thick-wire to BNC" connector. DEC refuses to believe that The 'HOSE' connectors are "N" and are MALE, so, if you ask for a FEMALE "N" to FEMALE "BNC" connector, you should get what you want. The old 50 ohm "N" FEMALE terminator you are removing can be most easily replaced with a "BNC" one at the end of your thinwire rather than converting back to "N" size. Note that most "BNC" terminators are MALE, as they are normally used to cap the last TEE which is generally FEMALExFAMALExMALE (MALE is the side port to your DESTA or controller card). So you will need a FxF barrel connector or a tee at the end to add the "BNC" terminator. If the HOSE is sacred and belongs to some other department, they may not like the idea of a thinwire extension, as that makes it very easy for users to mess up the backbone. Your DOT location in TECH square area puts you very near ELI's up near Inman Square on Hampshire St. If you havn't found it yet, DO! We paid a tad over 2k for a DEMPR there. All kinds of electronic 'junk', but they are DDA members, and DO know what the DEC stuff is worth. ================================================================================ Note 34.20 Ethernet transceivers and cables 20 of 22 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 9 lines 12-SEP-1989 22:45 -< Like a kid in a candy store >- -------------------------------------------------------------------------------- > Your DOT location in TECH square area puts you very near ELI's > up near Inman Square on Hampshire St. If you havn't found it > yet, DO! This is indeed a wonderful, er, place. I used to live only a few blocks away and practically haunted the place. This way I was able to get the best stuff first. It was, as a good friend pointed out, "within walking distance on the way over, but not on the way back." ================================================================================ Note 34.21 Ethernet transceivers and cables 21 of 22 EISNER::TACKETT "Galen Tackett" 4 lines 16-SEP-1989 19:40 -< Need info on heartbeat requirements >- -------------------------------------------------------------------------------- I've had trouble finding out from DEC the heartbeat requirements of various Ethernet devices. I suspect this has been discussed elsewhere on DECUServe. Could someone point me in the right direction? ================================================================================ Note 34.22 Ethernet transceivers and cables 22 of 22 EISNER::HARVEY "Jack Harvey" 5 lines 17-SEP-1989 09:58 -< DEC_NETWORKING 172 "Heartbeat" >- -------------------------------------------------------------------------------- >I suspect this has been discussed elsewhere on DECUServe. Indeed it has. Also, for a wealth of other discussion: SEARCH DECUSERVE_EXTRACTS:ALL_TITLES.TXT HEARTBEAT  ================================================================================ Note 35.0 If I had more memory, I might not forget so much 2 replies EISNER::TILLMAN "Brian Tillman, Lear Siegler, Inc." 7 lines 31-AUG-1987 14:41 -------------------------------------------------------------------------------- We currently have three 8600s with 16M, 20M, and 24M memory, each supporting about 70 logins (about 110 for BALSETCNT). We feel we should add memory, but are unsure of the measure of performance improvements we would expect to see, and without some measure, upgrades are not salable to management. We would expect to see some paging improvement (we don't see any swapping) and some disk improvement because we would increase our disk caching. Anyone out there with a resonable estimate? ================================================================================ Note 35.1 If I had more memory, I might not forget so much 1 of 2 EISNER::DROPPERS "Seton R Droppers" 15 lines 31-AUG-1987 20:59 -< Attempting to quantify response... >- -------------------------------------------------------------------------------- Your asking a DEC system to give you quantified answers... The best that I can give you is seat of the pants stuff from experience. We were running a 8600 with 16Mb and could get acceptable response with a load of 50 to 60 users. When we upgraded to 32Mb we were able to run about 70 to 80 users with the same response, the measure we had was more users at the same response, not increased response. We seem to see better response with about 30 to 50 users, then it levels off to about 70 users. P.S. We have since upgraded to 64Mb and 8650, we can now run 80 or more users, though I think the limit will be about 100. P.P.S. We might see better stats if we tuned harder than we have, particularly the IO. ================================================================================ Note 35.2 If I had more memory, I might not forget so much 2 of 2 EISNER::ERSKINE "Chris Erskine" 14 lines 1-SEP-1987 07:46 -< Additional Help >- -------------------------------------------------------------------------------- EMC has put together a manual called 'System Performance Tuning on DEC VAX/VMS'. They show a list price on the manual of $150 but I do not know of anyone who has had to buy it. They tend to give it out free of charge. Since they are in the memory business, the manual is of the flavor 'To increase performence, add memory'. They actually have done a good job for the basic memory type of tuning. If you can look at foreign memory for your system, if you talk to them or some of the other companies, they have supplied evaluation modules which can be used to help justify the performence gains. By taking hard numbers to management that show you need more memory, there is not much they can say.  ================================================================================ Note 36.0 PORTABLE TERMINALS 5 replies EISNER::KILLEEN "Jeff Killeen" 1 line 31-AUG-1987 22:03 -------------------------------------------------------------------------------- This topic is to discuss portable terminals ================================================================================ Note 36.1 PORTABLE TERMINALS 1 of 5 EISNER::KILLEEN "Jeff Killeen" 27 lines 31-AUG-1987 22:06 -< RANDOM CORP. - COLLEAGUE PLUS KEYPADS >- -------------------------------------------------------------------------------- This is the keypad layout of the control function keys on the Random Corp. Colleague Plus terminals. EDT KEYPAD F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 GOLD HELP FNDNXT DEL L ENTER DEL W DEL C SELECT FIND UND L SUBS UND W DEL C RESET F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 WORD EOL CHAR ADVANCE BACKUP CUT PAGE SECTION APPEND LINE C-CASE DEL EOL SPECINS BOTTOM TOP PASTE COMMAND FILL REPLACE O-LINE -------------------------------------------------------------------------------- NOTES KEYPAD F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 GOLD HELP NUM/CMD ONE_KEY PREV-N NEXT-UN PREV-S F11 F12 F13 F14 F15 F16 F17 F18 F19 F20 <--- | ---> ^ SELECT v | ================================================================================ Note 36.2 PORTABLE TERMINALS 2 of 5 EISNER::HANRAHAN "Jamie Hanrahan" 4 lines 15-SEP-1987 16:20 -< GRiD Compass VT100 emulation? >- -------------------------------------------------------------------------------- Anybody have experience using GriD Compass systems (e.g. 1101) as portable terminals, with the VT100 emulator? They're available used now for a song. And they run MS-DOS (though they're not PC compatible). ================================================================================ Note 36.3 PORTABLE TERMINALS 3 of 5 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 9 lines 16-SEP-1987 08:49 -< Yes we have some GriDs >- -------------------------------------------------------------------------------- We have several here. While the systems seem fairly good, I'm really turned off by the company. Back when we were considering these beasts, a sales rep came in to demo one. He didn't know the first thing about the machine. In fact, he had his demo stored in bubble memory and in his screwing around to figure out his machine (with which he wasn't having a whole lot of luck; I knew more about it from five minutes playing than he did), he erased the bubble memory. Bye bye demo! Anyway, is there something specific you'd like to know? We use our GriDs in VT100 emulation with MS-Kermit. ================================================================================ Note 36.4 PORTABLE TERMINALS 4 of 5 EISNER::EARLE "David Earle" 19 lines 18-JUN-1988 14:58 -< Problems with Random Colleague >- -------------------------------------------------------------------------------- I have a Random Colleague (not Plus), and have had some problems with it, such as: - On power up (or power down), I lose the program sequence in function keys 6 thru 8. On power up, I have to re-program. And I do save the channel data. - I also lose other setup attributes, such as auto repeat, repeat rate and audio level when power down occurs. Is there something I should be doing? - I never got the 'good' owners manual because I have an early model. Does any one else have one? - Does anyone know of an 'upgrade' that is availible to expand the function keys? Have you used it and is it worth it? Price and contact? Any other info would be helpful. Otherwise the terminal is a good one. ================================================================================ Note 36.5 PORTABLE TERMINALS 5 of 5 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 24 lines 18-JUN-1988 16:46 -< SOUNDS LIKE YOUR TERMINAL NEEDS SERVICE >- -------------------------------------------------------------------------------- > - On power up (or power down), I lose the program sequence in function > keys 6 thru 8. On power up, I have to re-program. And I do save the > channel data. > > - I also lose other setup attributes, such as auto repeat, repeat > rate and audio level when power down occurs. Is there something > I should be doing? If you are doing a save channel data then you are doing the right thing. I am sitting here typing on a RANDOM Colleague Plus and I have seen none of these problems. The only known problem is the VT52 setting does *NOT* work correctly. I would call them. > - I never got the 'good' owners manual because I have an early model. > Does any one else have one? It is worth getting a new owners manual just for the keypad overlay cards. > - Does anyone know of an 'upgrade' that is availible to expand the > function keys? Have you used it and is it worth it? Price and contact? Sounds like a hardware fix is needed - not an upgrade.  ================================================================================ Note 37.0 HARDWARE NEWS FLASHES 2 replies EISNER::KILLEEN "Jeff Killeen" 2 lines 1-SEP-1987 21:32 -------------------------------------------------------------------------------- This topic will be used for news flashes and other hot information ================================================================================ Note 37.1 HARDWARE NEWS FLASHES 1 of 2 EISNER::KILLEEN "Jeff Killeen" 14 lines 1-SEP-1987 21:39 -< PRICE BOOK WARNING!!!!! >- -------------------------------------------------------------------------------- DEC repriced the following on August 3RD. 1. RA, RC, RD disk drives 2. PDP-11 systems 3. TK Tape drives 4. controllers for above It is a 5 percent *INCREASE* ! The price lists that were shipped with the quarterly update of the Digital Reference Service are the *OLD* price lists. The updated price lists is dated August 3, 1987. The price list dated June 29, 1987 is *VERY* obsolete. ================================================================================ Note 37.2 HARDWARE NEWS FLASHES 2 of 2 EISNER::KENNEDY "Terry Kennedy" 21 lines 1-SEP-1987 23:01 -< Curiouser and curiouser... >- -------------------------------------------------------------------------------- I found out about this one myself, the hard way. Oddly enough, DEC's Add-on and Consulting Group quoted me the pre-increase price, but DEC Direct quoted the post-increase price on the same day. However, all of this applies to users without special agreements, which is the majority. Even if you qualify, DEC salespeople usually won't tell you about any special discounts. Over the last 6 years, I have had 4 official DEC salesreps for my account. Only the newest one told me that because my organization is a school, we are eligible for some substantial discounting. When I asked if this was new, she said that it has been like this for years. Also, above and beyond these discounts, there is special educational pricing for some software (mainly popular VAX languages). The best advice is to jump up and down to get the best price. DECdirect will only quote you the LIST price of an item. By the way, in my case, the agreement states that our pricing is from the Q2 catalog (September), less 14%, for the whole year, regardless of any price increases. If the prices go down, our price goes down, too. But all orders have to be placed through the local sales office. The moral of the story is, 'You get what you argue for'.  ================================================================================ Note 38.0 DECserver/DELNI problem 7 replies EISNER::NORTON 15 lines 2-SEP-1987 09:26 -------------------------------------------------------------------------------- We have 8 DECserver 100's connected to a DELNI, which hangs off an H4000. On power-up all the servers report a non-fatal error, and the little green lights on the front panels never stop blinking. The servers seem to function OK, but their 'heartbeat absent' counters all increment like mad. DEC field service says the heartbeat in this case comes from the DELNI, but replacing it and powerfailing the servers doesn't help. Any suggestions? ================================================================================ Note 38.1 DECserver/DELNI problem 1 of 7 EISNER::RICE "Gary Rice" 9 lines 2-SEP-1987 10:59 -< Check the H4000 >- -------------------------------------------------------------------------------- > The servers seem to function OK, but their 'heartbeat absent' counters > all increment like mad. Are you SURE that you used an H4000 transceiver? It sounds to me like you have an H4000-BA form of the transceiver instead. Gary ================================================================================ Note 38.2 DECserver/DELNI problem 2 of 7 EISNER::TILLMAN "Brian Tillman, Lear Siegler, Inc." 11 lines 2-SEP-1987 14:09 -< I'll bet -BA >- -------------------------------------------------------------------------------- RE: < Note 38.1 by EISNER::RICE "Gary Rice" > -< Check the H4000 >- >> The servers seem to function OK, but their 'heartbeat absent' counters >> all increment like mad. >Are you SURE that you used an H4000 transceiver? It sounds to me like >you have an H4000-BA form of the transceiver instead. I agree. The heartbeat is NOT supplied by the DELNI (at least when it is attached to a transceiver). The transceiver supplies it. ================================================================================ Note 38.3 DECserver/DELNI problem 3 of 7 EISNER::KENNEDY "Terry Kennedy" 12 lines 2-SEP-1987 20:23 -< DELNI / heartbeat >- -------------------------------------------------------------------------------- > I agree. The heartbeat is NOT supplied by the DELNI (at least when it > is attached to a transceiver). The transceiver supplies it. You need to be very cautious about using DELNI's with devices which generate/expect heartbeat. The DELNI passes the heartbeat along to all connected devices, without passing the signal which says that it IS a heartbeat. I suspect you don't really have a problem at all. However, it is best to either configure the whole network with or without heartbeat. 50/50 will cause you management headaches later. Terry ================================================================================ Note 38.4 DECserver/DELNI problem 4 of 7 EISNER::NORTON 8 lines 3-SEP-1987 09:09 -< Really really IS H4000 >- -------------------------------------------------------------------------------- >>Are you SURE that you used an H4000 transceiver? It sounds to me like >>you have an H4000-BA form of the transceiver instead. Yes, it really is an H4000. I'll try swapping that next after-hours. For some strange reason everybody got upset when I swapped the DELNI during first shift ;-} ================================================================================ Note 38.5 DECserver/DELNI problem 5 of 7 EISNER::NORTON 14 lines 17-SEP-1987 10:01 -< Eventual success - why? >- -------------------------------------------------------------------------------- > I'll try swapping that next after-hours. After swapping with another H4000, the problem still existed. About the only thing not changed at this point was the transceiver cable. The cable turned out to be two 5-meter BNE3's in series. After replacing them with a single 20-meter cable, the problem disappeared :-). Could a broken line in one of the cables clobber JUST the heartbeat signal? Or, is having two cables in series a no-no, even if they're not BNE4's? ================================================================================ Note 38.6 DECserver/DELNI problem 6 of 7 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 4 lines 17-SEP-1987 14:54 -< Not surprising >- -------------------------------------------------------------------------------- I've seen occasions where there was a connection problem using piggy backed cables as you've described where the node came up and you could SET HOST out from it but not into it. All kinds of strange things can occur with slightly faulty cable connections. ================================================================================ Note 38.7 DECserver/DELNI problem 7 of 7 EISNER::MERRIMAN "George Merriman - CCA/NY" 1 line 6-SEP-1988 20:56 -< also see notes 62.* in DEC_NETWORKING >- --------------------------------------------------------------------------------  ================================================================================ Note 39.0 Configuring a Microvax II File Server 2 replies EISNER::WALLIS "Barry Wallis - Fleetwood Enterprise" 39 lines 2-SEP-1987 12:45 -------------------------------------------------------------------------------- I have posted the following note in the MicroVAX_Systems conference. If anyone can help me, please post your replies there. ================================================================================ <<< EISNER::DUA0:[NOTES$LIBRARY]MICROVAX_SYSTEMS.NOTE;2 >>> -< MicroVAX_Systems >- ================================================================================ Note 17.0 Configuring a Microvax II File Server No replies EISNER::WALLIS "Barry Wallis - Fleetwood Enterprise" 28 lines 2-SEP-1987 12:13 -------------------------------------------------------------------------------- We are considering stand-alone workstations for our new CAD system. I have the following questions (which may be naive) concerning the intial hardware configuration for a Microvax II used strictly as a file server and 28 Vaxstation 2000s used as stand-alone CAD workstations: 1) How much memory is needed on the Microvax II? Is this different depending on whether the Vaxstations are diskless or diskfull? 2) What is the minimum / maximum amouont of disk space needed on the Microvax II for VMS (I know it is tailorable and I expect a range of answers to this questions)? 3) What would the "recommended" sizes be for the paging and swapping files if the Vaxstations are diskless? What about their sizes for the Vaxstations if they are diskfull? ***** And, we thank you for your support ***** Barry L. Wallis Fleetwood Enterprises, Inc. 3125 Myers Street Riverside, CA 92523 (714)351-3682 ================================================================================ Note 39.1 Configuring a Microvax II File Server 1 of 2 EISNER::ROECKEL 21 lines 15-SEP-1987 14:00 -< System Disk Size >- -------------------------------------------------------------------------------- >> 2) What is the minimum / maximum amouont of disk space needed on the >> Microvax II for VMS (I know it is tailorable and I expect a range >> of answers to this questions)? I don't know if this will help you out, but our MicroVAX II had an RD53 (70 Meg) drive for the system. We ran out of 'spare' disk space (required for loading updates) and had to go to an RD54 (159 Meg.) We have the following (roughly) on the system disk: VMS FMS Fortran77 Basic (w/ compiler) Pascal DECnet-VAX (w/ DECserver 100's & 200's) That about all that comes to mind at the moment. Bruce. ================================================================================ Note 39.2 Configuring a Microvax II File Server 2 of 2 EISNER::PERRY "Bob Perry - NLC - Skydiving Madman" 15 lines 20-OCT-1987 15:18 -< BIG little Vax's >- -------------------------------------------------------------------------------- We're about to create a Local Area Vax Cluster consisting of 3 uVAX II's. I intend to put 2-380 Mb maxtor drives on each control{_led by an Emulex QD21 controller. If I had the capitol(al?) available I'd have a total of 4 Maxtors on each system. They're quite inexpensive, and we've been using this disk/controller combination inside the company for about 2 years now. Oh, I forgot to mention that this LAVC will be used for CAD/CAM research (solids modeling, thermal analysis, etc). Fairly heavy stuff. We typically (will) have 5 or 6 users on a system. We currently have of the uVAX's in operation, although not clustered. Since it's relatively inexpensive, you should add as much static and rotating memory as you can afford. Our machines, typically, have 5-9 Mb of memory. BTW, the Emulex system is available for under $6,000. That's $11,000 for 1.2 Gb of storage and it all fits in the BA123 box.  ================================================================================ Note 40.0 PERIPHERAL DEVICE ANNOUNCEMENTS 56 replies EISNER::KILLEEN "Jeff Killeen" 2 lines 5-SEP-1987 12:03 -------------------------------------------------------------------------------- This topic will be used discuss peripheral device annoucements. All none system hardware devices. ================================================================================ Note 40.1 PERIPHERAL DEVICE ANNOUNCEMENTS 1 of 56 EISNER::KILLEEN "Jeff Killeen" 34 lines 5-SEP-1987 12:24 -< BYE BYE VT200 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! VT200 FAMILY RETIREMENT - 25-August-1987 o VT200 products are being retired o Final inventories are limited o Final orders now being accepted The VT220 text terminals and the VT240/VT241 graphics terminals are being retired. The VT300 family (VT320, VT330 and VT340) are the replacement products. Current Product Replacement Product VT220 VT320 VT240 VT330 VT241 VT340 All customers, especially OEMs, government contracts and large end users, should be encouraged to plan an orderly transition to the VT300 products now to avoid disruptions in their equipment procurement. There is no need to change existing VT200 orders. However, the VT300 family should be quoted for new orders. Final VT200 inventories are limited. Customers who anticipate an ongoing requirement for the VT200 products should be encouraged to submit orders now for their future needs, as Digital will not hold inventory. ================================================================================ Note 40.2 PERIPHERAL DEVICE ANNOUNCEMENTS 2 of 56 EISNER::KILLEEN "Jeff Killeen" 183 lines 5-SEP-1987 12:26 -< HELLO VT320 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! INTRODUCING THE VT320 TEXT VIDEO TERMINAL - 25-August-1987 The single-session VT320 is based on a new, two-piece, ergonomic design that is smaller than the already-trim VT220. All electronics for the terminal are housed in the same enclosure as the monitor. The on/off switch is mounted on the side of the monitor housing, within easy reach for the user. Similarly, the keyboard now connects to the side of the terminal. The VT320 incorporates a low-profile, integral tilt mechanism, built into the bottom of the terminal; a pedestal base with tilt and swivel capabilities is optional (VT3xx-CA). The VT320 video display is a 14-inch, antiglare, flat-surfaced monochrome screen, available in a new "paper white" phosphor, as well as in amber and green. The "paper white" phosphor, which features black characters on a white background, closely resembles the appearance of a printed page. All three display screens -- amber, green and paper white -- feature near- letter-quality fonts with crisp, fully formed characters. The screen resolution has been improved to 1200 pixels by 300 scan lines (in 80-column mode). In 132-column mode, the VT320 screen resolution is 1188 pixels by 300 scan lines. Fully compatible with the VT220, the VT320 offers several enhancements. These include a terminal/host-programmable 25th status line, improved keyboard features and a bidirectional printer port with modular DECconnect connector. The 25th display line is used to inform the operator of the terminal or host application status. The host system, under program control, may dynamically alter the use of the 25th status line by changing terminal setup to reflect the needs of the operating environment and/or the user application. Improved keyboard features include the following: o Compose key disable o Selectable backspace/delete key o User-configurable escape key option o Keyboard selectable brackets < > for data processing or typist mode With the introduction of the VT320, Digital now offers a completely new generation of video terminals -- text only, monochrome text and graphics, and color text/graphics -- to serve a full range of applications needs. The VT300 family offers complete compatibility with Digital's systems and networks. Therefore, connectivity to Digital's networking products -- gateways, clusters, terminal servers and external databases -- is assured. COMMUNICATIONS/CONNECTIVITY/MODEL INFORMATION There are three models of the VT320 available in the U.S.: the North American VT320, the International VT320 for North America, and the International VT320 for Australia. These models support Digital's DECconnect wiring strategy and provide bidirectional serial printer ports, which use the new Digital MMJ modular connectors. The VT320 does not support video out or 20 mA loop connectors. North American VT320 The North American VT320 (VT320-xx) provides a single DEC423 MMJ communication interface port for connection to a host system or terminal server. Also provided is a DEC423 to RS232 adapter with a 10-foot DEC423 DECconnect office cable (BC16E-10). This adapter cable allows easy connection to 25- pin serial cables, for those users who have not yet made the move to DECconnect. Modem control is through the DEC423 cable and 25-pin adapter for standard U.S. modems. It has an attached power cord that uses 120 volts and the English keyboard. The North American model has UL, CSA, and FCC Class A certification. This model will meet the needs of most customers. International VT320 for North American Use The International VT320 for North American use includes both DEC423 and RS232 communication interface ports. Full modem control is available through the RS232 25-pin connector. A DEC423 DECconnect office cable and DEC423 and RS232 adapter are not included. This 120-volt model features the English keyboard, a detachable power cord, multiple language set-up screens, and supports the NRCS (National Replacement Character Set). It qualifies for VDE, UL, CSA, GS, and FCC Class B certification. North American customers who require the higher FCC certification or do not wish to use the adapter for RS232 interface can use this model. International VT320 for Australian Use The International VT320 for Australia is a 240-volt model of the International model for North America. VT300 FAMILY PRODUCT POSITIONING The VT320 is a single session, basic text terminal for general-purpose computing -- designed to meet the needs of the Digital system user. The VT330 is Digital's premier monochrome text and graphics terminal. It should be sold to users who require high-performance, high-resolution graphics, as well as high-quality text display capabilities like dual sessions. The VT340 is Digital's premier color text and graphics dual-session terminal. It should be sold to users who require high-performance, high-resolution, full-color graphics, as well as high-quality text display capabilities. SOFTWARE SUPPORT The VT320 is fully supported by a complete range of operating systems and applications software from Digital and other vendors. Any software that currently supports text-only terminals in the VT100/VT200 families, or the VT52 video terminal, will support the VT320. VMS V4.6 fully supports the VT300 series terminals, including the VT320. VT300 series terminals can be used with earlier releases of VMS if they are set up as VT200 series terminals. PRICING/ORDERING INFORMATION North American Models VT320 120 V, with fixed power cord, DEC423 MMJ communication interface port, DEC423 to RS232 adapter, 10-foot DECconnect office cable, UL, CSA, and FCC Class A certification, documentation w/standard English Keyboard w/WPS English Keyboard MLP BMC VT320-AA white VT320-DA white $545 $3 VT320-BA green VT320-EA green VT320-CA amber VT320-FA amber International Models for North American Use VT320 120 V, with detachable power cord, DEC423 MMJ and RS232 communication interface ports, VDE, UL, CSA, GS and FCC Class B certifications, documentation w/standard English Keyboard w/WPS English Keyboard MLP BMC VT320-GA white VT320-NA white $595 $3 VT320-HA green VT320-PA green VT320-JA amber VT320-RA amber International Models for Australian Use VT320 240 V, with detachable power cord, DEC423 MMJ and RS232 communication interface ports, VDE, UL, CSA, GS and FCC Class B certifications, documentation w/standard English Keyboard w/WPS English Keyboard MLP BMC VT320-AZ white VT320-DZ white $595 $3 VT320-BZ green VT320-EZ green VT320-CZ amber VT320-FZ amber Accessories MLP VT3xx-CA VT320 Tilt/Swivel Base $20 IMPORTANT -- WHEN SELLING THE VT320 WITH SYSTEMS, TERMINAL SERVERS AND CPU UPGRADES, REDUCE THE PRICE BY $50 PER UNIT! SUPPORT SERVICES The VT320 is supported by a full line of service offerings, including a one- year "return-to-factory" hardware warranty (upgradable to on-site), BASIC, DECservice and Carry-In. Basic Monthly Charges (BMC) are $3. Prices for DECmailer and FRU Exchange are $50 and $80, respectively. AVAILABILITY Orderable at announcement; available in October 1987. ================================================================================ Note 40.3 PERIPHERAL DEVICE ANNOUNCEMENTS 3 of 56 EISNER::KILLEEN "Jeff Killeen" 44 lines 5-SEP-1987 14:11 -< LNO3 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! LOWER PRICES FOR LN03 AND LN03 PLUS LASER PRINTERS - 02-September-1987 o LN03 Reduced 17% o LN03 PLUS reduced 20% o One year on-site warranty included o LN03 PLUS Upgrade Kit reduced 31% Effective immediately, prices have been reduced on the LN03, LN03 PLUS, and LN03 to LN03 PLUS Upgrade Kit. The LN03 and the LN03 PLUS both print at 300 x 300 dots per inch, up to eight pages per minute, and are rated at up to 10,000 pages per month usage. The LN03 is ideal for general-purpose printing and word processing applications requiring text and simple business graphics output. The LN03 PLUS contains additional memory to support full-page bit map graphics used in CAD/CAM drawings, and scientific/technical environments requiring integrated text and graphics applications. PRICING/ORDERING INFORMATION Model No. Description Old Price New Price LN03-AA 8 ppm Desktop Laser Printer $3,495 $2,895 LN03S-AA Full Graphics, 8 ppm Laser Printer 4,995 3,995 LN03S-UA LN03 to LN03S Upgrade Kit 1,595 1,095 LN03-UA LN03 or LN03S to LN03R Upgrade Kit 2,495 2,995* * Price increase. The ScriptPrinter (LN03R) pricing was not reduced. This increase maintains the price differential between the LN03 and LN03R. Prices are effective now. All standard Digital discounts apply. ================================================================================ Note 40.4 PERIPHERAL DEVICE ANNOUNCEMENTS 4 of 56 EISNER::KILLEEN "Jeff Killeen" 168 lines 5-SEP-1987 14:12 -< LJ250/LJ252 COLOR PRINTERS >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! LJ250/LJ252 COMPANION COLOR PRINTERS - 25-August-1987 o Low-cost, leadership price/performance and high-quality printing o Supported by a full range of Digital and non-Digital operating systems and applications software o Fast, quiet, trouble-free disposable cartridge technology o Available September 1987 Announcing the LJ250 and LJ252 Companion Color Printers, two state-of-the-art, highly reliable color printers that combine fast, high-quality color text and graphics printing capabilities in a single desktop unit for Digital system users. The LJ252 parallel interface model allows users of IBM PCs, PC compatibles and other non- Digital systems to enjoy these same superior color text/ graphics printing capabilities. PRODUCT DESCRIPTION The Companion Color Printers can print up to seven pure colors -- including black -- and up to 255 mixed colors. Text printing is high-quality, while graphics display clear, bright colors with remarkable clarity and definition. The Companion Color Printers' print resolution, at 180 by 180 dots- per-inch, is better than most dot-matrix printers. Color graphics are fast, at typically 2-4 minutes per page (depending on image complexity). For black text printing, the speed is even greater: 167 characters per second (char/s) burst speed, 90 char/s throughput speed. And for both printing functions, the Companion Color Printers are unbelievably quiet -- less than 45 dBA. (It is even quieter than the LN03 laser printer!) The LJ250/LJ252 print text and graphics on paper and transparencies in the A (8.5" x 11") and European A4 (8.3" x 11.7") sizes, using single-sheet (friction) feeding or fanfold (sprocket) paper feeding. Specially treated paper and transparencies are available from Digital to help customers achieve optimum print quality and reliability from the Companion Color Printer's ink-jet technology. Print cartridges are integrated and incorporate the ink supply system with the print heads. Thus, the print heads are automatically replaced at the same time as the ink supply system, a snap- out/snap-in procedure that takes the user just seconds. This advanced, highly reliable design keeps the Companion Color Printers running at peak performance with minimal effort. Separate print heads for color and black inks ensure that supplies need not be replenished any more often than necessary. SOFTWARE SUPPORT o Compatible with Digital's LA75, LA210, LN03, LA50 and LA100 printers. o Full printing support for ReGIS-based graphics applications through the newly developed RETOS (ReGIS to Sixels Converter) host-based software (see supporting article in this issue entitled "Introducing RETOS -- VAX ReGIS to Sixels Converter). o Supported by many Digital operating systems and graphics utilities, including: VMS, Hardcopy UIS, VWS, ULTRIX-32 and RETOS. Also offered: o Built-in support for the Hewlett Packard HP-PCL protocol. Extensive software support from a library of Hewlett Packard applications and many popular third-party software packages is available now, allowing these packages to be used immediately on the VAXmate. When using the Companion Color Printer in the HP-PCL protocol mode, the user should select appli- cation software packages that support the HP-PCL Protocol, ensuring that the printer driver within these packages supports either the Digital Companion Color Printer or the HP PaintJet (HP3630A). Following is a list of software packages that support the HP-PCL Protocol. This list is for information purposes only and it is recommended that the user verify support with the software supplier. These software products have not been tested nor qualified by Digital and this list does not constitute an endorsement of any independent software by Digital. Further, all trademarks applicable to the following named software packages implied or otherwise are enacted. Word Processing/Desktop Publishing: Hewlett Packard's Executive MemoMaker V A.02; Microsoft Corporation's Microsoft Word V3.1; Multimate V3.31; Ashton- Tate's Multimate Advantage 2 V1.0; Software Publishing Corporation's PFS: Professional Write V1.04; Samna Corporation's Samna Word V4.0; Lifetree's Volkswriter 3 V1.0 and Volkswriter Deluxe+ V1.0; Word Perfect's Word Perfect V4.2; MicroPro International Corporation's Word Star 2000 V2.0. Business and Presentation Graphics: Boeing Computer Services' Boeing Graph V1.0; Ashton-Tate Graphic Product Center's Chart-Master V6.2; Computer Support Corporation's Diagraph V4.0; Media Cybernetics' Dr. Halo II ; Enertronics Research, Inc.'s EnerCharts V2.01 and Energraphics V2.01; Hewlett Packard's Graphics Gallery VA.01; Lotus Development Corporation's FreeLance Plus V2.0; Media Cybernetics' Halo DPE; Software Publishing Corporation's Harvard Graphics V2.0; Microsoft Corporation's Microsoft Chart V2.02; Media Cybernetics' Nimbus; Computer Support Corporation's Picture Perfect V4.0; Ashton-Tate Graphic Product Center's Sign-Master V5.1; Advanced Graphics Software's SlideWrite Plus V2.0; SPSS, Inc.'s SPSS Graphics V1.1. Technical Graphics: Autodesk, Inc.'s AutoCAD V2.52; Cadkey's CADKEY V2.11; Forsight Resources' Drafix I V1.0 and Drafix I Plus V2.0; Pacsoft's DTM-9000 V1.0; American Small Business Computer's Pro Design II V2.5; Computer Associates International's SuperProject Plus V2.0; Versacad Corporation's VersaCAD Advanced V5.2; Viewlogic Systems, Inc.'s Workview V2.1. Productivity: Lotus Development Corp.'s Lotus 1-2-3 V2.01; Ashton-Tate's Framework II V1.1; Word Perfect Corp.'s MathPlan V3.0; Microsoft Corp.'s Multiplan V3.02; Ashton-Tate's RapidFile V1.0; Lotus Development Corp.'s Report Writer V1 and Symphony V1.2. SUPPORT SERVICES The Companion Color Printers are supported by a full line of service offerings, including a one-year, carry-in warranty, BASIC Service, DECservice and Carry-in Service. Basic Monthly Charges (BMC) are $11. Prices for DECmailer and FRU Exchange are $175 and $262.50, respectively. Self-Maintenance Services Services and products will be available for self-maintenance customers from the Self-Maintenance Services Group. Please refer to the "helpful services" section of the current Self-Maintenance Services Price Book for detailed information, or contact their support line at (603) 884-5000. PRICING/ORDERING INFORMATION Model No. Description MLP BMC Seg. LJ250-CA Companion Color Printer $1,695 $11 XX Serial Interface LJ252-CA Companion Color Printer 1,695 11 XX Parallel Interface Accessories and Supplies: Option No. Description MLP LJ25x-AB Black Ink Cartridge $30 LJ25x-AA Color Ink Cartridge 38 LJ25x-DS Acrylic Desk Stand 40 LJ25x-DC Dust Cover 5 LJ25x-SW Paper Catcher 15 LJ25x-AC 8.5" x 11" Continuous Form 18 Paper LJ25x-AD 8.3" X 11.7 A4 Size 20 Continuous Form Paper LJ25x-AE 8.5" x 11" Cut Sheet Transparencies 69.50 LJ25x-AF 8.3" x 11.7" A4 Size Cut Sheet Transparencies 70.50 AVAILABILITY September 1987. ================================================================================ Note 40.5 PERIPHERAL DEVICE ANNOUNCEMENTS 5 of 56 EISNER::KILLEEN "Jeff Killeen" 183 lines 5-SEP-1987 14:13 -< LCG01 VS LJ250 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! LCG01/LJ250 COLOR PRINTER POSITIONING GUIDE - 25-August-1987 o A guide for matching the printer to the customer's application The LCG01 Color Printer and the LJ250/LJ252 Companion Color Printers comprise Digital's Color Printer Family. Positioning of the LCG01 Color Printer and the LJ250/LJ252 Companion Color Printers is critical to customer satisfaction; therefore, it is important to understand the differences in their target markets and environments, print quality, operating systems and protocols, operator involvement, and cost of ownership and consumables. The following positioning guide is designed to give you a convenient way of comparing the two printers. As you will see, the LJ250/LJ252 Companion Color Printers are designed to be color workstation printers that sit next to many individual users, have lighter duty cycles, low cost and can serve many general-purpose applications such as business graphics and presentations. On the other hand, the LCG01 Color Printing System offloads the host and is designed to be the shared resource color workstation printer that resides in a central or selected site among a cluster of workstations, has a heavy duty cycle, higher cost and serves specialized applications such as CAD/CAM/CAE. CHARACTERISTICS LJ250/LJ252 LCG01 Placement: Next to users Central or selected site Speed: 2-4 minutes per page 2 minutes per page Volume: 10 to 20 pages per day 20 to 100 pages per day Price: $1,695 $15,900 It is extremely important to understand the factors and definitions that go into determining print quality. For example, certain supplies are critical to achieving the maximum printer output quality of each printer. Print Quality LJ250/LJ252 LCG01 Quality: High quality text Letter quality text Resolution: 180 x 180 dots per inch 154 x 154 dots per inch Ink: LJ25X-AB Black ink LCGX1-JB Black ink cartridge cartridge LCGX1-JM Magenta ink cartridge LJ25X-AA Color ink LCGX1-JC Cyan ink cartridge cartridge LCGX1-JY Yellow ink cartridge LCGX1-JW Maintenance cartridge Paper: LJ25X-AC 8-1/2" x 11" LCGX1-PA Paper (8-1/2" x 11") Continuous form paper (500 sheets/pkg) LJ25X-AD A4 size Continuous form paper Transparencies: LJ25X-AE 8-1/2" x 11" LCGX1-AA Transparencies cut sheet transparencies (8-1/2 x 11") (100 sheets/pkg) LJ25X-AF A4 size cut sheet transparencies When considering the positioning of the Companion Color Printer and the LCG01, you must understand the similarities and differences of the two printers as to the operating systems and protocols under which they are supported, the requirements of the operator and the capabilities of each printer. LJ250/LJ252 LCG01 Operating Systems VMS Yes Yes ULTRIX Yes Yes MS-DOS Yes No Protocols/Applications Color Yes Yes Sixel Yes Yes ReGIS (RETOS) Yes Yes Naplps No Yes Gidis No Yes HP-PCL (IBM) Yes No UIS Yes Yes GKS No Yes Operator Involvement Paper Load: Individual cut sheet Cut sheet (cartridge) or continuous 100 PGS; transparencies 50 Transparencies Ink Load: Operator Operator LJ250/LJ252 LCG01 Cartridge Capacity: Black-500 pages Black and Color Color-150 pages 4000 pages Purging: Operator station Self-purging built-in Maintenance Fluid: Not required Self-dispensing Conclusion: Single user Multiple users operator attended minimal attendance Capabilities Print Speed - Text: 16 seconds per page 2 minutes per page (90 char/s for black @ 1500 characters per page) 2-4 minutes per page 2 minutes per page color printing (20 char/s for color) Print Speed - Graphics: 2-4 minutes per page 2 minutes per page Printheads: 2 Heads 4 Heads Passes: Black - 1 pass Page (single scan) Color - 3 passes Buffer: 2K Character Page (offloads host) After considering the above market characteristics of the Companion Color Printers and the LCG01, it is critical that you show your customer the differences between the cost of ownership and cost of consumables in order to support the justification for the purchase of one or the other printer. COST OF CONSUMABLES Special Paper: .07 per page .11 per page Special Transparencies: 1.39 per page* 1.29 per page Ink Cost: .06 per page (black) .04 per page (black) .12 per page (color) .04 per page (color) Maintenance Fluid: N/A <.01 per page Total Cost Per Page .13 per page (black) .15 per page (black) (Paper): .19 per page (color) .15 per page (color) Total Cost Per Page 1.45 per page (black) 1.33 per page (black) (Transparencies): 1.51 per page (color) 1.33 per page (color) * Includes protective sleeve LJ250/LJ252 LCG01 COST OF OWNERSHIP Purchase Price: $1,695 $15,900 One-Year Warranty: Yes Yes BMC: $11 $125 Installation: By user $270 Acquisition Cost: First Year $l,695 $16,170 Second Year $ 144 $ 1,500 MS is a trademark of Microsoft Corporation ================================================================================ Note 40.6 PERIPHERAL DEVICE ANNOUNCEMENTS 6 of 56 EISNER::KILLEEN "Jeff Killeen" 210 lines 5-SEP-1987 14:14 -< DHQ11 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! THE DHQ11 -- REPLACEMENT PRODUCT FOR THE DHV11 - 22-JUNE-1987 o Provides backward compatibility while improving reliability and performance o Supported by RS232 and DEC423 cabinet kits o Worldwide volume shipment to begin in mid Q4/FY87 The DHQ11-M occupies half the backplane space, uses significantly less power, and provides greater throughput capability than the DHV11-M; therefore, it is very attractive for MicroPDP and MicroVAX users. PRODUCT DESCRIPTION The DHQ11-M Q-bus communications controller provides eight asynchronous Direct Memory Access (DMA) communication lines from a dual size module. It provides RS232-C compatibility to the external equipment. All eight lines provide full modem control, which permits dial-up or leased-line operation with full- or half-duplex modems. The DHQ11 used with the CK-DHQ11-W* cabinet kit will also provide DEC423 compatibility with external equipment. Other features of the DHQ11 include a 256 character FIFO buffer on input and the ability to set all line parameters (e.g., baud rates, character lengths, etc.) via host software. While maintaining software compatibility with the DHV11, the DHQ11 provides the following enhancements: o Dual-height module o Reduced power requirements o Much greater reliability o Improved throughput performance The DHQ11-M Q-bus communications controller is the logical choice when connecting terminal users to a single MicroVAX II or MicroPDP-11 host computer system. However, when local terminal users need access to multiple computer systems, recommend the Local Area Network (LAN)-based terminal server solution. The DHQ11-M communication controller is ideal for local terminal-to-host computer applications within small- to medium-sized businesses such as law, real estate and insurance offices. It is also appropriate for local departmental computing in a small- to medium-sized corporation where the computer handles all the data processing requirements of one department. CONFIGURATION DHQ11-M Cabinet Kits The DHQ11-M is supported by two cabinet kits, an RS232 (full modem control) compatible cabinet kit and a DEC423 (without modem control) cabinet kit. When your customer's application requires terminal users to be located farther than RS232 specification distances (50 feet at 19,200 bit/s) from the computer system, the DEC423 cabinet kit offers a flexible solution. The DEC423 cabinet kit supports both the DHV11-M and DHQ11-M asynchronous communication controllers. DEC423 is based upon the RS423 signalling standard. It allows RS232 VT100/200/300 terminals or compatibles, to communicate at higher speeds from longer distances. This provides greater flexibility in locating terminals where they are needed (up to 1,000 feet from the computer when the H3105-A active adapter is used.) The DEC423 cabinet kit provides the following additional benefits: o A smaller distribution panel than the RS232 version cabinet kit. Takes up only a single panel on the CPU bulkhead. Therefore, additional space is available for more terminal connections. o Reduces the cable "clutter" from the back of the CPU cabinet. Only one 36-conductor cable connects from the bulkhead panel to the compact remote cable connector. o Provides EOS/ESD, (Electrical overstress and Electrostatic Discharge), which significantly reduces the risk of damage to equipment from static discharge, lightning or ac power impulses. Digital supports local terminal-to-host communications of 9600 bit/s speeds at 200 foot distances without the H3105-A active adapter that converts RS232 signals into RS423 signals. Distances beyond 200 feet and up to 1,000 feet at speeds of up to 9600 bit/s are supported with the H3105-A active adapter. FEATURES/BENEFITS o High Performance -- Transmits at up to 38.4 Kbit/s per line (operating system and user application dependent) for a total throughput of 60,000 char/s (10-bit characters; 8 data, 1 stop, no parity). o Direct Memory Access (DMA) on transmit -- Optimizes CPU I/O performance by transmitting data from memory to communication channel without interrupting the CPU on each transmitted character. o Automatic XON/XOFF Flow Control -- Allows the DHQ11-M to manage data flow in either direction on a per-line basis without program intervention. o Low Power Draw of only 1.4 Amps @ +5 V -- Provides more available power for additional peripheral and communication devices on the backplane. o Dual Size Card -- Half the size of the older DHV11-M. Customers can save backplane space by placing two DHQ11-Ms into a single backplane slot. o DEC423 support -- Allows the placement of RS232-compatible terminals where they are needed, up to 1,000 feet from the computer with the H3105-A active adapter. o Full Modem Control with the RS232 Cabinet Kit -- Supports full- and half- duplex, point-to-point modem communications. o High Reliability -- A High Mean Time Between Failure (MTBF) calculation of 60,000 hours results in a long, reliable product life. o Comprehensive Self-Test -- On-board self-test provides extensive testing of DHQ11 circuitry at each power-up or reset. SOFTWARE SUPPORT The DHQ11-M Q-bus asynchronous communications controller is supported by the following operating systems: RSX, MicroRSX, RSTS/E, Micro/RSTS, MicroVMS, MicroPower/Pascal, ULTRIX-32, ULTRIX-11, DSM-11 or VAXELN. SYSTEM SUPPORT The DHQ11 is designed to replace the DHV11 for all Q-bus systems and will be supported by the MicroVAX II, MicroPDP-11/23, MicroPDP-11/53, MicroPDP-11/73 and MicroPDP-11/83. SPECIFICATIONS Asynchronous Formats Half- or full-duplex supported format -> Single start bit Formats: -> 5-, 6-, 7- or 8-bit character size -> Odd, even or no parity -> 1, 1.5 or 2 stop bits Electrical Interface Standards The DHQ11-M will directly support: -> EIA RS232-C/CCITT V.28 w/the CK-DHQ11-A* Cab. Kit -> DEC423 w/the CK-DHQ11-W* Cab. Kit Diagnostics: The DHQ11 uses all existing DHV11 diagnostics. Operating Environment Power Requirements: -> +5 Vdc +/-5% at 1.4 A (typical) -> +12 Vdc +/-5% at 230 ma -> dc loads - 1.0 dc loads Storage Temperature: -> Short-term storage (up to 95% humidity) (2 months or less) -40oC to 66oC (-40oF to 151oF) -> Long-term storage (longer than 2 months) 5oC to 50oC (41oF to 122oF) Operating Temperature: -> 5oC to 50oC (41oF to 122oF) Relative Humidity: -> 10% to 95% noncondensing with a maximum wet bulb of 32oC (90oF) and a minimum dew point of 2oC (36oF) The DHQ11 complies with DEC STD 102 Class C, and conforms to all EMI/RFI standards as referenced in Section 7, Rev. C (FCC & VDE). PRICING/ORDERING INFORMATION Order No. Description MLP BMC DECservice Inst. DHQ11-M Base Module M3107 $1,520 $15 $18 $330 Cabinet Kits For RS232 Connection: This cabinet kit supports full modem control and contains two B size slot bulkhead distribution panels with 25-pin male connectors and associated cabling to connect the control module to the distribution panels. Data cables are not included. CK-DHQ11-AA 21" cable for BA123 Box $150 CK-DHQ11-AB 12" cable for BA23 Box $150 CK-DHQ11-AF 36" cable for H9642 Cab $150 For DEC423 Connection: This cabinet kit does not support modem control and consists of 1 B size slot bulkhead distribution panel with a 25-foot, 36-conductor cable, compact remote terminal concentrator and associated cabling to connect the control module to the distribution panel. Data cables are not included. CK-DHQ11-WA 21" cable for BA123 Box $400 CK-DHQ11-WB 12" cable for BA23 Box $400 CK-DHQ11-WF 36" cable for H9642 Cab $400 ================================================================================ Note 40.7 PERIPHERAL DEVICE ANNOUNCEMENTS 7 of 56 EISNER::KILLEEN "Jeff Killeen" 26 lines 5-SEP-1987 14:15 -< RX50 TO RX33 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! RX50 TO RX33 MIGRATION - 13-July-1987 o MicroPDP-11 customers encouraged to migrate from the end-of- life RX50 to newer RX33 The RX50 800 KB, dual-floppy diskette subsystem is now a mature product and is approaching its end-of-life. With the announcement of support for the RX33 1.2 MB, double-sided flexible diskette drive, MicroPDP-11 customers should now be encouraged to migrate. Because the RX50 is in product phase-down mode, customers who need a floppy device should purchase the RX33 when that option is supported. The dual-mode capability of the RX33 not only provides full backward read/write compatibility to existing RX50 media, but allows users to upgrade to the industry-standard 1.2 MB format, using double-sided, 96 tpi media. No attempt will be made to migrate systems nearing product phase down. These systems will continue to offer only the RX50 as the floppy device until their retirement. ================================================================================ Note 40.8 PERIPHERAL DEVICE ANNOUNCEMENTS 8 of 56 EISNER::KILLEEN "Jeff Killeen" 54 lines 5-SEP-1987 14:17 -< RL02 END OF LIFE >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! RL02 END-OF-LIFE - 13-July-1987 o RL02 product phase down underway o Last order accepted September 14, 1987; last shipment on December 24, 1987 o Cartridges will continue to be available o Software to still be distributed on RL02 media The RL02 has been a successful product for Digital with an installed base in excess of 100,000 units worldwide. However, the RL02 family is rapidly reaching end-of-life and we must communicate this to our customers. Shipments have declined significantly during each of the past two years, resulting in increased costs. Manufacturing capacity will be better utilized for newer, higher-volume products. ORDERING/SHIPPING INFORMATION To ensure that sufficient units are built to support ongoing customer requirements and long-term contractual needs, please submit orders by September 14, 1987. Final shipments will take place by December 24, 1987. Any exceptions will be handled on an individual basis by your Operations contact. Quick and accurate feedback to your Operations contact will enable our manufacturing plants to better size production requirements and meet the needs of our customers. AVAILABILITY Cartridges will still be available after end-of-manufacturing life and may be ordered through DECdirect. Software will continue to be distributed on RL02 cartridges. The RL02 will also continue to be offered embedded on the VAX 8600/8650 after the drive is no longer available as an independent option. SUPPORT INFORMATION Digital Field Service will continue to support the RL02 with its complete range of service products even after the end of the product's manufacturing life. The availability of spare parts and trained engineers is ensured as a part of the standard end-of-life planning. ================================================================================ Note 40.9 PERIPHERAL DEVICE ANNOUNCEMENTS 9 of 56 EISNER::KILLEEN "Jeff Killeen" 83 lines 5-SEP-1987 14:17 -< RD52 END OF LIFE >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! RD52 END-OF-LIFE - 10-July-1987 o RD52 is being phased out; RD32 is the primary replacement product o Place your RD52 orders by September 14, 1987; last shipment on December 24, 1987 The RD32 continues Digital's trend toward faster, higher-capacity storage in the half-height form factor. Compared to the RD52 drive, the RD32 provides 35% more capacity, 16% more speed and uses 60% less power. The RD32 utilizes three thin-film rigid media platters in a convenient half-height 5.25" package. PRODUCT POSITIONING RD31 RD32 RD52 RD53 Capacity (MB) 20 42 31 71 Access Time (Ms) 73 48 57 38 Transfer Rate (KB/s) <----------- 625 -----------> Form Factor <---------- 5.25" ----------> Controller RQDX2/3 RQDX3 RQDX2/3 RQDX2/3 Height <-- HALF --> <-- FULL --> Price $1,195 $2,300 $3,300 $3,800 Price/MB $50 $48 $97 $57 ORDERING/SHIPPING INFORMATION To ensure sufficient units are built to support ongoing customer require- ments and long-term contractual needs, please submit orders by September 14, 1987. Final shipments will take place by December 24, 1987. Exceptions will be handled on an individual basis by your Operations contact. Quick and accurate feedback to your Operations contact will enable our manufacturing plants to better size production requirements and meet the needs of our customers. SUPPORT INFORMATION Digital Field Service will continue to support the RD52 with its complete range of service products even after the end of the product's manufacturing life. The availability of spare parts and trained engineers is ensured as a part of the standard end-of-life planning. SYSTEM SUPPORT The following chart describes which newer drives are currently supported on certain systems. Refer to "Announcing Major Enhancements to the PDP-11 Family of Systems" in this issue for anouncement of additional drive support. MicroPDP-11/53 RD31, RD32, RD53, RD54 MicroPDP-11/73 RD32, RD53, RD54 MicroPDP-11/83 RD32, RD53, RD54 VAXmate RD31 PRO RD31 Rainbow RD31 MicroVAX II RD53, RD54 MicroVAX 2000/VAXstation 2000 RD32, RD53 ================================================================================ Note 40.10 PERIPHERAL DEVICE ANNOUNCEMENTS 10 of 56 EISNER::KILLEEN "Jeff Killeen" 181 lines 8-SEP-1987 19:42 -< DELQA >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! THE DELQA-M -- REPLACEMENT PRODUCT FOR THE DEQNA-M - 08-September-1987 o The DELQA-M is the Ethernet/IEEE 802.3 compatible replacement for the DEQNA-M o Compatible with latest versions of DEQNA-M software drivers o The DELQA will be available in three options, depending on system The DELQA Ethernet Controller is now available in options to support Q-bus MicroVAX II and MicroPDP-11 systems, or the new BA200 Series system enclosures that include the Industrial VAX and Industrial PDP-11. PRODUCT DESCRIPTION The DELQA Ethernet controller is available in various options, depending upon the system enclosure. DELQA-M Option -- The Ethernet Controller for Q-bus MicroVAX II and MicroPDP-11 systems. DELQA-SA and DELQA-SF Options -- The Ethernet Controllers for the new BA200 Series systems enclosures that presently consist of the Industrial Vax and Industrial PDP-11. The -SA version must be ordered with the system as part of the System Building Block configuration. The -SF version must be ordered as an independant add-on to previously shipped and/or installed systems. All DELQA options are compatible with the latest versions of the DEQNA-M software drivers, while providing the following added features: o Comparable throughput performance to the DEQNA o Hardware-based Maintanence Operation Protocol (MOP) o Supports both Ethernet V2.0 and IEEE 802.3 networks o Low power requirements o Greater reliability DELQA-M Option The DELQA-M is a dual-size board that provides connection for MicroVAX II and MicroPDP-11 Q-bus systems to both Ethernet V2.0 and IEEE 802.3 local area networks. It performs both physical and data link functions for the Ethernet node. CSMA/CD channel access functions are implemented in the hardware, while on-board firmware deals with data encapsulation/ decapsulation and data link management. The DELQA-M provides the hardware capability to support Maintenance Operation Protocol (MOP) which offers enhanced network management features including: remote circuit loopback, system identification messages, remote booting of diskless systems, maintaining data link counters and the IEEE 802.2 XID and test. The DELQA-M physically and electrically connects to the Ethernet coaxial cable by means of a CK-DELQA-** cabinet kit, transceiver cable (BNE3C or BNE3A series) and an H4000 transceiver or DELNI. The DELQA-M is also used in ThinWire environments by connecting to the DESTA station adapater. DELQA-SA/SF Options The DELQA-SA and DELQA-SF options are quad-size boards that presently support the DEC/CIM family of Q-bus systems which utilize the BA200 Series enclosures, including the Industrial VAX and Industrial PDP-11. The -SA and -SF options will also support other systems being planned for announcement in the near future. The DELQA-SA/SF options are mechanically different from conventional Q-bus controllers in that they do not require a cabinet kit. The bulkhead distribution panel is integral to the module handle. FEATURES/BENEFITS o Dual-Size Module -- Saves on scarce backplane space, providing more available backplane space for additional peripheral devices. o Low Power Consumption -- Only 2.5 amps are required from the internal CPU +5V power supply. Therefore, there is more available power for connecting additional peripheral and communication options. o Comprehensive Self-Test -- Executes extensive testing of DELQA circuits at each power-up or reset. o On-Board Maintanance Operation Protocol (MOP) -- MOP implemented in firmware, off-loads system software demand, thus reducing processor time required to support network management activities and ensures that expensive CPU cycles are not wasted. o Supports Ethernet/IEEE 802.3 Networks -- Provides the user with the added flexibility to use Ethernet V2.0 or IEEE 802.3 protocols. SOFTWARE SUPPORT The DELQA is designed to operate with the current DEQNA drivers that are included in the latest versions of the following operating systems or software packages: DECnet-VAX, LAVc, VAXELN, DECnet-ULTRIX, DECnet-RSX family, DECnet-Micro/RSX, DECnet/E, DSM-11, MicroVMS, ULTRIX-32, Micro/RSX, RSX-11M, RSX-11S, and RSX-11M-PLUS. SYSTEM SUPPORT The DELQA-M is designed to replace the DEQNA-M for all Q-bus systems and will be supported by the MicroVAX II, VAXstation II, VAXstation II/GPX, MicroPDP-11/53, MicroPDP-11/73 and MicroPDP-11/83. Planned dates are: SPECIFICATIONS Mounting Code: DELQA-M 1 dual slot DELQA-SA/SF 1 quad slot Power Requirements: DC amps drawn at +5 V: 2.5 DC amps drawn at +12 V:0.5 Bus Loads: AC 2.2, DC 0.5 I/O Connection Panel Insert: DELQA-M 1 Size A DELQA-SA/SF Panel is integral to module handle Operating Environment: The DELQA options comply with DEC STD 102 Class C and conform to all EMI/RFI standards as referenced in Section 7, Rev. C (FCC & VDE). Storage Temperature 0oC to 66oC (32oF to 151oF) Operating Temperature 5oC to 60oC (41oF to 140oF) Relative Humidity 10% to 95% non-condensing (complies with DEC STD 102 Class C) PRICING/ORDERING INFORMATION Order No. Description U.S. List Price BMC Inst. For Q-bus MicroVAX II and MicroPDP-11 systems, order the following: DELQA-M Ethernet/IEEE 802.3 $2,500 $15 $300 Q-bus Controller DELQA-M Cabinet Kits: CK-DELQA-YA Cab. Kit for BA123 and 150 N/A N/A BA11-M Box CK-DELQA-YB Cab. Kit for BA23 150 N/A N/A Micro-11 Box CK-DELQA-YF Cab. Kit for H9642 Cab. 150 N/A N/A For BA200 Series system enclosures, order the following: NOTE: Cabinet kits are not required. For System Building Block Configurations: DELQA-SA Ethernet/IEEE 802.3 2,500 15 300 Controller for BA200 series enclsoure For add-ons to previously shipped and/or installed systems: DELQA-SF Ethernet/IEEE 802.3 2,500 15 300 Controller for BA200 series enclosures ================================================================================ Note 40.11 PERIPHERAL DEVICE ANNOUNCEMENTS 11 of 56 EISNER::KILLEEN "Jeff Killeen" 282 lines 11-SEP-1987 00:26 -< RA70, RA82, AND TK70 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! NEW STORAGE FOR THE NEW SYSTEMS - 10-September-1987 MicroVAX, VAXserver, AND VAXstation SYSTEMS: RA70, RA82, AND TK70 o High-reliability, high-performance, high-capacity storage for the new members of the MicroVAX family o RA70 -- 280 MByte 5.25-inch Winchester disk drive o RA82 -- 622 MByte 14-inch Winchester disk drive o TK70 -- Streaming cartridge tape subsystem, ideal backup for RA70 and other Winchester disk drives Storage Systems announces the RA70, RA82, and TK70. Each complement the new MicroVAX, VAXstation and VAXserver family members in their own way (refer to individual sections below) as well as in ways common to all Digital storage products. All three products are designed to fit into the Digital Storage Architecture (DSA) and offer all the features and benefits of the DSA: o INVESTMENT PROTECTION for existing and future applications and equipment o NO SOFTWARE CHANGES NEEDED when adding new DSA mass storage o DATA AVAILABILITY and DATA INTEGRITY features maximize the users' access to data and help avoid costly unrecoverable errors All three storage products provide enhanced storage-related performance to MicroVAX 3xxx-based systems. Each product comes with a one-year warranty which includes DECservice, Digital's premier on-site service offering. NOTE: Refer to the system level articles for selling these products with the systems. RA70 PRODUCT DESCRIPTION o High-capacity, high-performance 280 MByte 5.25-inch disk drive o Initial availability on MicroVAX 3xxx-based systems ONLY The Digital-designed and manufactured RA70 disk drive is well suited for use in office and laboratory environments. The RA70 launches a new generation of Digital high technology disk products, This 280 MByte formatted (350 MByte unformatted*) disk drive is designed for I/O intensive applications and features full DSA/SDI functionality, dual port, high reliability, high performance, and high capacity in a 5 1/4-inch form factor. The RA70 HDA includes a linear actuator with eleven minicomposite data heads and one thin film servo head. Servo information is provided via a dedicated servo head and is also embedded in each data sector. The HDA also includes an air filtration and circulation system. Like all 5.25-inch disk drives, the RA70 does not have its own power, cooling, or packaging. Power and cooling for the RA70 must be supplied by an external source. At this time, RA70 power and packaging support is available on the MicroVAX 3500/3600 system enclosures only. RA82 PRODUCT DESCRIPTION o High capacity and high transfer rate for large contiguous files o Availability on MicroVAX 3600 and VAXserver 3600/3602 computers ONLY The RA82 14-inch, 622 MByte (855 MByte unformatted*) disk drive -- the highest capacity RA-disk drive -- also features full DSA/SDI functionality, dual port, high reliability, high performance. *Unformatted capacity provided for comparison purposes; only formatted capacity is accessible to the user in any disk device. RA70 FEATURES RA70 BENEFITS Two field replaceable units If the RA70 experiences a failure with the module set, it can be quickly and easily replaced without removing the HDA. Avoids loss of 280 MBytes of data and the trouble of going through a restore process Faster throughput Faster response time for I/O intensive applications Thin film media Areal density of 30.4 million bits per square inch, Only 5.25-inch disk drive fully DSA Higher availability than previously compatible, including SDI, dual available porting RA70 and RA82 FEATURES RA70 and RA82 BENEFITS Dedicated and embedded servo system Provides increased track density High reliability High data availability; low service costs 170-bit error correction code (ECC) Prevents data loss Bad block replacement (BBR) Suspect blocks retired before media defects cause data loss Radial connection Increased data availability, and ability to service individual drives while system is running INTERNAL POSITIONING ON MicroVAX 3600 AND VAXserver 3600 On the MicroVAX 3600 and VAXserver 3600, the RA70 and RA82 provide a range of storage configurations suited to a range of I/O intensive applications. Choose an appropriate combination of RA70s for faster access times, and RA82s for higher capacity and higher transfer rates of large contiguous files. For example, in applications such as CAD/CAM, the need for quick access to symbols and keywords can be met by RA70s, and the need for high capacity and high transfer rates for large, contiguous files can be met by RA82s. SPECIFICATIONS RA70 RA82 Media Characteristics Formatted Capacity (MBytes) 280 622 Unformatted Capacity (MBytes) 350 855 Areal Density (Mbits/Sq. In.) 30.4 13.3 Number of Actuators 1 1 Number of Data Heads 11 15 Performance** RA70 RA82 Average seek time (ms) 19.5 24 Average access time (ms) 27.0 32.3 Peak transfer rate (MBytes/sec) 1.4 2.4 ** Does not include improvements resulting from DSA/SDI controller optimizations. TK70 STREAMING CARTRIDGE TAPE SUBSYSTEM PRODUCT DESCRIPTION o TK70 offers 3X capacity (296 MB) and 2X speed (125/90 KBytes per second) of TK50 o Newly designed, easy to understand front panel o TK70 can read TK50 cartridges o Available with MicroVAX 3500/3600, VAXserver 3500/3600/3602, and VAXstation 3500 Support of the TK70 on other systems will be announced at a later date. Designed to be used on high-end Q-bus systems, the TK70 will satisfy your customers' needs for disk backup, software distribution, data collection, and data interchange between microsystems. TK70 is included with MicroVAX 3500/3600, VAXserver 3500/3600/3602 and VAXstation 3500. The subsystem consists of the TK50-sized (5.25-inch form factor) tape drive, dual-height Q-bus controller and cables. Power supply and cabinetry are supplied by the system box in which the TK70 is mounted. Unlike the TK50, TK70 is not currently available in desktop or rack-mount packages. The new subsystem uses an enhanced version of the CompacTape cartridge. The CompacTape II cartridge (model number TK52-K) appears virtually identical to the TK50-K and contains the same magnetic media, but is manufactured to tighter tolerances to accommodate the high-density 296 MB format of the TK70. TK52-K cartridges should be used only with TK70 drives; TK50 users should continue to order TK50-K cartridges. Even though a brand new TK50-K cartridge might work on a TK70 drive in TK70 mode, this is not a supported configuration and performance cannot be guaranteed. The new TK70 can read (but not write) TK50 formatted data, so libraries of older cartridges can be read on the new subsystem, and OEMs who distribute software on TK50-K cartridges can continue to do so, whether their customers have TK50 or TK70 drives. Digital will continue to distribute software on TK50-K cartridges in TK50 95 MByte format. FEATURES Five unique characteristics make the TK70 a leadership product: o 296 MB CAPACITY* allows the TK70 to back up the RA70 disk or any of Digital's RD family of disks onto a single cartridge -- with no operator intervention. Only two cartridges are required to back up a 95% full RA82 Disk. Compared to other products in high-volume production, this is the highest capacity cartridge in the industry. o FASTER PERFORMANCE helps users complete backup operations and other tape-related applications in approximately half the time they would need with a TK50. Raw data transfer rate is 125 KB/sec. Formatted transfer rate is 90 KB/sec. This translates to an RA70 backup time (on MicroVAX 3500/3600, VAXserver 3500/3600/3602) of less than one hour. o SOLID RELIABILITY results from our use of rugged parts in the TK70 and from a highly intelligent controller that can automatically recover from most fault conditions. o A REDESIGNED FRONT PANEL, using multiple clearly-labeled lights and an audible signal, clearly tells the user what the drive is doing and what operations the user can safely perform. This major advance, over competitive products and over the TK50, helps ensure that even novice users will operate the TK70 correctly. o DATA INTEGRITY is outstanding on TK70, as on TK50, because the subsystem uses a uniquely powerful error correcting code (ECC), a cyclic redundancy check (CRC) for error detection, and a read-after-write procedure to verify that data has been written correctly. In addition, calibration bursts and automatic gain control guarantees media interchangeability between subsystems. No field adjustments, no preventive maintenance, and no head cleaning in normal use provide ease of use and low cost of ownership. *Actual capacity may be slightly less in certain situations. See Performance section in this article. PERFORMANCE With a peak transfer rate for user data of 90 KB/second, the TK70 streaming tape subsystem can back up an RA70 on a MicroVAX 3500 or 3600 in less than one hour. Actual performance on some systems may be slower and will depend on several variables, such as disk speed, load on the system and the operating system and tape utility being used. Test results, actual performance data and more complete data on effective capacity will be published soon. In the interim, customers can be told about several features built into the TK70 to enhance its performance: - 64 KB buffer in the controller - the ability to pre-fetch commands and data from the CPU - the ability to do error correction without stopping to reposition the tape All of these features are discussed in the TK70/TK50 Family Information Sheet. To enhance streaming performance with certain host configurations, the TK70 can also write redundant blocks (a string of zeroes), thereby allowing the drive to continue streaming and reducing the need to reposition. Although this reduces cartridge capacity somewhat, the effect is small and it still leaves sufficient capacity to perform single-cartridge backups of Digital's mini-Winchester disks. SPECIFICATIONS Function TK50 TK70 Capacity (Formatted) 95 MB 296 MB Form Factor 5.25" 5.25" Transfer Rate (RAW) 63 KB/s 125 KB/s Transfer Rate (DATA) 45 KB/s 90 KB/s Cartridge Type TK50-K TK52-K TK50 Compatibility READ/WRITE READ Tracks 22 48 Tape Length 600' 600' Tape Speed 75 ips 100 ips Recording Density 6.7K bpi 10K bpi Error Rate 10*E-11 10*E-11 Power Requirements: Drive +12V 2.4A 2.4A Drive +5V 1.4A 1.4A Controller +5V 3.0A 3.5A Power Consumption (Drive/Contr.) 36W/15W 36W/17.5W AC Bus Loads 4.1 4.1 ================================================================================ Note 40.12 PERIPHERAL DEVICE ANNOUNCEMENTS 12 of 56 EISNER::KILLEEN "Jeff Killeen" 7 lines 11-SEP-1987 00:28 -< YEA!!! >- -------------------------------------------------------------------------------- > o A REDESIGNED FRONT PANEL, using multiple clearly-labeled lights and an > audible signal, clearly tells the user what the drive is doing and what > operations the user can safely perform. This major advance, over > competitive products and over the TK50, helps ensure that even novice > users will operate the TK70 correctly. YEA!!!!!!!!!!!!!!!!!!!!!! ================================================================================ Note 40.13 PERIPHERAL DEVICE ANNOUNCEMENTS 13 of 56 EISNER::KILLEEN "Jeff Killeen" 9 lines 11-SEP-1987 22:36 -< TK70 / RA70 PRICES >- -------------------------------------------------------------------------------- All the following prices are for components (no power supplies). The RA70 is the same form factor as the RD54 and mounts the same way. TQK70 $1,700 TK70 $5,600 KDA50-QA $6,825 RA70 $9,000 ================================================================================ Note 40.14 PERIPHERAL DEVICE ANNOUNCEMENTS 14 of 56 EISNER::KILLEEN "Jeff Killeen" 329 lines 10-OCT-1987 13:43 -< H7317 AND H7318 POWER SYSTEMS >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! ANNOUNCING THE H7317 POWER DISTRIBUTION SYSTEM PLUS AND THE H7318 POWER CONDITIONING SYSTEM PLUS - 2-October-1987 o The next generation of environmental power products for medium and large data processing systems The H7317 Power Distribution System PLUS and H7318 Power Conditioning System PLUS are the newest member of Digital's family of enviromental monitoring and power solutions products. The PDS PLUS and PCS PLUS are flexible power distribution, voltage regulation and transient surge suppression systems... P L U S a microprocessor-based data acquisition system, which enables total electronic environmental surveillance of the entire computer room. PRODUCT DESCRIPTIONS The H7317 PDS PLUS and H7318 PCS PLUS systems are enhanced units of the H7227 PDS and H7228 PCS systems. The power conditioning and distribution function remain the same. But the present monitoring system has been removed and replaced with a state-of-the-art microprocessor-based environmental monitoring system. The new monitoring system will do more than just power monitoring, conditioning and distribution. The total electronic environmental surveillance of any computer site and its supporting equipment can be accomplished by this device. The step beyond monitoring is achieved by real-time sampling of internal and externally attached sensors. Any physical environmental changes are sensed and then compared to user pre- established thresholds. When an abnormality is sensed, the user is alerted. This system has the capability of handling 112 data sensor/probes. Each data probe is individually programmed with up to four threshold levels. Each threshold level is programmed separately for delays and alarm action. Each probe has an identifying name that is easily read on a user friendly display. The system has an optional RS232 communications port for interfacing to a terminal or host system. FEATURES/BENEFITS o Provides customers with a central reporting station for various alarm systems o Provides customer computer room with additional environmental monitoring o Allows customer to preprogram a system's reaction to various alarm systems o Single connection point to building power o Flexible power distribution with electrical noise protection o Single point grounding o State-of-the-art technology o Reduces cost of ownership through higher system availability o High reliability and quality o Available for 60 Hz and domestic receptacles o UL and CSA listed SPECIFICATIONS The following technical specifications apply to both PCS and PDS PLUS. Output: 120/208 Volts AC WYE connected 3 Phase Input Main Circuit Breaker: Three pole breaker rated for unit by power capacity and input voltage. It can be operated manually or by internal or external shunt trip mechanism. Isolation Transformer: Designed specifically for digital computers. Electrically shielded delta to WYE copper construction. Conforms to MIL-T-27. Efficiency: 96% Minimum Display: LCD Monitoring: Microprocessor controlled with 43 Internal points and up to 69 External points Alarm Thresholds: Up to four threshold levels for analog probes and two thresholds for digital probes. Printer: 40 character, built in to the unit Dimensions: Height 42", Width 30.5", Depth 36" CONFIGURATION SELECTION GUIDE A System Building Block approach can be used in the selection of the correct options needed for your particular installation. Select H7318 PCS PLUS or H7317 PDS PLUS unit. (J-Box is supplied with unit.) Suffix variant will depend on input voltage and KVA size required (see latest Environmental Product catalog for selection and sizing information). REPO and ENVIRONMENTAL MONITORING STATIONS (EMS) Remote Emergency Power Off button is required by the National Electrical Code. One should be ordered for each exit within the computer room: H7317-JA REPO button, wall mounted with 50 foot connecting cable. REPO button is illuminated to indicate that it is connected and active. Select one of following for each exit to have a wall-mounted REPO unit with Temperature and Humidity monitoring (EMS). Temperature and Humidity readings are displayed on an LCD display. Unit can be preprogrammed to two over temperature levels. First levels provides audible alarm with indicator and second stage will automatically power the system down. Temperature and Humidity readings are not reported to PCS/PDS PLUS unit for logging. H7317-JB REPO w/Temperature monitoring H7317-JC REPO w/Temperature and Humidity monitoring Select one of the following to have the Temperature and Humidity readings on the EMS LCD reported to the PCS/PDS PLUS for logging. H7317-JD REPO w/Temperature monitoring H7317-JE REPO w/Temperature and Humidity monitoring (50-Foot connecting cables are included.) H7317-KB BUILDING INTERFACE ALARM UNIT (BIA) This unit is used to interface up to eight dry switch contacts. A dry switch contact is a switch that is provided by most alarm systems and allows an easy way of tying into their system. The sensing of the dry switch contact can be fire alarm system, Halon systems, Smoke Detectors, door alarms or any alarm device that has a dry contact. The BIA unit also has three Relays for outgoing alarm signals which can be used to control air conditioners or used for remote alarm signalling such as light, bell or autodialer. H7317-KE REMOTE SENSING UNIT (RSU) The RSU allows the interfacing of various types of sensors and probes to be connected to the PCS PLUS/PDS PLUS system to provide extensive environemtal monitoring in real-time mode. The RSU has a built-in microprocessor which gathers input information from up to twenty electrical interface adapters and tramsmits it to the PCS PLUS/PDS PLUS unit via a data communications cable. Up to six H7317-KE units can be attached to one PCS PLUS/PDS PLUS unit for the interfacing of 69 external probes or sensors. A unit requires a data communications cable. Select one for each RSU from the listing below. o First RSU -- Select one only BC31B-50 (50ft), BC31B-A0 (100ft), BC31B-B0 (200ft). o Second RSU -- Select a BC31B cable of desired length and 1 H7317-KY Auxillary Power Adapter. o Third through sixth RSU -- All require a H7317-KY Auxillary Power Adapter. Cables for third through the sixth RSU are BC31F-50 (50ft), BC31F-A0 (100ft), BC31f-B0 (200ft). The RSU can be located up to 5000 ft from the PDS PLUS/PCS PLUS unit. Connector Kits for data communication cables are available. H3280-AB for PCS PLUS/PDS PLUS unit to RSU and H3280-AD for RSU to RSU. PROBES, SENSOR AND ELECTRICAL INTERFACE ADAPTERS FOR THE H7317-KE RSU o Wall mounted temperature probe (H7317-KC) -- monitors temperature and displays actual readings on PCS PLUS/PDS PLUS LCD display. Four user threshold levels can be programmed. (High, high-high, low and low-low). Each level can provide different alarm reactions as desired by the customer. Multiple temperature sensor can be attached to the RSU for expanded computer room monitoring of hot spots. o Water Detector (H7317-KF) -- is used to detect water under computer room floors. They are ideal for placement near water cooled air conditioner systems. They can be connected in parallel for mulitple sensing on one alarm circuit. ELECTRICAL INTERFACE ADAPTERS The EIA's electrical interface adapters are interfaces which convert external sensor voltage or signal levels to proper data levels required by the RSU. There are primarily two types: one is the digital input which has only two states and can be programmed for two alarm actions -- alarm active or alarm normal. The second is the analog type. This means that input sensor levels can be converted to usable measurements such as temperature levels, voltage, etc. The analog inputs can be programmed for four threshold levels (high, high-high, low, low-low). o H7317-KJ EIA Dry Contact Sensing -- this interface adapter is used to sense switch contact closure or opening of any alarm system (digital type). o H7317-KL EIA 12 Volt AC/DC Interface -- allows sensing of 12 volts AC or DC present or not present (digital type). o H7317-KM EIA 24 Volt AC/DC Interface -- allows sensing of 24 volt AC/DC present or not present (Digital type). o H7317-KN EIA Frequency Sensing Interface -- allows the measurement of frequency from 0-400 Hz. Requires an AC signal level of no greater than 12 volts (analog type). o H7317-KP EIA 5 Volt Interface -- this adapter will detect the presence of 5 volts DC (digital type). o H7317-KR 4 to 20 milliamp Interface -- allows the interfacing of many off-the-shelf industrial probes that utilize the 4 to 20 milliamp standard interface (analog type). o H7317-KS Analog Buffer 0 to 12 Volts AC input -- this EIA allows the measurement of AC voltages (analog type). Requires voltage step down from primary voltage to max 12 volts AC. OUTPUT RELAYS There is one relay position in each RSU which can be used as additional outgoing alarm circuits. There are two types of relays: H7317-KV, a four pole relay and H7317-KW, a mechanical latching type. When operated, it will stay operated until reset electrically. POWER ADAPTERS H7317-KZ 12 Volt Auxiliary Power Adapter provides power for the additional RSUs or for RSUs that are greater than 200 feet from the PDS PLUS/PCS PLUS unit. They also provide battery backup for the RSUs during power failures. SERVICE OFFERING Purchase price includes installation and one-year warranty. DSMC and BMC will be published in the USPL. PRICING/ORDERING INFORMATION H7318 POWER CONDITIONING SYSTEM PLUS (PCS PLUS) INPUT VOLTAGE U.S. KVA Poles 208 Volts 480 Volts List Price 15 42 H7318-AA H7318-HA $22,995 30 42 H7318-AB H7318-HB 23,995 50 42 H7318-AC H7318-HC 24,995 50 84 H7318-AJ H7318-HJ 25,995 75 84 N/A H7318-HD 26,995 100 84 N/A H7318-HE 27,995 H7317 POWER DISTRIBUTIUON SYSTEM PLUS (PDS PLUS) INPUT VOLTAGE U.S. KVA Poles 208 Volts 480 Volts 600 Volts List Price 15 42 H7317-AA H7317-HA H7317-CA $12,995 30 42 H7317-AB H7317-HB H7317-CB 14,995 50 42 H7317-AC H7317-HC H7317-CC 16,995 50 84 H7317-AJ H7317-HJ H7317-CJ 17,995 75 84 N/A H7317-HD H7317-CD 18,995 100 84 N/A H7317-HE H7317-CE 19,995 N/A = NOT AVAILABLE Other input voltages of 220, 240, 440, AND 460 available upon request PDS PLUS/PCS PLUS REPO ENVIRONMENTAL MONITORING STATIONS (EMS) Order No. Description U.S. List Price H7317-JA Remote Emergency Power off Station (REPO) $250 H7317-JB REPO with/Temperature monitoring 770 H7317-JC REPO with/Temp and Humidity monitoring 975 H7317-JD REPO with Temperature monitoring 770 H7317-JE REPO with Temp and Humidity monitoring 975 H7317-JF REPO with Temperature monitoring (RSU) 835 H7317-JG REPO with Temp and Humidity monitoring (RSU) 995 PDS/PCS PLUS OPTIONS Order No. Description U.S. List Price H7317-KB BIA (Building Interface Alarm) unit $495 H7317-KE RSU (Remote Sensing Unit) 935 H7317-KE OPTIONS Order No. Description U.S. List Price H7317-KC Temperature probe wall mounted w/EIA card $195 H7317-KF Water detector, under floor probe w/EIA card 495 H7317-KJ EIA 1 Dry contact closure sensor 65 H7317-KK EIA 2 1 milliamp current sensor 65 H7317-KL EIA 3 12 Volt AC/DC interface 65 H7317-KM EIA 4 24 Volt AC/DC interface 65 H7317-KN EIA 12 Frequency electrical interface 175 H7317-KP EIA 15 5 Volt electrical interface 65 H7317-KR EIA 17 4-20 milliamp electrical interface 165 H7317-KS EIA 8 Analog buffer 0-12 Volts AC 190 H7317-KT EIA 9 Isolated buffer 0-5 Amps AC 190 H7317-KV Four Pole output relay module 65 H7317-KW Four Pole output relay mechanical latching 250 H7317-KY 12 Volt auxiliary power adapter for H7317-KE RSU 450 H7317-KZ AC Power unit 120 Volt input 16 Volts output 44 Cable Options Order No. Description U.S. List Price BC31A-50 RS422 Unit to Unit communications cable 50 ft $ 65 BC31A-A0 RS422 Unit to Unit communications cable 100 ft 105 BC31A-B0 RS422 Unit to Unit communications cable 200 ft 170 BC31B-50 Unit to RSU communications cable 50 ft 65 BC31B-A0 Unit to RSU communications cable 100 ft 115 BC31B-B0 Unit to RSU communications cable 200 ft 190 BC31C-50 Unit to Unit REPO cable 50 ft 75 BC31C-A0 Unit to Unit REPO cable 100 ft 85 BC31C-B0 Unit to Unit REPO cable 200 ft 165 BC31D-50 Cable unit to RS232 connector 55 BC31F-50 RSU to RSU communications cable 50 ft 85 BC31F-A0 RSU to RSU communications cable 100 ft 140 BC31F-B0 RSU to RSU communications cable 200 ft 220 Cable Connector Options Order No. Description U.S. List Price H3280-AA Connector for RS422 Unit To Unit Communications $ 55 H3280-AB Connectors for Unit to RSU communications 75 H3280-AC Connectors for Unit to Unit REPO in/out 65 H3280-AD Connectors for RSU to RSU communications 85 ================================================================================ Note 40.15 PERIPHERAL DEVICE ANNOUNCEMENTS 15 of 56 EISNER::KILLEEN "Jeff Killeen" 180 lines 10-OCT-1987 13:45 -< DFA01 MODEMS >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s via the RONNIE network !___!___!___!___!___!___!___! ANNOUNCING THE DFA01 AND DFA01/VAXPAC SYSTEM - 2-October-1987 o DFA01 -- Internal modem for use in MicroVAX 3500/3600 systems and PDP-11 microcomputers o VAXPAC -- Allows VMS users to connect to a remote system using asynchronous lines o DFA01 hardware and VAXPAC software -- a state-of-the-art implementation of 2400 bits/s technology communications package for the VMS-based MicroVAX 3500/3600 system user PRODUCT DESCRIPTION DFA01 The DFA01 is two 300/1200/2400 direct connect modems with an integral DZQ11 interface on a single quad module. Telephone modular jacks are provided on the DFA01 for direct connection to the telephone network. Phone cables are also supplied. The DFA01 is to be sold in the U.S. and Canada only. DFA01 Hardware Product Features: o High performance dual modems plus interface on a quad card o Operates at 2400/1200/300 bits/s (V.22 bis, 103J, 212A) o Built-in autodialer compatible o Q-bus DZQ11 interface o MJ Telephone Interface (RJ11C) o Most modem options and parameters are software selectable o Self-test capability o Remote diagnostic capability o Built-in Network Management features VAXPAC VAX Public Access Communications software provides the VMS user with a mechanism for connecting to a remote system using asynchronous lines. DFA01/VAXPAC MicroVAX 3500/3600 System Product Features: DFA01/VAXPAC MicroVAX 3500/3600 System -- A VMS-based asynchronous communications system consisting of DFA01s and VAXPAC VMS software providing easy-to-use, cost-efffective remote access for multiple users. o Modem selection and auto dial Telephone Management features o Session logging o File transfer (Kermit is supplied with the packet, as are "hooks" for third-party products) o Stores numbers for standard and personal connections o Remote terminal connection makes any terminal appear as a terminal on a remote system FEATURES/BENEFITS Cost Per User o Low cost per user -- DFA01s can be shared by all system users. For example, 30 Microvax 3500/3600 users can share 3 DFA01s and VAXPAC for as little as $310 per user. Based on these guidelines, the DFA01/VAXPAC system price is one of the lowest cost modem communications system alternatives in the industry. DFA01 hardware-only price compares favorably with most 2400 bits/s modems plus DZQ interface and RS232 cables. Ease of Use o Easy-to-use human interface -- menu-driven software. Both remote terminal access and file transfer are supported. o Neat packaging -- No additional cables or boxes required. Modem setup is simplified by easy-to-use human interface software. PRICING/ORDERING INFORMATION The decision table below lists all DFA01 hardware and DFA01/VAXPAC part numbers and prices for MicroVAX 3500/3600 System products. U.S. No. DFA01 No. VAXPAC* Cost Per User Part No. List Price Modules Modems Software (30 Users) (License & H-Kit) Factory Installed: DFA01-AA $ 1,650.00 1 2 No DJ-DFA01-AA 6,300.00 1 2 Yes $210.00 DJ-DFA01-BA 7,800.00 2 4 Yes 260.00 DJ-DFA01-CA 9,300.00 3 6 Yes 310.00 DJ-DFA01-DA 10,800.00 4 8 Yes 360.00 * Please note that software installation charges are not included. U.S. No. DFA01 No. VAXPAC* Cost Per User Part No. List Price Modules Modems Software (30 Users) (License & H-Kit) Field Installed (Price includes hardware field installation): DFA01-AF $ 1,902.29 1 2 No DJ-DFA01-AF 6,552.29 1 2 Yes $218.41 DJ-DFA01-BF 8,159.72 2 4 Yes 271.99 DJ-DFA01-CF 9,767.15 3 6 Yes 325.57 DJ-DFA01-DF 11,374.58 4 8 Yes 379.15 NOTE: There are no restrictions on the number of DFA01/VAXPAC MicroVAX 3500/3600 System users who may share a pool of modems. The number of simultaneous remote connections is limited by the number of modems installed. * Please note that software installation charges are not included. Customers buying a new MicroVAX 3500/3600 VMS system should order as follows: TABLE 1 ____________________________________________________________________________ IF: AND THEN ORDER ____________________________________________________________________________ The customer is also wants one or DJ-DFA01-XA purchasing a MicroVAX more DFA01s and VAXPAC Factory installed. 3500/3600 VMS System software as part of configuration The customer is wants additional DFA01-AF purchasing a MicroVAX modems after he has 3500/3600 VMS System already purchased a modem/software package ============================================================================ Aftermarket VMS-Based Packages: TABLE II ____________________________________________________________________________ IF: AND THEN ORDER ____________________________________________________________________________ The customer has wants an add-on DJ-DFA01-XF already purchased modem and software Field installed. a system, but without package a DFA01 or VAXPAC The customer has wants an add-on modem DFA01-AF already purchased without software a system, but without a DFA01 or VAXPAC ============================================================================ PDP-11-Based System Packages -- Software Not Available: TABLE III ____________________________________________________________________________ IF: AND THEN ORDER ____________________________________________________________________________ The customer is wants a modem DFA01-AA at the purchasing a PDP-11 time of system sale based system The customer has wants a modem DFA01-AF already purchased a as an add-on option PDP-11-based system NOTE: Only RSX-llM-PLUS V4.0, Micro/RSX and VMS 4.6 has support for the DFA01 at this time. ============================================================================ ================================================================================ Note 40.16 PERIPHERAL DEVICE ANNOUNCEMENTS 16 of 56 EISNER::KILLEEN "Jeff Killeen" 91 lines 10-OCT-1987 14:58 -< TK50 FCO'S >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! TK50 -- FIELD CHANGE ORDERS (FCOs) - 30-September-1987 o TK50 FCO's improves reliability, eliminates known problems and enhances ease of use o TK50 FCO's -- free of charge to customers with a Digital maintenance contract o Currently being implemented by Field Service Repair Centers Over the past two and a half years, the TK50 has become the most widely used tape drive in its class, and the majority of users have been very pleased with the device. As happens with any new technology, a few customers have experienced problems and these have attracted a lot of attention. Tape Engineering has studied these reported problems and has implemented several enhancements to improve user satisfaction. The improvements have two objectives: o Correct a few real deficiencies o Make the drive more tolerant of the mistakes often made by untrained users Correct operation of the drive is extremely important and trained users are the most valuable enhancement to the TK50's operating reliability. CARTRIDGE IMPROVEMENTS Users may encounter problems if a cartridge has been dropped onto a hard surface. The shock can knock the tape leader out of position, making automatic threading of the tape impossible. We have now added a shock hook to the cartridge to make the tape leader more resistant to displacement if the cartridge is dropped. Problems can also occur if a user inserts a cartridge most of the way into the drive and then pulls it back out before the cartridge snaps in place in the drive. An anti-teasing notch has been added to the cartridge shell to prevent users from accidentally getting the cartridge locked out of the drive. Cartridges with both improvements began shipping in April 1987 and can be identified by a "9" as the first digit of the 5-digit serial number on the bottom of the cartridge. DRIVE IMPROVEMENTS Users have reported a variety of problems with runaway leaders (a flapping sound in the drive) and with cartridges occasionally getting stuck in the drive. Most of these problems were caused by user mistakes; a few were caused by problems in the drive. A new FCO includes mechanical and microcode changes to the drive and makes the drive much more tolerant of common mistakes. The enhancements also allow a user to recover from errors that would previously have required a field service call. The FCO is currently being implemented by Field Service and expectations are that it should take between 9-12 months to complete. This FCO is free of charge to those customers who have purchased a Digital maintenance contract. The current strategy is to initially replace field units that need to be repaired and later to replace those field units which are working properly but do not include the FCO revisions. An up-to-rev drive will be labeled Rev E (or higher) on the back of the drive baseplate, which can be checked by a Field Service engineer. CONTROLLER IMPROVEMENTS In some cases, users have gotten an error message stating that the drive was "not software enabled" or identifying the unit with a phantom device number. Other users have had problems with system recognition of the TK50 after the drive is powered down and then powered up again (in a table-top package) while the system remains operating. TK50 controller code improvements eliminate these problems and dramatically improve reliability in PDP-11 configurations by eliminating bus-related problems. This FCO was released many months ago and most users should have received it if they were experiencing these problems. An up-to-rev controller will be labeled Rev F1 or higher. The basic message to any customers who have complained about the TK50 is that: o A few problems have been identified and fixed o Correct operation is important and users must be trained o The subsystem is reliable ================================================================================ Note 40.17 PERIPHERAL DEVICE ANNOUNCEMENTS 17 of 56 EISNER::KILLEEN "Jeff Killeen" 208 lines 10-OCT-1987 14:58 -< TK70 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! TK70 SUPPORT FOR MicroVAX II SYSTEMS - 29-September-1987 o The TK70 cartridge tape subsystem recently announced for MicroVAX 3xxx series systems, is now available for MicroVAX II systems o TK70 is a major advance in reliability, capacity, performance and ease of use o TK70 can read TK50 cartridges TK70, the second member of Digital's low-end cartridge tape drive family, is now available as a component of MicroVAX II and MicroVAX 3xxx series systems and as an option for MicroVAX II configurations. The phenomenal growth in the use and power of MicroVAX systems over the past few years, as well as the increased capacity of the fixed-media disks used on those systems, has created a need for a higher-capacity, faster- performing, highly reliable tape subsystem to back up those disks and protect user data. TK50 users have also told us that they would like a front-panel that was easy for novice users to understand, so they would be more likely to use the drive correctly. PRODUCT DESCRIPTION The TK70 meets all of those needs -- and more -- in a compact, elegant package. The TK70 is a major advance in capacity, performance, reliability and ease of operation. Designed to be used on high-end Q-bus systems, it will satisfy your customers' needs for disk backup, software distribution, data collection and data interchange between microsystems. The subsystem consists of the TK50-sized (5-1/4" form factor) tape drive, the dual-height Q-bus controller and cables. Power supply and cabinetry are supplied by the system box in which the TK70 is mounted. Unlike the TK50, TK70 is not currently available in desktop or rack-mount packages. The new subsystem uses an enhanced version of the CompacTape cartridge. The CompacTape II cartridge (model number TK52-K) appears virtually identical to the TK50-K and contains the same magnetic media, but is manufactured to tighter tolerances to accommodate the high-density 296 MB format of the TK70. TK52-K cartridges should only be used with TK70 drives; TK50 users should continue to order TK50-K cartridges. Even though a brand new TK50-K cartridge might work on a TK70 drive in TK70 mode, this is not a supported configuration and support cannot be guaranteed. The new TK70 can read (but not write) TK50 formatted data, so libraries of older cartridges can be read on the new subsystem, and OEMs who distribute software on TK50-K cartridges can continue to do so, whether their customers have TK50 or TK70 drives. Digital will continue to distribute software on TK50-K cartridges. FEATURES/BENEFITS Feature Benefit o 296 MB maximum formatted capacity* TK70 backs up any of Digital's mini-Winchester disks onto one cartridge -- no operator inter- vention is needed. It also can distribute large programs and databases with ease. o Robust parts/intelligent controller High reliability; the TK70 will be available when you need it. *Actual capacity may be slightly less in certain situations. PERFORMANCE and SPECIFICATIONS were described in the announcement article in the September 2 Special Issue. Feature Benefit o Redesigned front panel Provides clear information and helps users operate the TK70 correctly. o Read-after-write, error correction High data integrity; this combina- code (ECC) and cyclic redundancy tion of features is a major advance (CRC). among microsystem tape drives. o 90 KB/sec formatted data transfer Fast performance; can back up RA70 rate (125 KB/sec unformatted rate) on MicroVAX 3500/3600s in less than one hour. o 64 KB Cache Enhances performance o No field adjustments, Ease of use and low cost of owner- No preventive maintenance, ship. No cleaning of heads needed in normal use. o Pocket-sized rugged cartridge Easy transportability for data and software. o Calibration bursts and automatic Guarantees media interchangeability gain control between subsystems o TK50 read compatibility New drive can read archives of old cartridges. TK50-mode can be used for software distribution to both TK50s and TK70s. o DSA compatibility No software changes needed; invest- ment protection. CONFIGURATION INFORMATION Option I/O Panel Units TK70-** N/A TQK70-** 1 "A" panel POSITIONING TK70 is positioned between TK50 and TU81-Plus, in terms of performance and price. Its capacity per cartridge is much higher than any of our other magnetic tape subsystems. TK70 is the device of choice for MicroVAX 3500 and 3600 based systems, and for large MicroVAX II based systems in most situations. Generally, if the system has an RD54 (159 MB) or larger disk, TK70 is your best choice. For smaller systems TK50 is a more cost-effective choice. TU81-Plus is best for larger configurations. It is also recommended for boot members of Local Area VAXcluster systems if the cluster has many disks to be backed up to the boot member, or if the customer needs industry-standard nine-track tape for interchange purposes. The TU81-Plus offers significantly faster performance that is better matched to the needs of larger systems, and provides industry-compatible data recording at 6250 bpi (bits per inch) and 1600 bpi. ORDERING INFORMATION The TK70 will be available for Q-bus systems only. A controller and drive must be ordered separately to configure a TK70 subsystem. Controller Part Numbers (order one): U.S. Order No. Description List Price BMC TQK70-AA M7559 controller board, plus 30" cable for BA23 $1,700 $10 TQK70-BA M7559 controller board, plus 30" cable for BA123 1,700 10 TQK70-SA M7559 controller board, cable, filler panel 1,700 10 w/labels, dual blank module, for BA213 TQK70-SF Field installation kit for BA213: M7559, cable 1,700 10 (17-01363-01), filler panel w/labels, dual blank module, installation guide Drive Part Numbers (order one): TK70-AA TK70 drive for BA23 and BA123 box (MicroVAX II) $5,600 $30 TK70-EA TK70 drive for BA2xx boxes (MicroVAX 3xxx 5,600 30 series systems) TK70E-SA TK70 drive with BA2xx box mounting hardware 5,600 30 TK70E-SF Field installable TK70 for BA2xx 5,600 30 Documentation: User information for the TK70 will be included with each computer system manual. A documentation set will be shipped with each TK70 drive. Documentation part numbers are as follows: Description Option No. TK70 Tape Drive Subsystem User's Guide EK-OTK70-UG-001 TK70 Tape Drive Subsystem Technical Manual EK-OTK70-TM-001 QUESTIONS AND ANSWERS Q. Will the TK70 replace the TK50? Not in the foreseeable future. The two products complement one another and there is a substantial price difference between them. TK70 is aimed at larger MicroVAX systems with more disk storage to be backed up, while TK50 remains the device of choice for smaller, low-cost systems and for UNIBUS systems that need to interchange data with MicroVAX systems. Q. What happens if I use a TK50 cartridge in a TK70? There is no problem reading a TK50 cartridge in a TK70. If you try to write on a TK50 cartridge with a TK70 drive, several things may happen. As background, you should know that TK50 and TK70 drives write calibration tracks on tape the first time a cartridge is loaded into a drive, and TK50 calibration tracks are different from those written by a TK70. If the cartridge was ever mounted in a TK50, the TK70 will sense the TK50-format calibration tracks on the tape and will automatically write-protect the cartridge. If the TK50 cartridge is brand new, the TK70 will write on it, but may reject it part way through the process if excessive data errors occur. This can happen if the TK50 tape does not meet the tighter tolerances required by the high-density TK70 data format. ================================================================================ Note 40.18 PERIPHERAL DEVICE ANNOUNCEMENTS 18 of 56 EISNER::KILLEEN "Jeff Killeen" 38 lines 10-OCT-1987 15:03 -< BA23-UC UPGRADE KIT >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! ANNOUNCING AVAILABILITY OF THE BA23-UC - 28-September-1987 o BA23 6-button front panel Upgrade Kit for MicroPDP-11 systems o Available today MicroPDP-11/73 and MicroPDP-11/83 customers can now upgrade their existing BA23-based systems to accommodate dual RD32 disk drives. The RD32, 42 MB, half-height Winchester disk drive is the replacement product for the RD52 (31 MB disk). Support for the RD32 on the MicroPDP-11/73 and the MicroPDP-11/83 was announced on June 29, 1987. Since that time, all MicroPDP-11/73 and MicroPDP-11/83 systems shipping in a BA23 enclosure have the same front panel as the MicroPDP-11/53 -- a six-button front panel. PRODUCT DESCRIPTION The BA23-UC Upgrade Kit consists of a 6-button front panel, labels and the installation manual. The 6-button front panel contains an extra "Write/Protect" and "Ready" button to accommodate the second RD32 drive. Prior to June 29, all MicroPDP-11/73 and MicroPDP-11/83 systems shipped with a 4-button front panel. ORDERING/AVAILABILITY INFORMATION U.S. Model No. Description List Price Seg. BMC Inst. BA23-UC BA23 6-Button front $200 XX N/A P/C panel upgrade kit ================================================================================ Note 40.19 PERIPHERAL DEVICE ANNOUNCEMENTS 19 of 56 EISNER::KILLEEN "Jeff Killeen" 80 lines 10-OCT-1987 15:04 -< LG01/LG02 MODELS FOR THE BA213 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! LG01/LG02 PRINTER VARIANTS - 21-September-1987 FOR MicroVAX 3500/3600 in BA2xx ENCLOSURE o LG01 and LG02: - The only Digital line printers available for use with the new BA2xx enclosures o New LPV11-SA controller can support two line printers INTRODUCTION The LG01 600 lpm text printer and the LG02 600 lpm text and graphics printer can now be connected to the new BA2xx enclosure MicroVAX 3500 and MicroVAX 3600. PRODUCT DESCRIPTION NEW FEATURE The new LPV11-SA controller provides the same functionality as the LPV11 controller and operates under VMS; however, the LPV11-SA can support two LG01 or LG02 line printers at the same time. The two ports on the LPV11-SA control module provide backplane space savings for those customers who need to connect two LG0x line printers to one MicroVAX 3500 or 3600. The LG01 text printer offers data processing and correspondence mode printing, underlining, bolding, strike through and multiple character/inch print selections. The LG01 can be field upgraded to the LG02. The LG02 text and graphics printer offers barcodes, custom forms creation, subscript/superscript, landscape mode and many other graphic functions. Both of these line matrix printers handle a full range of continuous fanfold paper, including 1-6 part forms. SPECIFICATIONS The controller for these LG0x variants is a full quad-sized board which draws 1.75 Amps (2.2 Amps maximum) at +5 Volts (8.75 Watts), no 12-Volt draw. PRICING/ORDERING INFORMATION Following are the model numbers, descriptions and pricing for the four new LG01 and LG02 printer variants. Model No. Description LG01-EA 600 lpm Text printer + LPV11-SA controller + BC27L-30 cable 60 Hz/120 V LG01-JA 600 lpm Text printer + BC27L-30 cable 60 Hz/120 V LG02-EA 600 lpm Text/Graphics printer + LPV11-SA controller + BC27L-30 cable 60 Hz/120 V LG02-JA 600 lpm Text/Graphics printer + BC27L-30 cable 60 Hz/120 V U.S. Model No. List Price BMC DSMC Inst. LG01-EA $12,900 $133 $158 $570 LG01-JA 12,900 118 140 570 LG02-EA 15,900 133 158 570 LG02-JA 15,900 118 140 570 NOTE: The variants designated -JA have no controller and are for owners of the -EA variants who are adding a second printer. For further information, please consult your Digital Representative. ================================================================================ Note 40.20 PERIPHERAL DEVICE ANNOUNCEMENTS 20 of 56 EISNER::KILLEEN "Jeff Killeen" 121 lines 10-OCT-1987 15:04 -< TSV05 MODELS FOR THE BA213 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! TSV05 INDUSTRY STANDARD MAGTAPE - 14-September-1987 FOR MicroVAX AND VAXserver 3500 AND 3600 SYSTEMS o TSV05 Magtape with enhanced controller for new BA200 series Q-bus system including MicroVAX 3500 and 3600 o TSV05 H9642 style cabinet variant matches MicroVAX 3600 cabinet and now includes side panels o Integrated TSV05 option for VAXserver 3602 system o New cable kits for BA200 series enclosures included o Available October 1987 PRODUCT DESCRIPTION The new TSV05 option variants consist of the standard TS05 1600 bits/in phase encoded tape transport system with an enhanced Q-bus controller. The controller differs mechanically from the conventional Q-bus controller in that the bulkhead distribution panel is integral with the TSV05 Control module handle. Thus, a pair of identical cables is all that is required to connect the tape transport to the control module. TSV05-SA, SB, SC, SD This TSV05 variant family provides a rack-mount transport with quad-size module and a 16-foot cable kit. Customers purchasing this variation are generally mounting the TSV05 into an existing Digital or EIA-SE-102 Standard Cabinet. TSV05-SE, SF, SH, SJ This TSV0O5 variant family offers a top-mount TS05 transport in a Digital 40" cabinet. This arrangement provides convenient top cover access for tape head cleaning, and allows expansion for two 10.5-inch devices. A quad-size module, cabinet side panels, a 16-foot cable kit, and an integral power controller at the cabinet base are included. TSV05-SK, SL, SM, SN This new variant family provides a convenient integrated TSV05 option for the top opening of the second VAXserver 3602 system cabinet. This option can be factory or field installed and includes top access cover, quad-size module and necessary cabling. FEATURES/BENEFITS Feature Benefit Compact Increased expansion space Auto loading Saves operator time Low noise level Acceptable in office evironment SPECIFICATIONS Hardware Requirements -- See Current Systems and Options Catalog Software Requirements -- A VMS layered Driver (Q0Z99/Q1Z99) is required for any VMS 4.X version. ULTRIX-32 has the I/O Driver embedded in the operating system. Configuration Information Power Bus Loads Mounting Requirements @+5 V @+12 V Watts AC DC Module 1 Quad Slot 6.5 0 32.5 3.0 1.0 Tape Transport See current Systems and Options Catalog. PRICING/ORDERING INFORMATION U.S. Model No. Description Seg. List Price BMC TSV05-SA BA2XX QBUS MTAPE W/O CAB 120 V 090 $10,800 89 TSV05-SB BA2XX QBUS MTAPE W/O CAB 240 V 090 ** TSV05-SC BA2XX QBUS MTAPE W/O CAB 100 V 090 ** TSV05-SD BA2XX QBUS MTAPE W/O CAB 220 V 090 ** TSV05-SE BA2XX QBUS MTAPE H9642 CAB 120 V 090 12,800 89 TSV05-SF BA2XX QBUS MTAPE H9642 CAB 240 V 090 ** TSV05-SH BA2XX QBUS MTAPE H9642 CAB 100 V 090 ** TSV05-SJ BA2XX QBUS MTAPE H9642 CAB 220 V 090 ** TSV05-SK BA2XX MAGTAPE FOR 36XX CAB 120 V 090 11,200 89 TSV05-SL BA2XX MAGTAPE FOR 36XX CAB 240 V 090 ** TSV05-SM BA2XX MAGTAPE FOR 36XX CAB 100 V 090 ** TSV05-SN BA2XX MAFTAPE FOR 36XX CAB 220 V 090 ** Additional Cables: CK-TS05-15 BA2XX/TSV05 8 FT. CABLE KIT 090 $600 N/A CK-TS05-16 BA2XX/TSV05 16 FT. CABLE KIT 090 600 N/A **Consult your Digital Rperesentative SERVICES/SUPPORT The TSV05 comes complete with a full one year, on-site DECservice warranty. Following this full-service warranty period, Standard Digital Services will be available. Installation is included when the tape subsystem is purchased as part of a system order. Add-On Installation is available for those customers purchasing separately. AVAILABILITY The TSV05-S(X) variants and Cable Kit will be available in October 1987. ================================================================================ Note 40.21 PERIPHERAL DEVICE ANNOUNCEMENTS 21 of 56 EISNER::KENNEDY "Terry Kennedy" 14 lines 12-OCT-1987 21:37 -< BA23-UC where are you? >- -------------------------------------------------------------------------------- > o BA23 6-button front panel Upgrade Kit for MicroPDP-11 > systems > > o Available today DEC Direct won't accept an order for this part (BA23-UC) - you have to order it through your local Sales Office. The local Sales Office doesn't want to sell it because a) It is a very low cost item b) It is only available with Field Service installation. Funny - they trust us to replace backplanes, but not write protect switches? If anyone knows a way around this, please post a reply here. ================================================================================ Note 40.22 PERIPHERAL DEVICE ANNOUNCEMENTS 22 of 56 EISNER::KILLEEN "Jeff Killeen" 109 lines 13-OCT-1987 23:10 -< DESPR - THIN WIRE TO THICK WIRE REPEATER >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! ThinWire ETHERNET SINGLE-PORT REPEATER - 13-October-1987 o Single-port ThinWire repeater ideal for applications requiring high-speed, flexible networking at the desk o Available now In response to customer demand, Digital announces a ThinWire Ethernet single-port repeater (DESPR), a low-cost ThinWire connection to standard Ethernet backbone. PRODUCT DESCRIPTION DESPR delivers 10 Mbps Ethernet performance and full DECnet functionality to the desktiop and work area for connection of PCs, workstations, network servers and low-end computing devices. DESPR is ideally suited for applications requiring high-speed, flexible networking at the desk. Designed for localized office and work areas requiring Ethernet connectivity for isolated PCs and workstations, DESPR allows customers to daisy chain up to 29 stations. It connects a single ThinWire coaxial cable segment of up to 185 meters in length to any standard Ethernet backbone via an Ethernet transceiver like the H4000-BA or DELNI. It can be rack-mounted, wall- mounted or used as a table-top device. FEATURES/BENEFITS Feature Benefit o Single-port ThinWire (BNC) connection o Low-cost initial investment for to a standard Ethernet backbone. corporate network connections. o Connects isolated, small groups of PCs or workstations to a standard Ethernet backbone via ThinWire. o Allows connection of devices that only use ThinWire (BNC) connectors. o Table top, rack or wall mount. o Configuration flexibility. o Diagnostic Capabilities - Automatic o Protects the backbone network, Segmentation, Segmentation LEDs enhances fault isolation and serviceability. CONFIGURATION Two Major User Configurations The ThinWire Ethernet products will serve users in the floor area of a facility as contrasted with standard Ethernet products which are targeted for use between floors and buildings at a customer site. The ThinWire Ethernet products are ideally suited for work area networks, e.g., small stand-alone local area networks, where multiple PC/Workstation users in a work area share resources, such as printers and storage devices. PRICING/ORDERING INFORMATION The DESPR product will be shipped in early Q2/FY88. U.S. Product Order No. Description List Price P/L Available DESPR-AA ThinWire Ethernet $875 XX Q2/FY88 Single-port Repeater, 120V* DESPR-AB ThinWire Ethernet *** XX Q2/FY88 Single-port Repeater, 240V** * DESPR-AA is shipped with a power cord that can be used in the U.S., Canada, Mexico and Japan. ** DESPR-AB is not shipped with a power cord -- it must be ordered separately. See the list below for the correct power cords for each country. ***Consult AQS. POWER CORDS U.S. Order No. List Price Country BN02-2E * United Kingdom, Ireland BN03-2E * Austria, Belgium, Finland, France, Germany, Netherlands, Norway, Sweden, Hungary, Poland, Czechoslovakia, Portugal, Spain BN04-2E * Switzerland BN05-2E * Australia, New Zealand BN06-2E * Denmark BN07-2E * Italy ================================================================================ Note 40.23 PERIPHERAL DEVICE ANNOUNCEMENTS 23 of 56 EISNER::KILLEEN "Jeff Killeen" 163 lines 13-OCT-1987 23:11 -< DF242-CA MODEM WITH SECURITY AND MNP >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! THE DF242-CA SCHOLAR Plus MODEM - 13-October-1987 o Callback and Access Security - Prevent unauthorized use of your database o Supports Hayes "AT" Command Set and Digital Modem Command Language (DMCL) o MNP and X.PC On-Line Error Correction protocols provide error-free communication over standard telephone lines o Available now Unauthorized dial-up access to a host system or database is a growing concern of many companies. Additionally, transfer of data must be done efficiently and accurately over the varying quality of voice-grade telephone lines. Modems must be easy to install, maintain, operate, and not require constant changing of soft and hard switches. The DF242-CA offers a new security system to protect databases, error correction to assure accurate data transfers at speeds up to 2400 bits/s, and the simplified Digital Modem Command Language (VAX/VMS-like commands) for ease in setting modem parameters. To complete the package, up to 30 telephone numbers can be stored and linked by a predefined label for simplicity in dialing. PRODUCT DESCRIPTION The DF242-CA Scholar Plus is a stand-alone desktop modem that brings many new features to the Digital family of modems. The DF242-CA offers a unique security access feature. Four levels of security can be set: 1. Password Access -- A predefined password must be keyed before the host system is accessed. An incorrect password causes the modem to disconnect. 2. Password and Callback -- The password is predefined along with a telephone number. The modem requests the password, and then disconnects. The modem recalls the telephone number associated with the password, and initiates a call to that known number. 3. Password, Callback, Telephone Number Checked -- The password and telephone numbers are preset. The modem requests that the dial-in caller key in a password and callback phone number correctly prior to disconnect and callback to the known location. 4. Password, Request for Phone Number -- The password is requested. The user can then type an alternate phone number or the number assigned to that password as the callback number. For complete system security an audit trail of dial attempts is maintained, and even the modem parameter settings are secured by password so that unattended modems cannot be modified. All modems share the same purpose and operating principles, but differences between modems emerge as the telephone line quality deteriorates. The Scholar Plus, with the industry-standard MICROCOM Networking Protocol (MNP) on-line error correction, will automatically retransmit if errors are found. When connecting to the Tymnet Value-Added Network, X.PC error correction assures error-free transmission. When a "no answer" or "busy" signal is received (the built-in speaker allows call progress monitoring), the modem auto-dialer will automatically link to an alternate number after a predetermined number of rings. A user can store up to 30 numbers -- local or long distance -- each number can be up to 36 digits in length. These numbers can be assigned "labels" or names rather than typing the phone number itself. The Digital Modem Command Language (DMCL) is an alternative available in the Scholar Plus. DMCL lists, in a simplified manner, the different parameters available. The user has a choice of changing one parameter, many, or all parameters at any point in time. On-line ease-of-operation features include speed buffering up to 9600 bits/s (terminal speed does not have to be changed to match modem speed); automatic speed adjustment, up or down, to match the speed of the remote modem; and additionally, the modem parameters may be reviewed for accuracy, off-site, by Digital Field Service. Once set, the parameters are stored in nonvolatile RAM, so they are not forgotten when the Scholar Plus is switched off. The industry standard HayesR "AT" command language is also an option on the DF242-CA for experienced modem users; they will not have to learn a new command structure. To use the security and number-linking features, the DMCL mode is required. Information Sheet: ED-30591-78 User's Guide: EK-DF242-UG-001 FEATURES/BENEFITS Feature Benefit Callback/Access Security Data protected from unauthorized use Audit Trail Unauthorized attempts to access host system can be monitored MNP Class 2 and 3 error correction Accurate data transfer over dial-up lines, compatibility with other industry standard modems Automatic fallback and speed detect No need to change terminal or modem speeds when communicating with compatible modems No user intervention Speed Buffering to 9600 bits/s No need to adjust terminal speed to modem speed "AT" command set and Easy to set and change, parameters Digital Modem Command Language are stored in nonvolatile RAM Linkage to alternate telephone No need to redial when a line is numbers (DMCL mode only) busy or no answer Remote parameter review for off-site Digital Field Service can check, problem solving support from off-site, parameter settings Internal Audible Speaker Allows call progress monitoring SPECIFICATIONS o 2400 bits/s, CCITT V.22 bis o Async. 5,6,7,8 data bit, or sync. 1200 bits/s, CCITT V.22 Bell 212A 300 bits/s, Bell 103J o Data Rates: -High Speed, 2400 bits/s sync. and async. -Low Speed, 1200/300 bits/s async. 1200 bits/s sync. o Full duplex, all speeds o Wall-mounted power supply o EIA RS232-D, RS423 Interface Compatibility NOTE: Security, error correction and speed buffering are available in asynchronous mode only. PRICING/ORDERING INFORMATION U.S. Order No. Description List Price DF242-CA 2400 bits/s Scholar Plus modem $795 with wall mounted power supply, telephone extension cord User's Guide NOTE: Data cable is not supplied, Digital part number BC22E is recommended for asynchronous operation. AVAILABILITY Available now in the U.S. and Canada. Bell is a trademark of Bell Telephone Companies Hayes is a registered trademark of Hayes Microcomputer Products Inc. MICROCOM is a registered trademark of Microcom, Inc. TYMNET is a registered trademark of Tymnet, Inc. ================================================================================ Note 40.24 PERIPHERAL DEVICE ANNOUNCEMENTS 24 of 56 EISNER::KILLEEN "Jeff Killeen" 409 lines 13-OCT-1987 23:14 -< DECSERVER 500 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! ANNOUNCING THE DECserver 500 - 12-October-1987 o The DECserver 500 provides high-density flexible packaging for terminal interconnect solutions o Attractive cost-per-line benefits for large configurations o Compatible with all current terminal servers through the use of LAT protocol o Limited availability now; volume availability late Q2/FY88 THE DECserver SOLUTION The DECserver 500, the newest member of the DECserver family of Ethernet- based terminal interconnection products, expands the product set to include servers ranging from 16 to 128 lines. With the DECserver 500, the customer now has the capability of combining support for multiple signalling types, such as RS232-C devices (for use with modems, data switches and processors that do not support the LAT protocol) and DEC423 devices (commonly used with a DECconnect installation) in the same package. The DECserver 500 V1.0 software, coupled with the DSRVS-A*/B* hardware and the desired types and quantities of controller cards, provides a highly flexible, cost-effective, high-end terminal server. Previously, the highest density terminal server available was the 32-line Ethernet Terminal Server (DECSA-DA). The DECserver 500 is intended to replace all future orders for DECSA-DA products and to be used where even higher density would be very desireable, such as limited space SERs (Satellite Equipment Rooms). The required load software, which resides on at least one processor on the LAN, is down-line loaded upon power-up or re-initialization of the DECserver 500. It implements the same LAT architecture used for all of the existing Digital terminal server products. A license is required for each DECserver 500 and will be INCLUDED in the price and part number of the DECserver 500, a change from the previous ordering methods for the DECserver 100 and 200. PRODUCT DESCRIPTION With its modular growth capability, the DECserver 500 will be a flexible terminal network switch that will support up to 128 asynchronous lines in a DEC423, data-only configuration. With the two communication controller cards available at FCS, CXA16 (DEC423 data only) and CXY08 (RS232 modem control), this product will support customers requirements for data-only DEC423 and modem control RS232-C concurrently. Remote connection via dial-in modems also is fully supported with the CXY08 option card. The DECserver 500 is available in two basic versions. One for rack mounting (DSRVS-AA/AB), which includes the rack-mount brackets, and one for stand- alone configurations (DSRVS-BA/BB), which includes suitable cabinetry enclosures. The required number of line controller cards are ordered separately. FEATURES/BENEFITS The DECserver 500 will provide your customer essentially the same user features and benefits provided by both versions of the DECserver 200 combined. Configurations must include the correct option card(s) for the DECserver 500, either a CXY08 (RS232) for the modem-control features you might use a DECserver 200/MC for, and the CXA16 (DEC423) for the DECserver 200/DL features. This multiple interface structure provides our customers the connectivity they need and the flexibility they require, all at an attractive cost per line. All features pertaining to modem control, modem support and modem-signal flow control apply to the CXY08 eight-line controller card hardware only. The CXY08 (RS232) provides foreign-host capability to connect host systems that do not support the LAT protocol, thereby extending the access of a single user's terminal to all host processors, both Digital and non-Digital. Dial-out modems can also be connected in this fashion to provide outbound service to remote processors, again from the same terminal used for local connections. The CXA16 (DEC423) should be ordered for all connections that do not need to implement modem control or foreign-host connections. This 16-line device is the superior choice for terminal and Digital printer connections, providing greater distance (up to 1,000 feet), and twice the line density (up to 128 lines) within your DECserver 500 when used in place of the CXY08 (64 lines maximum). The DECserver 500 allows VMS and MicroVMS (V4.6 and later) host-initiated connections to asynchronous devices of many types to be established. Devices can range from high-quality laser printers and X-Y plotters to barcode readers and badge scanners. Digital provides a specific Terminal Server Print Symbiont for VMS and MicroVMS hosts, which can initiate connections to asynchronous printers connected to DECserver 200 or DECserver 500 ports. This allows the printers to be distributed throughout a facility and accessed transparently by many nodes on the Ethernet, thereby reducing the effective cost per user for these types of devices. DECserver POSITIONING DECserver 500 versus DECserver 200 Functional Positioning DECserver 500 DECserver 200 Lines/Server 16-128 8 Expandability 8-line increments /mc none (new box required) 16-line increments data Modem control Yes, if desired via Yes (with /MC) 8-line /mc cards CXY08 Customer installable No Yes Connection to non-LAT hosts Yes (with CXY08) Yes (with /MC) Support nonkeyboard Yes Yes Devices Physical packaging 25" h 4-5/8" h 17-1/4" w 19-1/4" w 12" d 12-5/8" d Economical Cost-Per-Line The DECserver 500 provides an economical solution when greater than 64 terminal connections are considered, either in the short term or in the future. As an example, the following table compares the cost-per-line for the DECserver 200/DL with the DECserver 500, utilizing DEC423 terminal connections. The DECserver 200/DL cost-per-line is fixed at $420. The DECserver 500 becomes less expensive at or above 64 lines. Cost Per Line No. of Lines DECserver 500 DECserver 200/DL 48 $502 $420 64 419 420 80 369 420 96 336 420 112 313 420 128 295 420 SOFTWARE DECserver 500 software is down-line loaded over the LAN from a Phase IV DECnet End-Node Load Host. The required node ACCESS software is provided with the standard VMS/MicroVMS operating system kit. This software supports the simultaneous operation of up to 128 asynchronous devices -- data only, or 64 devices with full modem control (and combinations in between), at speeds from 75 bits/s to 19.2K bits/s. The software supports the following features: o Full modem control and non-LAT host support (CXY08 only) o XON/XOFF software flow control o Both CTS/RTS and DSR/DTR hardware flow control (CXY08 only) o Data transparency mode for PC file transfer (i.e., Kermit) applications o Ability to pass break character and error notifications o Multiple sessions (up to 256 sessions per DECserver 500) o Security features such as port access passwords and service grouping o On-line Help features, both tutorial and reference o Automatic load balancing over VAXclusters (including LAVc) SUPPORT Non-LAT hosts can be connected to the network via the DECserver 500 using the CXY08, provided they support XON/XOFF, ASCII Standards and EIA RS232-C/CCITT V.24/V.28 interfaces. The CXA16 controller card can be used with any DEC423-compatible device or RS232 device using either the DECconnect passive or active adapter, depending on the line distance desired. Terminal Devices DECserver 500 software supports the following Digital terminal devices: o LA12, LA34, LA36, LA38 o VT100, VT101, VT102, VT125, VT131 o VT200 series o VT300 series Supported terminal parameters: Character size: 7 or 8 bits per character Parity: even, odd, mark, space or none The automatic line-speed detection (Autobaud) feature is supported for either 7-bit characters with even parity or 8-bit characters with no parity. NOTES: 1. The DECserver 500 supports the Professional, Rainbow, VAXmate and DECmate personal computers when used in terminal emulation/file transfer mode (via an integral capability such as "T" on the Rainbow, or using layered software such as Kermit, WPS or POLY-TRMTM). 2. This product CANNOT be used to connect PCs (including IBM personal computers) on a network as asynchronous DECnet nodes. DECnet-Rainbow, DECnet-VAXmate, DECnet-DOS and PRO/DECnet software require serial connections provided by DECnet routing nodes such as the DECnet Router Server or DECrouter 200; A DECserver CANNOT satisfy this requirement. 3. This product is NOT WARRANTED to support non-Digital terminal devices. However, terminals supporting VT200-like characteristics and personal computers supporting IBM PC, IBM Personal Computer XT and ATTM characteristics SHOULD operate with this product. Printers DECserver 500 software supports the following Digital asynchronous printers when accessed from VMS and MicroVMS systems: o LA12, LA34, LA36, LA50, LA75, LA120, LA210 o LQP02, LQP03 o LXY12-DA, LXY22-DA o LN01S, LN03, LN03S, LGP01, LJ250 Modems DECserver 500 software supports the following Digital asynchronous modems with the CXY08* option card: o DF03, DF112, DF124, and DF224 full-duplex asynchronous modems for either dial-in or dial-out use. Also supported are private or leased line modem applications. *Please make sure that the CXY08 hardware is ordered if modem support is desired. PRICING/ORDERING INFORMATION DECserver 500 Hardware When purchasing a DECserver 500, order the DSRVS-xx plus a minimum of two option cards. This two-option-card requirement will be monitored through the XCON/XSEL ordering system. The customer will be permitted to choose from the following two options: CXA16-AA or CXY08-AA. These two controller cards can be mixed in any combination up to a total of eight per DECserver 500. Note that the CXY08 and CXA16 controller cards INCLUDE the BC19N and H3104-B cabinet kits, respectively. The DECserver 500 hardware is available in the following packages and INCLUDES the DECserver 500 software license: U.S. Order No. Description List Price DSRVS-AA/AB DECserver 500 rack-mount system box $15,250 capable of supporting a combination of up to 8 communication controller cards: either the CXA16 (DEC423) or CXY08 (RS232). The DECserver 500 also includes processor, Ethernet controller, rack-mount brackets and software license. DSRVS-BA/BB Same as the above model except 15,950 has set of enclosures for use as stand-alone floor model. ***THE PURCHASE OF A MINIMUM OF TWO COMMUNICATION CONTROLLER CARDS WILL *** BE REQUIRED AT THE TIME OF THE INITIAL ORDER OF A DECserver 500 UNIT. THE ORDERING SYSTEM WILL REFUSE ALL ORDERS NOT SUBMITTED IN THIS MANNER. Communication Controller Cards (Available at FCS) U.S. Order No. Description List Price CXA16-AA/AF 16-line asynchronous communication $2,725 device supporting DEC423 signalling standard. Includes associated cabinet kit material (H3104-B). CXY08-AA/AF 8-line asynchronous communication 1,885 device supporting RS232-C signalling standard. Includes associated cabinet kit material (BC19N). NOTE: All cards ordered at the same time the DECserver 500 is ordered should be of the -AA variant. Order the -AF variant for future add-on field orders to add additional lines and/or line types. Power Cord Kits (Non U.S. Versions) NOTE: U.S. versions of the DECserver 500 hardware (DSRVS-Ax) INCLUDE the required power cord kit. Units intended for operation in other countries should be ordered WITH the proper power cord kit from the following list. The U.S. power cord kit is listed to retrofit DECserver 500 models ORIGINALLY ordered for non-U.S. operation only. U.S. Order No. Description List Price BN20A-2E US/Canada $12 BN20B-2E Japan 12 BN20C-2E New Zealand, Australia 12 BN20D-2E Belgium, Finland, Holland, Germany Sweden, Norway, France, Spain * BN20E-2E UK, Ireland * BN20F-2E Switzerland * BN20H-2E Denmark * BN20J-2E Italy * BN20K-2E India, So. Africa * BN20L-2E Israel * *Consult AQS Cable Information Use the following SHIELDED cables with each of the physical lines on the BC19N distribution cables included with each CXY08 option card (see DECserver 500 configuration diagram): These cables are not included with the option card and should be ordered on an as-needed basis. BC22D Null modem cable for local terminal or printer connections BC22E Full modem straight through cable for modem connections BC22F Full modem straight through cable for modem connections BC22R Recommended null modem cable for host systems and other devices, including those which use CTS/RTS flow control BC17D Null modem cable for host systems and other devices which do not use CTS/RTS flow control Use the following DEC423 cables and components with each of the MMJ (DEC423) lines on the H3104-B cable included with the CXA16 option card (see DECserver 500 configuration diagram): These cables are not included with the option card and should be ordered on an as-needed basis. BC16E Terminal data office cable for local terminal/printers H8571-A Passive office cable to 25-pin adapter (DEC423 to RS232 within 250 feet) H3105 Active adapter for office cables servicing RS232 devices greater than 250 feet away from DECserver 500 The DECserver hardware requires both a transceiver drop cable and Ethernet connection, H4000, DESTA, DELNI or CHIPCOM modem to connect to the Ethernet physical channel. Software The DECserver 500 software license is INCLUDED in the purchase price of the DECserver 500 hardware. The DECserver 500 cannot be down-line loaded using DECserver 100, DECserver 200 or Ethernet Terminal Server software. A DECserver 500 Load Host Kit (H-Kit) must be purchased separately for the DECserver 500 to operate; however, this load kit may be used for multiple DECserver 500s (1 kit per 12 servers recommended). The DECserver 500 Load Host Kit is eligible for inclusion in a System Startup Package. The following media and documentation kits are available for the DECserver 500: U.S. Order No. Description List Price QLZ46-H(M,3,5) VMS Load Host Kit $700 QRZ46-HM RSX Load Host Kit 700 QRZ46-HH RSX Load Host Kit 900 QYZ46-H(3,5) MicroRSX Load Host Kit 700 NOTE: The VAX/VMS software order number (QLZ46-xx) refers to the DECserver 500, where the software is licensed and runs, rather than the VAX/VMS Load Host, on which the software is initially installed. Also, VAX-11/725, MicroVAX I and VAXstation I processors are NOT supported as down-line load hosts. Additionally ULTRIX-32 and TOPS 10/20 operating systems are NOT supported as load hosts. PREREQUISITE SOFTWARE DECserver 500 software will operate with all Digital service nodes that support the LAT protocol. The DECserver 500 software license applies to the DECserver 500 on which the server software runs, not to service host node CPUs in the network. Supported Service Node Operating Systems: ULTRIX-32, V1.1 or later ULTRIX-32m, V1.1 or later VMS, V4.2 or later MicroVMS, V4.2 or later RSX-11M-PLUS with DECnet-11M-PLUS, V3.0 or later Micro/RSX V3.0 with DECnet-Micro/RSX V1.0 or later TOPS-20, V6.1 or later TOPS-10, V7.03 or later Digital Operating Systems Supported As Load Hosts: VMS with DECnet-VAX, V4.4 or later MicroVMS with DECnet-MicroVMS, V4.4 or later RSX-11M-PLUS, with DECnet-RSX, V3.0 or later Micro/RSX with DECnet-Micro/RSX, V3.0 or later Refer to the System Software Order Table (SSOT) for more information on the appropriate load host(s). AVAILABILITY The DECserver 500 is immediately available in limited quantities with volume availability in Q2/FY88. poly-TRM is a trademark of Polygon Associates, Inc. Personal Computer XT and Personal Computer AT are trademarks of International Business Machines Corporation ================================================================================ Note 40.25 PERIPHERAL DEVICE ANNOUNCEMENTS 25 of 56 EISNER::KILLEEN "Jeff Killeen" 56 lines 21-OCT-1987 16:52 -< LQP45 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! USING DECmate/WPS VERSION 2.3 WITH THE LQP45 - 21-October-1987 o The LQP45 Letter Quality Office Printer is the only available printer for those customers needing letter quality impact printing. It replaces the LQP02 and LQP03 printers. FEATURES These features apply to operation with the DECmate. Paper Handling Bidirectional forms tractor Dual bin cut sheet feeder with envelope feeder Printwheels Monospacing, proportional spacing multinational print- wheels Use the Reference Diskette V2.3 that is in the DECmate/ WPS V2.3 kit to print out the character sets on the printwheels. With each printwheel, print Document Number 40, Character Set Reference Document, on the Reference Diskette to show the characters according to the DECmate keyboard layout. Printer Cable BCC05-10 cable is included with the purchase of a DECmate III or DECmate III Plus. OPERATIONAL DIFFERENCES Basically, the DECmate/WPS customer, who has used an LQP02 or LQP03, will know how to use the software with the LQP45. There are, however, some differences. o Replacement Characters -- The LQP45 printer is a multinational printer and provides a wide range of characters. These characters can be accessed by using printer control commands within documents. For this reason, replacement characters are not implemented on the LQP45 printer. o Proportional Spaced Printing -- When using proportional spaced printing, always use a ragged right margin setting in the ruler. Proportional spacing with a justified right margin setting is not supported. o Cancelling a Document -- The message "Non Existent Printer" can appear on the Stop Printer Menu when attempting to queue a document to print immediately after cancelling a print operation. The LQP45 has a large print buffer and may still be printing what remains in the buffer as the next print operation begins. To avoid this, wait until the LQP45 has finished printing, before queuing another print operation. NOTE: Some operational discrepancies have been discovered with pre-released firmware in the LQP45. Testing has not been completed at this time. ================================================================================ Note 40.26 PERIPHERAL DEVICE ANNOUNCEMENTS 26 of 56 EISNER::KILLEEN "Jeff Killeen" 25 lines 21-OCT-1987 16:53 -< MODULES HANDLES 101 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! Training Classes for New Module Handle Design - 15-October-1987 Half-day training classes have been scheduled to show you how to design or specify module handles for the new packaging used in our MicroVAX 3500 and 3600 and future Q-bus products. Whether you design your own modules or incorporate a third-party's product, you will find the half-day session useful in learning how to adapt existing and new module designs. You'll see a thorough presentation on the new packaging and get some hands-on familiarity with the handle kit. An engineering staff member will answer your questions and refer you to specialists for more detailed considerations. The classes are Oct. 21 in Irvine, Oct. 27 in San Francisco, and Nov. 2 in Houston. Call your Digital sales representative to register. ================================================================================ Note 40.27 PERIPHERAL DEVICE ANNOUNCEMENTS 27 of 56 EISNER::KILLEEN "Jeff Killeen" 122 lines 28-JAN-1988 16:14 -< MEMORY PRICE INCREASES >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! PRICES INCREASE FOR CERTAIN VAX-11/7xx, - 28-January-1988 MicroVAX II AND PDP-11 MEMORY o List prices for certain MicroVAX II, PDP-11 and VAX-11/7xx series memory products are being increased 15% to offset rising suppliers costs o Memory Volume Products increase by 25% o Effective immediately As a result of recent, significant semiconductor Direct Random Access Memory (DRAM) cost increases worldwide, Digital will increase the list prices of certain memory products by 15% and MVDA (Memory Volume Products Section) prices by 25%. These higher prices for DRAMs are expected to continue for the next 12 months and have been precipitated by the increased value of the Yen and the effects of the U.S.-Japan Trade agreement. All memory manufacturers who source components from DRAM suppliers are affected by these increases. These changes are consistant with Digital's policy to price its memory options competitively, yet reflect changes in the cost to manufacture these products. System prices and service prices are not affected by these changes. HARDWARE SERVICE All Digital VAX memory products have a one-year warranty that includes on-site service. Additionally, during subsequent years, the Basic Monthly Charge (BMC) for VAX memory is zero ($0) when the memory is installed on a system covered by a Digital Field Service agreement. (This $0 BMC program does not apply to PDP-11 memory.) Also, customers installing multiple memory boards will pay only one installation charge per call plus $60 for each additional memory board installed. PRICING/ORDERING INFORMATION U.S. Model No. Description List Price Seg. BMC Instal. MS630-BB 4 MB 256K MOS MicroVAX II $ 1,618 XX N/C $325 Memory MS630-BF 4 MB MicroVAX 11 Expander 1,625 XX N/C 325 For BA2xx U.S. Model No. Description List Price Seg. BMC Instal. MS630-CA 8 MB 256K MOS MicroVAX II $ 3,183 XX N/C $325 Memory MS630-CF 8 MB MicroVAX II Expander 3,183 XX N/C 325 For BA2xx MS730-CA 1 MB ECC 11/730 1,065 XX N/C 455 MS750-CA 1 MB ECC 11/750 1,065 XX N/C 455 MS750-HB 8 MB + Controller Upgrade 9,245 X N/C 679 11/750 MS780-FA 2 MB ECC/64KMOS 11/780 1,993 XX N/C 500 MS780-JA 8 MB ECC/256KMOS 11/780 3,861 XX N/C 512 MS780-HC 11/780 256K Memory 8 MB + 13,805 XX N/C 610 Controller MS780-HD 11/780 256K Memory 8 MB + 13,805 XX N/C 610 Controller MS11-PB 1 MB ECC MOS Memory 1,565 XX 48 320 MSV11-QA 1 MB Q-bus MOS Parity 960 XX 29 335 MSV11-QB 2 MB Q-bus MOS Parity 1,200 XX 36 335 MSV11-QC 4 MB Q-bus MOS Parity 1,680 XX 72 350 MSV11-JD 1 MB PMI ECC 1,805 XX 16 290 MSV11-JE 2 MB PMI ECC 2,400 XX 32 328 MSV11-JF 2 MB 11/83 Memory BA2xx 2,400 XX 32 328 MSV11-JH 1 MB 11/83 Memory BA2xx 1,805 XX 16 290 SVA VOLUME PRICES - OEM AND END USER All new volume prices are 25% above current volume prices U.S. -------------ANNUAL MEGABYTE QUANTITY------------- Option No. List Price 10-24 25-49 50-99 100-199 200-499 500-999 1000+ MS730-CA $1,065 $1,050 $ 900 $ 893 $ 866 $ 854 $ 840 $ 828 MS750-CA 1,065 1,050 900 893 866 854 840 828 MS780-FA 1,993 1,890 1,701 1,606 1,559 1,536 1,513 1,489 MS780-JA 3,861 N/A 3,685 3,481 3,379 3,328 3,276 3,225 MS11-PB 1,565 N/A N/A 1,501 1,329 1,213 1,184 1,155 MSV11-QB 1,200 N/A N/A 1,160 1,026 938 915 893 U.S. -------------ANNUAL MEGABYTE QUANTITY------------- Option No. List Price 10-24 25-49 50-99 100-199 200-499 500-999 1000+ MSV11-QA $ 960 N/A N/A N/A $ 903 $ 854 $ 838 $ 820 MSV11-JD 1,805 N/A N/A N/A 1,733 1,638 1,606 1,575 MSV11-JE 2,400 N/A N/A $2,321 2,094 1,979 1,941 1,904 MS630-BB 1,618 N/A $1,608 1,470 1,379 1,286 1,240 1,195 MS630-CA 3,183 N/A 3,019 2,888 2,756 2,625 2,494 2,363 AVAILABILITY The memory options are available now at the new prices. ================================================================================ Note 40.28 PERIPHERAL DEVICE ANNOUNCEMENTS 28 of 56 EISNER::RICE "Gary Rice" 13 lines 29-JAN-1988 10:19 -------------------------------------------------------------------------------- > As a result of recent, significant semiconductor Direct Random Access > Memory (DRAM) cost increases worldwide, Digital will increase the list > prices of certain memory products by 15% and MVDA (Memory Volume Products > Section) prices by 25%. These higher prices for DRAMs are expected to > continue for the next 12 months and have been precipitated by the > increased value of the Yen and the effects of the U.S.-Japan Trade > agreement. YES, Fujitsu announced a price increase from $10.50 to $12.00 for their 1 megabit chips. Definitely 25% . . . Gary ================================================================================ Note 40.29 PERIPHERAL DEVICE ANNOUNCEMENTS 29 of 56 EISNER::KENNEDY "Terry Kennedy" 9 lines 29-JAN-1988 20:40 -< Such a deal... >- -------------------------------------------------------------------------------- They (DEC) really didn't want to get into price wars with the discount vendors anyway... If you were really buying 500 boards a pop (and getting the great 8% volume discount) you still couldn't beat Brand X's quantity one price. As an aside, why can't the company that gets 64 Mb on a card give us a 4 Mb PMI memory? This is available from the Brand X companies, after all. Is it just possible that DEC is *not* the technology leader in memory products? [grin] ================================================================================ Note 40.30 PERIPHERAL DEVICE ANNOUNCEMENTS 30 of 56 EISNER::ERSKINE "Chris Erskine" 5 lines 1-FEB-1988 06:56 -< Maybe at one time >- -------------------------------------------------------------------------------- > companies, after all. Is it just possible that DEC is *not* > the technology leader in memory products? [grin] Is DEC a leader at anything anymore. (Except trying to take your money) ================================================================================ Note 40.31 PERIPHERAL DEVICE ANNOUNCEMENTS 31 of 56 EISNER::HASSINGER "Bob Hassinger" 7 lines 1-FEB-1988 11:01 -< Certainly DEC is still the leader in certain areas... >- -------------------------------------------------------------------------------- > Is DEC a leader at anything anymore. Certainly. I think they are still the leading supplier and developer of VMS operating system software and they have little competition in the production of computers that execute the VAX instruction set ;-). Bob H ================================================================================ Note 40.32 PERIPHERAL DEVICE ANNOUNCEMENTS 32 of 56 EISNER::DEC_NEWS_1 "News about DEC from DEC" 102 lines 28-APR-1988 20:41 -< SUPER HSC - ABLE TO LEAP TALL I/O REQUESTS..... >- -------------------------------------------------------------------------------- VERSION 3.70 OF HSC SOFTWARE NOW AVAILABLE - 28-April-1988 o Provides an HSC performance improvement of up to 64% o Provides faster disk drive failover o Adds support for VAXsimPLUS o Corrects several known problems PRODUCT DESCRIPTION Version 3.70 of the HSC Software contains significant improvements to Version 3.50, particularly in the areas of I/O performance and system availability, as well as correcting a number of known problems. BENEFITS Version 3.70 can provide a significant system performance increase when the application's current limitation is the number of requests per second that can be handled by the HSC subsystem. Bug fixes, support for VAXsimPLUS and the reduction in disk failover time, all contribute to increased data and system availability. V3.70 ENHANCEMENTS o Performance improvements--An HSC70 can now handle up to 1150 I/O requests per second, an increase of 64% when compared to the 700 requests per second achievable with V3.50. Similarly, an HSC50 can handle up to 550 I/O requests per second with V3.70 software, an increase of 46% when compared to the 375 requests per second with V3.50. The performance penalty frequently experienced when shadowing with a high proportion of WRITE operations has been offset. The performance of an HSC with full shadowing enabled is now typically comparable to that of the same HSC running V3.50 software without shadowing. o Reduction in disk failover time -- Failover of a dual-ported disk to another controller may be automatically initiated as a result of a failure in either a drive port or a controller channel port. The time between the disk's failure to respond to a controller request and that disk again becoming available to the cluster has, in the past, been typically around 63 seconds. Prior to V3.70, by far the greatest contribution to that time (about 48 seconds) was from the HSC itself. That 48 seconds has now been reduced to about four seconds, which reduces the total failover time to about 18 seconds, and provides higher availability of the disk data and of the whole cluster. o Support for VAXsimPLUS--An extension to VAXsim (VAX System Integrity Monitor) allows prediction of failures on Digital's RA-series Winchester disk drives. VAXsimPLUS analyzes error-log information from the VMS operating system, predicts what is wrong with the drive and notifies the customer to call Field Service. When used in conjunction with the autocopy feature of VMS Shadowing, a suspect drive can be automatically copied to a spare drive which will be used to keep the customer's system running until the failing drive can be repaired. o Bug fixes--Most significant among the various known problems fixed in servers and utility programs was the correction of two problems that caused unnecessary reboots of an HSC and impacted the availability of all drives on that subsystem. One problem did occur when an HSC was heavily involved in error recovery related to a particular disk drive. If the error recovery took too long, all the evidence would indicate to the host that the HSC had failed even though it had not. The host would then initiate a reboot of the HSC. Since the drive with the problem failed over to another HSC, the scenario could then be repeated. In V3.70, failed drives are marked off-line if the error recovery does not appear to be succeeding in a timely manner, thereby eliminating unnecessary reboots. Another internal resource problem which produced similar symptons in very heavily loaded clusters has also been corrected in V3.50. A reboot of the whole cluster was sometimes needed to recover from this problem, impacting overall system availability. The correction of these two problems provides a significant increase in cluster availability. PREREQUISITES The HSC software will only run on HSC50 or HSC70 hardware. PRICING/ORDERING INFORMATION Orders for HSC hardware include HSC software license and H-kit, thus they do not need to be ordered separately. In addition, existing HSC customers who have Digital maintenance contracts will automatically receive the update at no charge as part of their integrated service warranty or agreement. For self-maintenance customers, the software H-kits may be ordered separately as follows: Order No. Description U.S. List Price QX926-37 HSC70 SOFTWARE SMS RX33 $70/month QX926-3G HSC50 SOFTWARE SMS TU58 70/month QX926-H7 HSC70 SOFTWARE UPD RX33 605 QX926-HG HSC50 SOFTWARE UPD TU58 595 QX926-HZ HSC SOFTWARE RTC UPD 315 QX926-I7 HSC70 SOFTWARE ISS RX33 380 QX926-IG HSC50 SOFTWARE ISS TU58 380 AVAILABILITY Version 3.70 of the HSC Software is available now. ================================================================================ Note 40.33 PERIPHERAL DEVICE ANNOUNCEMENTS 33 of 56 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 6 lines 30-APR-1988 11:15 -< Several <> all >- -------------------------------------------------------------------------------- >> VERSION 3.70 OF HSC SOFTWARE NOW AVAILABLE - 28-April-1988 >> o Corrects several known problems But not all? :-} Alan ================================================================================ Note 40.34 PERIPHERAL DEVICE ANNOUNCEMENTS 34 of 56 EISNER::DEC_NEWS_1 "News about DEC from DEC" 66 lines 31-JUL-1988 23:10 -< RA81 enters "end-of-life planning phase" >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! END-OF-LIFE OF THE RA81 DISK DRIVE - 14-July-1988 ******************************************************************************* o The RA81 disk drive has entered the end-of-life planning phase o Last order date for the RA81 is December 31, 1988 o Customers should plan to migrate from the RA81 to SA482/ RA82 o Field end-of-life inputs are required by November 1, 1988 Storage Systems has begun the product phase down (PPD) process for the RA81 disk drive. The date for removal from the price book is December 31, 1988. The recent price reduction on the RA82 has made that product significantly more attractive, now being priced below the RA81, but with much greater capacity, performance and reliability, and lower maintenance costs. NOTE: Minimum VMS for the SA482/RA82 is V4.5. Shadow sets may be done only with identical disks, e.g., an RA81 with an RA81, etc. PRODUCT COMPARISON When migrating your customer to the RA82, there are four specifics you should use to compare the RA82 with the RA81. The RA82: RA81 RA82 1. Is 10-15% faster Avg. Seek Time (ms) 28 24 Avg. Access Time (ms) 36 32 2. Has 36% more capacity Capacity (MB,F) 456 622 3. Has lower monthly maintenance cost BMC $95 $59 4. Costs less on box price as well USLP $17,640 $17,000 as price/MB Price/MB $38.68 $27.33 RESOURCE INFORMATION Installed Systems Marketing is offering the "DSA Migration Plus" program to migrate your customer from the RA80, RA81 and RA60 to the newer SA482/RA82 disk technology. For more information on this program contact your Digital Representative. AVAILABILITY After Last Order Date: Storage Systems has committed to support the Field Service requirements for RA81 head disk assembly and module spares through the end-of-service life of that product. A limited number of refurbished RA81 units will continue to be available from the Traditional Product Line (TPL) Group for an undetermined timeframe. Final RA81 build will be completed in FY89 and all units inventoried against customers' RA81 end-of-life requirements. ================================================================================ Note 40.35 PERIPHERAL DEVICE ANNOUNCEMENTS 35 of 56 EISNER::DEC_NEWS_1 "News about DEC from DEC" 79 lines 31-JUL-1988 23:23 -< THE LP29-QA, 2000 LPM PRINTER >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! THE LP29-QA, 2000 LPM PRINTER - 22-April-1988 ******************************************************************************* FOR MicroVAX II SYSTEMS o New Q-bus model in the LP29 Printer series o Suited for high volume print applications o Used with MicroVAX II processors o Available in May, 1988 The LP29-QA is the latest LP29 printer family member that can be connected to Digital's MicroVAX II processors. Capable of printing up to 2000 lpm, the LP29-QA is a companion product to the LP29-SA variation (MicroVAX 3500/3600 systems) completing the LP29 Q-bus offerings. The LP29-QA connects to the MicroVAX II system via the LPV11 controller. It is through the MicroVAX II system that the LP29 can be accessed to DECnet. The primary benefit of access to DECnet through the MicroVAX II is increased availability to more users on the network. Both the LP29-QA and LP29-SA are Digital unique products for Q-bus systems. The LP29 series of printers are the only printers on the market today that will operate at 2000 lpm with Digital's systems. Digital unique features include: optimized 64 character band, and double buffer interface. PRODUCT DESCRIPTION The LP29-QA is the MicroVAX II, Q-bus variation of the fastest and most reliable impact line printer offered by Digital. It prints at speeds up to 2000 lpm. o MicroVAX II Compatability. The LP29-QA is the only 2000 lpm printer available with the MicroVAX II. o DECnet Compatibility. Through its MicroVAX II connection users are able to access the printer via the network. CONFIGURATION INFORMATION The LP29-QA is qualified on the MicroVAX II. SOFTWARE SUPPORT The LP29 is supported on VMS. For details on specifications and LP29 features, refer to the LP29 December 14, 1987, announcement. PRICING/ORDERING INFORMATION U.S. Model No. Description Seg. List Price Instal. BMC LP29-QA 2000 lpm Bandprt with 090 $38,500 $625 $394 LPV11 Controller, 120 V AVAILABILITY LP29-QA is available in May 1988. SUPPORT SERVICES The LP29 is offered with a one-year warranty. It is supported by a full line of onsite service offerings, including Basic and DECService. Basic monthly charge is $394. Installation is included when the LP29 is ordered as part of a system order; when purchased separately, there is an installation charge of $625. ================================================================================ Note 40.36 PERIPHERAL DEVICE ANNOUNCEMENTS 36 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 112 lines 1-AUG-1988 16:32 -< Bar codes for LN03 and LA75 printers >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! BAR CODE CAPABILITIES FOR LN03 AND LA75 PRINTERS - 28-April-1988 ******************************************************************************* o Barcode font cartridges for LN03 Laser Printers and the LA75 Companion Printer o Available for shipment in Q1/FY89 Two new font cartridges have been developed to provide fast, easy and inexpensive generation of high quality bar code printing. Customers are finding new uses daily for accurately identifying, counting, tracing and tracking items. They are finding bar coding an effective way to improve service and cut expenses. PRODUCT DESCRIPTION The industry standard "3 of 9" produces machine readable output in which each character is represented by a code printed from bar and space elements. These cartridges contain characters defined by Uniform Symbology Specification USS-39 and meet Mil Spec 1189A for Bar Code 3 of 9. LN03 Bar Code 3 of 9 Font Cartridge (LN03X-DE) The LN03X-DE font cartridge contains two bar code heights. The 18 point height contains bars only; the 36 point height contains bars with characters below. These two heights can be combined to create a variety of bar heights. +---------------------------------------+ | Insert LN03X-DE samples | | | | 18 point height 36 point height | | (bars only) (bars and characters) | +---------------------------------------+ LA75 Bar Code 3 of 9 Font Cartridge (LA75X-AC) The LA75X-AC cartridge contains only a single bar code height and contains bars only. Multiple passes can be performed to create bars of varied height. +---------------------------------------+ | Insert LA75X-AC sample | | | | one height - bars only | +---------------------------------------+ Both fonts can be used by DECmate WPS, WPS-PLUS and in VMS applications. Modifications to the WPS-PLUS Printer Table Utility (PTU) are required for all LN03 font cartridges. Like all Digital Font Cartridges, a user manual is included with the Bar Code 3 of 9 Cartridges. These cartridges do not support the Multinational Character set (MCS), the National Replacement Character set (NRC) or lower-case alphabetic characters. FEATURES The Bar Code 3 of 9 Cartridges contain the following characters: o Uppercase alphabet A-Z o Numerals 0-9 o Punctuation marks: period (.), hyphen/dash/minus (-), dollar sign ($), plus sign (+), slash (/), space, per cent sign (%). o Asterisk (*) is used as the start/stop character. Bar coding, which has proven to be one of the most accurate means of data entry, is being used in libraries and throughout health care, retailing and manufacturing industries such as automotive, primary metals, paper, aerospace and government agencies. Industry sources state this technology is expanding worldwide. Many of our customers' bar code printing requirements can be satisfied with our LA75 dot matrix and LN03 laser printers, especially for label applications. (Gummed label stock in various sizes is commercially available for both printers.) The Bar Code 3 of 9 Font Cartridges permit bar codes to be generated within the printer upon receiving the appropriate ASCII character command, thus eliminating the need for special software. The availability of bar code printing adds another feature and benefit to the LN03 and LA75 printer products and opens the door to new sales opportunities. Bar coding is an effective way to improve service and cut expenses. Mention this feature to your customers' purchasing department. Further information about other bar code symbologies and some recently announced bar code software products can be found in the November 1987 issue of DESKTOPics. PRICING/ORDERING INFORMATION These font cartridges can be ordered from DECdirect (1-800-DIGITAL) or the Electronic Store. Order No. Description U.S. List Price LN03X-DE LN03 Bar Code 3 of 9 Font Cartridge $299 LA75X-AC LA75 Bar Code 3 of 9 Font Cartridge 53 AVAILABILITY These cartridges will be available for customer shipment in Q1/FY89. ================================================================================ Note 40.37 PERIPHERAL DEVICE ANNOUNCEMENTS 37 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 147 lines 1-AUG-1988 16:33 -< CISCE-AA 24 port CI start coupler >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! CISCE-AA 24-PORT CI-STAR-COUPLER - -5-July-1988 ******************************************************************************* EXPANDER o CISCE increases the maximum number of VAXcluster nodes that can be connected to the CI-bus from the previous limit of 16 up to 24 o U.S. area only PRODUCT DESCRIPTION The CISCE Option is used to interconnect up to three SC008 Star-Coupler options (-ACs, -AD), which are not included, to form a 24-Port CI- Star Coupler. To avoid introducing a single point of failure, the CISCE is comprised of two active amplifier/coupler boxes, one each for the "A" and "B" paths of the CI- (Computer Interconnect) bus. The CISCE provides room to grow at the top for installed, large (12 nodes or more) CI-VAXcluster system customers who simply need to add more nodes to their CI-VAXcluster systems without increasing their available I/O bandwidth. The ability to add either CPU or mass-storage capacity serves to: o Allow larger CI-VAXcluster systems in most application environments, o Enhance customer satisfaction with large CI-VAXcluster systems, o Provide an opportunity to sell additional hardware, software, and services. The actual number of CI-VAXcluster system nodes allowed in a specific configuration depends on the application mix running during peak processing periods and on the actual utilization of CI-ports and the CI-bus itself. Digital will assist customers in evaluating their system needs through the "Large VAXcluster System Expansion Planning Process." PREREQUISITES CI-Bus Adapters CPUs: - CIBCA - (CI750, or CI780, or CIBCI) in combination with a CINLE-AA/AB Enhancement Kit(s). (This is determined by a Field Service site audit.) Prerequisite Software: HSCXX: - V370 Software - No charge CINLE-BA/BB Enhancement Kit(s), or equivalent FCOs. (This is determined by the Field Service site audit.) VMS: - V4.7 or later Large VAXcluster System Expansion Planning Process--Assures customers that their planned expansion will deliver the intended benefits. The process measures I/O activity on the CI-bus of installed VAXcluster systems during peak processing periods, and projects what the I/O loading will look like after the expansion. This insures better, continued satisfaction with the VAXcluster system. The process involves the following steps: o The customer together with the account team must complete the Large VAXcluster System Expansion Planning Questionnaire, and provide basic information about the present configuration and planned expansion. o Clusters Sale Support runs a VAXcluster Evaluation Package on the customer's cluster during periods of peak processing activity to measure the I/O load on CI resources and identify potential problems. The results of this measurement are evaluated to determine the maximum, reasonable expansion of the VAXcluster. o Field Service conducts a site audit to determine what upgrades are needed to CI adapters in VAX processors and HSC disk controllers. o Area Clusters Sale Support conducts a technical review and generates a recommended course of action. o Area Installed Base Sales Management conducts a business review, and decides whether or not the order can be certified. The greater than 16 node capability should only be offered to customers with installed CI-VAXcluster systems with 12 nodes or more (VAX plus HSC), and have stated a concern for being able to grow beyond 16 CI-Nodes. This includes: o Rapidly expanding CI-VAXcluster system customers who require confidence that there is room to grow at the top. o Rapidly expanding CI-VAXcluster system customers who are faced with having to subdivide into multiple VAXcluster systems, losing the major benefit of clustering, and are considering a competitive mainframe alternative. o Customers faced with having to consolidate multiple VAXcluster systems to reduce the operational costs associated with multiple CI-VAXcluster systems, i.e., multiple operators, application and layered software licensing fees, etc. NOTE: CISCE can only be sold to installed CI-VAXcluster systems with 12 nodes or more. It will not be sold as a component of a new VAXcluster system. o Increased CI-Star Coupler connectivity to allow additional CPU or mass-storage capacity to be added to the CI-VAXcluster system. o An understanding of how their current application mix utilizes the CI resources during peak processing periods. o An understanding of what hardware and application software will be added. o An understanding of the potential impact of adding more nodes on time-critical applications, if any. NOTE: The CISCE should not be proposed in situations where customers require: - Increased available total CI I/O bandwidth. - Reduced latencies or improved response times. DOCUMENTATION CISCE-AA Installation and User's Guide EK-CISCE-UG CINLE-AA Installation Guide EK-CINLU-IN CINLE-BA Installation Procedures EK-CINLE-IN AVAILABILITY CISCE-AA and CINLE-xx are available now. PRICING/ORDERING INFORMATION Hardware Model No. Description CISCE-AA 24-PORT CI-STAR-COUPLER EXPANDER CINLE-AA/AB CPU CI-NODE LINK EHANCEMENT KIT CINLE-BA/BB HSCXX CI-NODE LINK ENHANCEMENT KIT ================================================================================ Note 40.38 PERIPHERAL DEVICE ANNOUNCEMENTS 38 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 166 lines 1-AUG-1988 16:35 -< DECconnect RWE - Remote Wall Enclosure >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! DECconnect ENHANCEMENT - 21-July-1988 ******************************************************************************* REMOTE WALL ENCLOSURE o Simplify and reduce the cost of local workgroup network connectivity PRODUCT DESCRIPTION The DECconnect Remote Wall Enclosure (RWE), H3131-A, is a wall-mounted metal enclosure which houses passive termination equipment to support a radial wiring topology for either 16-high or 16-low speed data lines running over unshielded twisted pair (UTP) cabling. The RWE provides small local workgroups with a simple, easy to implement network connectivity solution for low-speed terminal via terminal servers (DECserver 200/500) and high- speed personal computer/Workstation (WS) via Ethernet multiport repeaters (DEMPR) to an enterprise-wide Ethernet backbone. LOW-SPEED TERMINAL APPLICATIONS Support of low-speed, data-only, (no modem control) terminal network connectivity is accomplished by mounting a 16-user termination patch panel kit, H3107-C, into the RWE. Unshielded twisted pair office cabling terminates at Modified Modular Jack (MMJ) connectors in the H3107-C and radiate out to DECconnect faceplates or directly to office equipment. A pair of 36 conductor cables lead from the RWE mounted H3107-C to local or remotely located (up to 300 feet) DECserver 200/DL terminal servers or to DECserver 500 CXA16 line cards. Terminal servers (DECserver 200/DL) can be wall mounted in close proximity to the RWE by using a H039 wall mount kit (1 per DECserver 200). The BC16C-10 cables, which come with the DECserver200/DL, can be used to connect the RWE mounted termination patch panel to the terminal server. If the terminal servers are remotely located, then the customer can either use additional BC16D 36 conductor extension cables or field terminate bulk 36 conductor cable using the H8101 field termination tool kit and H8227 field termination connector kit. The 36 (18 twisted pairs - twp) conductor cable can be run up to 300 feet to a DECserver 200/DL or to a CXA16 line card in a DECserver 500. CONFIGURATION GUIDELINES FOR LOW-SPEED For low-speed communications o The maximum length for a 36 conductor cable run between a RWE and a terminal server is 91.5 meters (300 feet.) High-Speed Twisted-Pair Ethernet Applications Support of high-speed (10 Mbps) Ethernet communications is accomplished by mounting an Unshielded Twisted-Pair Card Cage Kit, H3131-B, into the RWE (H3131-A). Up to two Unshielded Twisted-Pair Adapter kits, H3330-AA, can then be mounted into the H3131-B. Each H3330-AA kit contains four adapter boards. Each adapter board supports two unshielded twisted-pair Ethernet connections that run to DECconnect faceplates or office equipment. Two H3330-AA support up to 16 radially distributed unshielded twisted-pair communications lines. ThinWire Ethernet cables run from the back of the H3330-AA to DEMPRs which are wall-mounted next to the RWE using the H039 wall-mount kit. SEE HARDCOPY FOR ART CONFIGURATION GUIDELINES FOR HIGH SPEED For high-speed communications o The maximum length for twisted-pair cable running Ethernet communications is 70 meters (230 feet). o Twisted-pair cables terminate at Twisted-Pair Ethernet Adapters, which, in turn, use a short section of ThinWire to connect to their respective devices. The total length of ThinWire cable (the cable leading from the RWE to the DEMPR plus the cable leading from the office adapter to the end-user equipment) must not exceed 10 meters (33 feet). FEATURES/BENEFITS Feature Benefit o Compact wall-mounted enclosure o Uses no floor space o Aesthetically acceptable for office applications o Local cable termination o Simplifies moves, adds and changes Small departmental MIS, network system, telecommunication and facilities managers. Many are looking for a simple, low cost means of connecting low density, terminal or personal computer/workstation users to an Ethernet backbone. Customers need a low-cost, easy-to-install, aesthetically acceptable means of locating network hardware and cable termination components. These components offer access to, and management by, a local MIS/network resource. The H3131-A (RWE) and H3107-C termination panel are for low-speed video terminal to Ethernet applications. When the terminal servers are placed in close proximity to the RWE, use the H039 wall-mounting bracket (one per DECserver). The BC16D-10 cable that comes with each DECserver is used to connect the server to the H3107-C panel. This configuration will support up to a maximum of 16 lines. When terminal servers are remotely located (>10 feet) from the RWE, sell the H3131-A, H3107-C and an appropriate length of H8247 or H8248 36 conductor (18 twp) bulk cable. High-speed Twisted-pair Ethernet Applications. Because unshielded twisted-pair (UTP) adapter products are only available in kits of eight, customers have a choice of eight or 16 lines. For eight (8) line configurations For sixteen (16) line configurations sell the: sell the: (1) H3131-A RWE, (1) H3131-A RWE (1) H3131-B card cage kit, (1) H3131-B card cage kit (1) H039 wall-mount bracket (2) H039 wall-mount brackets (1) H3118-AC (kit of 16 8' (1) H3118-AC ThinWire cables)* (2) H3310-AA (1) H3310-AA kit of eight UTP (2) H3330-AA office adapters (2) DEMPR (1) H3330-AA kit of four dual-ported, rack-mount UTP adapters (1) DEMPR - multiport repeater *Less costly than individual BC16M cables The RWE product is one means of illustrating how simply and cost effectively Ethernet network connectivity can be provided to local nonenterprise-wide networked terminals, personal computers and workstation users. This product offers customers one of the lowest cost, most manageable and simple-to-implement Ethernet network connectivity solutions available--which supports Digital's distributed processing strategy using Ethernet. PRICING/ORDERING INFORMATION U.S. Model No. Description List Price H3131-A Remote Wall Enclosure $ 175 H3131-B Card Cage Kit 40 H8227-A Connector Kit, 36 Conductor Cable 35 H8101 Tool Kit, Termination 36 cond cable 150 H8247-A Spool, 36 Cond. Cable, PVC, 250 feet 325 H8247-B Spool, 36 Cond. Cable, PVC, 1,000 feet 1,060 H8248-A Spool, 36 Cond. Cable, FEP, 250 feet 980 H8248-B Spool, 36 Cond. Cable, FEP, 1,000 feet 3,700 H3118-AC (16) ThinWire Patch Cables 8 foot long 145 H3310-AA TWP Ethernet Office Adapter Kit (8) 392 H3330-AA TWP Ethernet SER/OCC Adapter Kit (8) 600 H3107-C Low-speed 16 MMJ Termination Panel 58 H039 Wall-mount Kit, DECserver, DEMPR 47 AVAILABILITY All products are available for shipment from stock for July FY89. ================================================================================ Note 40.39 PERIPHERAL DEVICE ANNOUNCEMENTS 39 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 29 lines 1-AUG-1988 16:37 -< Bye bye to (DF112/124)-AA/AM modems >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! END-OF-LIFE MODEMS REMOVED FROM PRICE FILE - 06-June-1988 ******************************************************************************* o DF112-AA/AM, DF124-AA/AM placed on Maintenance-Only status o Encourage customers to migrate to new Scholar Plus modems - more functionality, newer technology at little or no price increase Three new modems, the Scholar Plus series, have been announced and are recommended as upgrades for users of the above-mentioned retired modems. The Scholar Plus modems feature access and call-in security, MNP/X.PC error correction and Hayes "AT" compatibility. Customers will benefit from the added functionality and newer technology of the Scholar Plus modems. The following new Scholar Plus modems are available on FastShip from DECdirect as well as through the regular ordering channels. o DF212-CA: 1200 bps Compact Desktop Modem o DF242-CA: 2400 bps Compact Desktop Modem o DF124-CA: 2400 bps Leased-line/Dial-up Desktop Modem o DF124-CM: 2400 bps Leased-line/Dial-up Rackmount Module ================================================================================ Note 40.40 PERIPHERAL DEVICE ANNOUNCEMENTS 40 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 232 lines 1-AUG-1988 16:39 -< DHB32 now available in 423 version >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! DEC423 VERSION AVAILABLE FOR DHB32 - 18-July-1988 ******************************************************************************* VAXBI COMMUNICATIONS CONTROLLER o Enables up to 16 terminals or printers to directly connect to a VAXBI system using DECconnect o Volume Availability -- Now The DEC423 version of the DHB32 VAXBI Communications Controller is now available for providing DECconnect wiring for asynchronous devices. PRODUCT DESCRIPTION The DHB32 is a host-based DMA communications controller that multiplexes up to 16 asynchronous devices to communicate with Digital's VAXBI processors. It provides a direct connect solution, offering efficiency and data security. The DHB32 consists of the following hardware: o DHB32 Module: This is a single VAXBI-sized board which plugs directly into one VAXBI backplane slot. o Cabinet Kit which includes a distribution panel, a loopback connector for diagnostic tests, six adapter cables to connect the module to the distribution panel and a user guide. The distribution panel is the interface for up to 16 asynchronous devices and is available in two versions, depending on which cabinet kit option is ordered: o Version 1: 16 EIA232 Ports--Two DHU11-type panels, each containing eight 25-pin connectors that support full modem control. This enables up to 16 asynchronous terminals, modems, and serial printers to connect through these ports. o Version 2: 16 DEC423 Ports (2 selectable as EIA232)--A single Octal panel contains two 36-pin Type III connectors to link up to 16 terminals plus two 25-pin connectors for alternate EIA232 connections on Ports 14 and 15. Switches on the panels select EIA232 or DEC423 on these ports. Each 36-pin connector supports eight users and allows access using DECconnect wiring strategy from a cable concentrator in the office environment via eight MMJ (Modified Modular Jacks) connectors. The DHB32 is intended for any customer who requires direct terminal connection to a VAXBI system. Primary markets include laboratory, education and medical organizations where high performance and fast response time is desired. Other markets include government, defense and finance organizations where a secured environment with controlled access is necessary. While most terminal connections are currently being configured in networks using DECservers, the following cases show where a direct connect solution is needed: o Where the customer perceives that the desired level of security can only be met with a direct connect approach. o Where the customer perceives that a direct connect solution is needed to provide the desired level of response time, for example, realtime process control. o Where a CPU upgrade is required for an installation where communications wiring is already in place for a host-based approach. The DHB32 offers an alternative for customers who need a direct connect solution. It provides a single vendor solution for system needs and gives a choice of RS232 or DEC423 connections in this environment. Where Are Your Resources? Information Sheet: YM-AA120-00 User's Guide: EK-DHB32-UG FEATURES/BENEFITS Feature Benefit o Direct VAXBI connection - Ease of installation and operation - Data security, controlled (physical) access o Simple configuration - System connection cost savings Feature Benefit o High performance - Increased user efficiency o Use of DMB32 driver - Software compatibility SOFTWARE DESCRIPTION The DHB32 emulates the asynchronous portion of the DMB32 using similar architecture and the same VMS driver. The functionality offered by the eight asynchronous ports on the DMB32 is extended to 16 ports on the DHB32. Unlike the DMB32, there is no synchronous or parallel printer support on the DHB32. PERFORMANCE The DHB32 offers guaranteed 19,200 bps throughput on Ports 0 and 1 and 9,600 bps throughput on each of the remaining 14 ports simultaneously. The total aggregate throughput is 17,500 characters per second. Maximum port speeds of 38,400 bps are possible; however, the present VMS operating system (V4.7 or better) supports up to 19,200 bps only. Customers can develop a custom driver in order to achieve higher throughput speeds. SYSTEM CONFIGURING RULES The following is the maximum number of DHB32s which can be supported per system. Note that panel unit availability on each system may limit the actual number of DHB32s which can be configured. o VAX 82xx/83xx Config. 1 Maximum two DHB32s, OR one DHB32 and one DMB32 per system o VAX 82xx/83xx Config. 2 Maximum four DHB32s OR six DMB32/DHB32s per system o VAX 8530/8550 Maximum two DMB32/DHB32s per VAXBI channel and four DMB32/DHB32s per system o VAX 8700/8800 Maximum two DMB32/DHB32s per VAXBI channel and eight DMB32/DHB32s per system o VAX 62xx Maximum four DMB32s per VAXBI channel and eight DMB32/DHB32s per system but for internal VAXBI channels, a total of two DMB32s or two DHB32s (DEC423) or one of each allowed; or one DMB32 and one DHB32 (EIA232) allowed The following is the configuring information/power requirements for the DHB32: Mounting |----- dc Amps drawn @ ------| VAXBI Panel Option Requirements 5V 12V -12V -5.2V -2V Nodes Units DHB32 EIA232 1 VAXBI slot 5.56 0.42 0.42 0.00 0.00 1 8 DHB32 DEC423 1 VAXBI slot 5.56 0.42 0.42 0.00 0.00 1 4 SPECIFICATIONS Signal Compatability o For EIA232 Distribution Panel: - Electrically and mechanically compliant with EIA-232-D (replaces RS232-C) and compatible with CCITT V.28/V.24 standards. Two panels are provided, each with eight ports. o For DEC423 Distribution Panel: - All 16 ports conform to DECconnect DEC423 (Type III connectors). Ports 14 and 15 are switchable between DECconnect DEC423 and EIA 232-D. PRICING/ORDERING INFORMATION Hardware: U.S. Discounts Order No. Description List Price E/U OEM DHB32-M DHB32 Module $6,200 Q Q EIA232 Cabinet Kits: CK-DHB32-AM For VAX 82xx/83xx 1,000 S S CK-DHB32-AJ For VAX 85xx/87xx/88xx internal 1,000 Q Q VAXBI channels; and VAX 62xx internal and external channels CK-DHB32-AN For VAX 85xx/87xx/88xx external 1,000 Q Q VAXBI channels DEC423 Cabinet Kits: CK-DHB32-LM For VAX 82xx/83xx 600 S S CK-DHB32-LJ For VAX 85xx/87xx/88xx internal 600 Q Q VAXBI channels; and VAX 62xx internal and external channels. CK-DHB32-LN For VAX 85xx/87xx/88xx external 600 Q Q VAXBI channels NOTE: The DHB32-M module is used with each of the cabinet kits listed above. Software: The DHB32 uses the asynchronous portion of the DMB32 device driver which is available with VMS V4.6 and later. ULTRIX-32 support for the DHB32 will be announced at a later date. VOLUME AVAILABILITY Both versions of the DHB32 are available now. SUPPORT SERVICES Digital Field Service offers a wide variety of services to ensure proper maintenance of the DHB32. o The DHB32 comes with a one-year on-site warranty (DECservice). o Contract Service Offerings: DECcall Add-on Option BMC RET/EVENT DSMC CMC Instal. DHB32-M $40 N/A $48 N/A $580 (NOTE: No incremental cost for cabinet kits.) o Non-Contract Service Offerings: Part No. DECmailer T1044 (module) $1,240 (NOTE: Individual cabinet kits are not field replaceable.) ================================================================================ Note 40.41 PERIPHERAL DEVICE ANNOUNCEMENTS 41 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 86 lines 1-AUG-1988 16:41 -< Reduced prices on RA82,RD32,RD54,RX33 disks >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! 5-1/4" Winchester Reduced by 38% - 11-May-1988 ******************************************************************************* - Multi-Drive Family of 14" Winchesters Reduced by As Much as 28% DIGITAL ANNOUNCES SIGNIFICANT DISK DRIVE PRICE REDUCTIONS Maynard, Mass. -- April 28, 1988 -- Digital Equipment Corporation today announced repricing for a series of disk drives that deliver customer savings of from 19 percent to 51 percent over the company's previous list prices. Aimed at sustaining Digital's competitive price/performance for highly-integrated storage products, today's new pricing involves the Company's 5-1/4 inch RD54 and RD32 family of hard disk drives, the RX33 flexible diskette drive and SA842/RA82 family; Digital's families of 14 inch hard disk drives. While the new pricing schedules establish a set of competitive prices, all of the repriced drives carry Digital integration features beyond those typically available from so-called plug-compatible option products from third party suppliers. All of the repriced drives, for example, feature 100 percent compatibility with Digital's DSA architecture, which is the company's highly integrated, long-term approach to providing achitectural and technological storage compatibilities and investment protection for all customers. Today's announcements include a 38 percent reduction -- to $4,500 -- for Digital's 5-1/4 inch RD54 hard drive featuring 159 megabytes of formatted storage capacity. Similarly, the Company's 5-1/4 inch RD32 hard drive, with a formatted capacity of 42 megabytes, has been reduced by 33 percent to $1,615. Digital's RX33 floppy disk drive, featuring a formatted capacity of 1.2 megabytes, has been reduced by 51 percent to $338. Digital's SA482/RA82 product family have been reduced by 19 percent to 28 percent depending on the configuration. These products not only feature a one-year warranty, but also, because of their long-term reliability, are offered with the industry's lowest monthly maintenance service charge for large disks of only $59. The Company's four-spindle SA482-AA/DD drive, which offers a formatted capacity of 2.488 gigabytes, is now available for $66,000. The three-drive, three-spindle SA482-LA/LDD, now available for $51,000, offers 1.8666 gigabytes of formatted storage. And the Company's two-drive, two-spindle SA482-HA drive with 1.244 gigabytes for formatted storage capacity, is now available for $34,000. Digital's RA family of 14 inch hard drive products, which come with a standard 3-high cabinetry configuration, have also been repriced. The RA82-EA/DE, now available for $51,000, offers 1.86 megabytes of formatted capacity. The RA82-DA/DD, featuring 1.24 gigabytes of formatted storage capacity, is available for $34,000. And, the RA82-CA/CD, now at $17,000, offers 622 megabytes of formatted capacity. All prices are U.S. list, quantity-one. Volume discounts are available to qualified customers. Availability for all drives is immediate. Digital's recent announcement of VAXsimPLUS service technology represents a significant advancement in diagnostic service technology, resulting in a major improvement in storage system availability. This advanced service technology provides an "early warning" capability that alerts customers to potential problems. The PLUS represents the addition of expert-based software for predicting potential failures in RA series Winchester disks prior to any data loss. VAXsimPLUS service technology can automatically invoke VMS Volume Shadowing software to create a "real-time" copy of the suspect disk. The duplicate disk is then used to keep the system operating while the failing component is replaced and the disk is returned to operation virtually eliminating disk-related downtime. Available this May, VAXsimPLUS service technology is provided at no charge to customers whose VAX/VMS systems are covered by a Digital system warranty or a standard service agreement. Digital Equipment Corporation, headquartered in Maynard, Massachusetts, is the world's leading manufacturer of networked computer systems and associated peripheral equipment, and is the leader in systems integration with its networks, communications, software and services products. ================================================================================ Note 40.42 PERIPHERAL DEVICE ANNOUNCEMENTS 42 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 30 lines 1-AUG-1988 16:44 -< Bye-bye LA100-CA/CB. Buy BA/BB with LA10X-FL. >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! LA100-CA/CB RETIREMENT - 13-June-1988 ******************************************************************************* o Final inventories limited o Replacement products suggested The LA100-CA AND LA100-CB KSR dot matrix printers, with multiple font capability, are being retired. The replacement products are the LA100-BA and LA100-BB plus an LA10X-FL multiple font option. The LA100-BA/BB product is identical to the retiring product but it does not contain the built-in multiple font functionality. The LA10X-FL font must now be ordered separately. Current Product Replacement Product LA100-CA LA100-BA + LA10X-FL LA100-CB LA100-BB + LA10X-FL Current inventories of the LA100-CA/CB are planned for depletion by the end of June 1988. The LA100-BA and LA100-BB replacement products will continue to be available for approximately one more year. ================================================================================ Note 40.43 PERIPHERAL DEVICE ANNOUNCEMENTS 43 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 63 lines 1-AUG-1988 16:45 -< LP27-SA for MicroVAX 3500/3600 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! LP27-SA FOR MicroVAX 3500/3600 - 13-June-1988 ******************************************************************************* 1200 LPM PRINTER o New Q-bus model in the LP27 printer series o Suited for print applications of 200,000 pages/month o Used with MicroVAX 3500/3600 o Available now PRODUCT DESCRIPTION The LP27-SA is the latest LP27 printer family member that can be connected to Digital's MicroVAX 3500/3600 processors. Capable of printing AT 1200 lpm, the LP27-SA is a companion product to the LP27-QA variation (MicroVAX II), completing the LP27 Q-bus offerings. The LP27-SA connects to the MicroVAX 3500/3600 via the LPV11-SA controller. It is an ideal companion printer for these processors, offering reliable printing for applications requiring 200,000 pages/month. Both the LP27-QA and LP27-SA are Digital 1200 lpm printers designed for Q-bus systems and include all the cables and controllers necessary for connection. CONFIGURATION INFORMATION The LP27-SA is qualified on the MicroVAX 3500/3600 and is supported by a maximum cable length of 30 feet. SOFTWARE SUPPORT The LP27 is supported on both VMS and ULTRIX. PRICING/ORDERING INFORMATION U.S. Discount Order No. Description Seg. List Price Instal. BMC E/U OEM LP27-SA 1200 lpm Bandprt with 090 $29,950 $630 $267 Y 2 LPV11-SA controller, 120 V AVAILABILITY LP27-SA is available now. SUPPORT SERVICES The LP27 is offered with a one-year warranty. It is supported by a full line of on-site service offerings, including Basic and DECservice. Installation is included when the LP27 is ordered as part of a system order; when purchased separately, there is an installation charge (see above). ================================================================================ Note 40.44 PERIPHERAL DEVICE ANNOUNCEMENTS 44 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 40 lines 1-AUG-1988 16:48 -< Prices reduced on RA82 preconfigured 8250/8350s >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! PRICE REDUCED ON RA82 - 22-April-1988 ******************************************************************************* -BASED VAX 8250/8350 PRECONFIGURED SYSTEMS o $3,000 price reduction on VAX 8250/8350 preconfigured systems o New pricing effective immediately Effective immediately, all VAX 8250 and VAX 8350 Preconfigured Systems are reduced in price by $3,000. The VAX 8250/VAX 8350 Preconfigured Systems all utilize the 622 MB high-capacity RA82 disk drive, which was announced in November of last year. All these systems are now reduced in price by $3,000 to incorporate the recent price reduction of the RA82 Old U.S. New U.S. Model No. Description List Price List Price SV-8A47A-GL/GM 8250/1 PRECONFIGURED SYSTEM PP $126,000 $123,000 SV-8A47A-GK/GN 8250/1 PRECONFIGURED SYSTEM PDUP 145,950 142,950 SV-8A47N-GL/GM 8250/2 PRECONFIGURED SYSTEM PPL 136,500 133,500 SV-8A47N-GK/GN 8250/2 PRECONFIGURED SYSTEM PDUP 156,450 153,450 SV-8B47A-GL/GM 8350/1 PRECONFIGURED SYSTEM PPL 152,250 149,250 SV-8B47A-GK/GN 8350/1 PRECONFIGURED SYSTEM PDUP 177,450 174,450 SV-8B47N-GL/GM 8350/2 PRECONFIGURED SYSTEM PPL 162,750 159,750 SV-8B47N-GK/GN 8350/2 PRECONFIGURED SYSTEM PDUP 187,950 184,950 ================================================================================ Note 40.45 PERIPHERAL DEVICE ANNOUNCEMENTS 45 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 156 lines 1-AUG-1988 16:49 -< Prices reduced for SA482/RA82 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! PRICE REDUCTION FOR THE SA482/RA82 - 28-April-1988 ******************************************************************************* o Prices on all variants of the SA482/RA82 product family are reduced by 19% to 28%, depending on configuration o All variants of the SA482/RA82 product family are on Fastship status o Actual MTBF higher than any quoted large drive spec in the industry Digital announces major price reductions on all variants of the SA482/RA82 product family. The reduction gives this product family significant competitive advantage on the price/performance and price/MByte metrics. Also, the RA82 is now priced at or below the RA81 while delivering greater capacity, performance and reliability. This will provide your customer with greater incentive to migrate from the RA81 to the RA82. The new price structure is as follows: U.S. Order No. Description List Price SA482-AA/AD Storage Array, 2.488 GB, four spindles $66,000 SA482-LA/LD Storage Array, 1.866 GB, three spindles 51,000 SA482-HA/HD Storage Array, 1.244 GB, two spindles 34,000 RA82-EA/ED Three RA82-AA/AD, 3-hi cabinet 51,000 RA82-DA/DD Two RA82-AA/AD, 3-hi cabinet 34,000 RA82-CA/CD One RA82-AA/AD, 3-hi cabinet 17,000 RA82-AA/AD One 622 MB disk, add-on 17,000 RA82-AE/AF One 622 MB disk, no SDI cable 16,621 NOTE: Discount Structure, BMC, DSMC and installation charges are unchanged. PRODUCT FAMILY SUMMARY The SA482/RA82 storage product family is summarized in the following variant table: SA482/RA82 PRODUCT FAMILY VARIANTS Single U.S. # of Drives (4-hi) SA482 (3-hi) RA82 Add-on List Price Per Drive 4 -AA - - $66,000 $16,500 3 -LA -EA - 51,000 17,000 2 -HA -DA - 34,000 17,000 1 - -CA -AA 17,000 17,000 Note that the price is based on the number of drives in a configuration, regardless of the kind of cabinet used. The SA482 is an exception, with an implied spindle price below the other variants. This pricing structure essentially eliminates the charge for a cabinet, in addition to reducing the price on the RA82 itself. IMPORTANT MARKETING MESSAGES Product Availability: A significant increase in manufacturing capacity has made the RA82 readily available for immediate delivery. Inventory is on hand and all variants of the SA482/RA82 product family are on FASTship delivery status. Reliability: The SA482/RA82 product family has been subjected to the most rigorous reliability testing and evaluation of any disk product ever produced by Digital. The resulting product and process improvements have produced Digital's most reliable large disk drive. Recent data from the On-going Reliability Testing (ORT) lab and from the field indicate an actual MTBF for the RA82 is significantly higher than the quoted specs of any large disk drive in the industry. Furthermore, the vast majority of customer comments on the RA82 reliability have been very favorable. Maintenance: As a demonstration of our confidence in the reliability of the RA82, we have set the Basic Monthly Charge (BMC) at $59. This is the lowest BMC of any large disk in the industry and makes a significant contribution to a low cost of ownership on the RA82. DSA Compatibility: Digital guarantees 100% DSA compatibility with all its storage products. Each drive and controller is subjected to intensive, detailed testing and evaluation with all other DSA drives and controllers and with all products which interface to the subsystem, including software. It is only after this long, thorough and rigorous evaluation that any storage product is qualified by Digital to be 100% DSA compatibile. Digital makes this promise on all currently offered DSA products and on DSA products which will be offered in the future. This is the essence of the solution offered by the Digital Storage Architecture for your customers' needs in compatibility, expandability and investment protection. Extensive in-house testing of other "DSA compatible" products in the marketplace has revealed serious shortcomings to DSA compatability. Performance: Although third-party disk drives typically display performance specifications which appear to be superior to the RA82, a thorough in-house evaluation of these products has produced a very different conclusion. In most applications, third-party disk drives have shown no significant performance advantage over the RA82. Warranty: All members of the SA482/RA82 product family come with a one-year warranty which includes DECservice, Digital's premier on-site service offering. Data Integrity: Only Digital storage products deliver data with the kind of integrity that meets the rigid Digital specifications and protects your customers against data loss and degradation. This is accomplished with Digital's industry-leading 170-bit error correction code, which can correct up to eight 10-bit bursts in any block. Data Availability: Through the Digital Storage Architecture, the RA82, like all other DSA disk drives, provides high data availability through automatic bad block replacement and revectoring. This feature is invisible to the user, and there is no shrinkage in available storage capacity over time. Consistent Product: The RA82 spindle is the basic storage unit in all variants of the SA482/RA82 product family. The add-on RA82-AA/AD spindle is the upgrade part for either the RA82-xx three-high cabinet variants or the SA482-xx Storage Array (four-high) cabinet variants. Support: All variants of the SA482/RA82 product family are supported on the VAXBI, UNIBUS VAX Systems and VAXclusters (CI). All RA82 variants are also supported on both MicroVAX II and MicroVAX 3600 systems. Version 4.4 is the minimum VMS for all variants of the SA482/RA82. There is no 18-bit support for this product family. COMMONLY ASKED QUESTIONS CAN AN SA482-HA OR SA482-LA BE UPGRADED WITH AN RA82-AA? The RA82-AA/AD single spindle is the upgrade unit for both the 3-high cabinet and the Storage Array (4-high) cabinet variants. CAN MY CUSTOMER ORDER A MIXED CABINET OF RA81s AND RA82s? Digital does not offer a specific product containing a mix of RA81 and RA82 drives. However, both drives are compatible with all existing 3-high and 4-high cabinets and may be installed in the field. The one exception is the Storage Array cabinet, which may accommodate only RA82 drives. IS THE RA82 FIELD ADD-ON COMPATIBLE WITH EXISTING THREE- AND FOUR-HIGH CABINETS? The RA82-AA/AD add-on drive may be field installed in any existing 3-high, 4-high or Storage Array (4-high) cabinet. CAN THE CUSTOMER MOUNT AN RA60 IN AN EMPTY SLOT IN THE SA482-HA OR SA482-LA? The RA60 cannot be mounted in the Storage Array cabinet. However, it can be mounted in any other 3-high or 4-high cabinet with any combination of RA81 or RA82 drives. ================================================================================ Note 40.46 PERIPHERAL DEVICE ANNOUNCEMENTS 46 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 50 lines 1-AUG-1988 16:50 -< Bye-bye RC25 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! RC25 END OF LIFE - 06-June-1988 ******************************************************************************* o RC25 product phase down is underway o Last order accepted July 1, 1988; drives no longer available after that time o Orders filled on a first-come, first-served basis, while quantities last o Last shipments on September 30, 1988 o Cartridges will continue to be available The RC25 has been a successful product for Digital during the past four years. However, the RC25 family is rapidly reaching end of life and we must communicate this to our customers. Shipments have declined significantly during each of the past three quarters, resulting in increased costs. Manufacturing capacity will be better utilized for newer, higher volume products. ORDERING/SHIPPING INFORMATION To ensure sufficient units are built to support ongoing customer requirements and long-term contractual needs, please submit orders by July 1, 1988, with final shipments scheduled by September 30, 1988. Any exceptions will be handled on an individual basis by your Operations Specialist. Cartridges will still be available after end-of-manufacturing life. They may be ordered through DECdirect. There is no family of follow-on fixed/removable Digital products planned. Please position your accounts to move towards the RD series disks with TK tape back-up products, or use Digitals's high-end storage RA series of fixed and removable disks. SUPPORT INFORMATION Digital Field Service will continue to support the RC25 with its complete range of service products even after the end of the product's manufacturing life. The availability of spare parts and trained engineers is ensured as a part of the standard end-of-life planning. ================================================================================ Note 40.47 PERIPHERAL DEVICE ANNOUNCEMENTS 47 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 65 lines 1-AUG-1988 16:52 -< RF-VT240 Tempest terminal >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! RF-VT240 TEMPEST TERMINAL SOLUTION - 06-June-1988 ******************************************************************************* o Available 30 days after receipt of order o Full TEMPEST certification to NACSIM 5100A PRODUCT DESCRIPTION The TEMPEST RF-VT240 terminal consists of a detachable, low-profile TEMPEST keyboard (RF-LK201), a system box, and a 12" black and white monitor. It has a built-in serial printer port, down-line loadable character sets, programmable function keys and VT100/220 capabilities. The RF-VT240 features VT240 functionality, which includes full bit-mapped graphics with 800-by-240 pixel resolution, in both ReGIS and Tektronix 4010/4014 protocols. For TEMPEST connection to the host CPU or printer, a shielded cable is required. RESOURCES For information on all of our TEMPEST products, please refer to the TEMPEST Reference Catalog (ED-30844-31/88), which is available from your local Sales Communications Center. POSITIONING The TEMPEST RF-VT240 provides the terminal solution for those environments requiring TEMPEST compliance. PRICING/ORDERING INFORMATION U.S. Discount Order No. Description List Price BMC E/U OEM RF-VT240-AA Monitor, system unit and $4,195 $72 Y B keyboard RF-VT240-BA Monitor, system unit and 4,195 72 Y B WPS keyboard RF-BC09D-XX* Host CPU shielded cable |- |- connection U.S. Discount Order No. Description List Price BMC E/U OEM RF-BC09E-XX* Draft printer shielded |- |- cable connection *Specify length "XX" - 10, 15, 25, 50 or 100 foot |-Price depends on length. See U.S. Price List. AVAILABILITY The RF-VT240 TEMPEST Graphics System is available 30 days after receipt of order. Tektronix is a registered trademark of Tektronix, Inc. ================================================================================ Note 40.48 PERIPHERAL DEVICE ANNOUNCEMENTS 48 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 215 lines 1-AUG-1988 16:54 -< SA600 storage array - RA90s >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! THE SA600 STORAGE ARRAY - 21-July-1988 ******************************************************************************* o Capacity for the largest corporate databases o Performance for demanding TP applications o Mainframe storage in a 5.5 sq. ft. footprint o Maximum system up-time with VAXsimPLUS predictive maintenance o Customer investment protection with Digital Storage Architecture/Standard Disk Interconnect (DSA/SDI) compatibility o Greater customer satisfaction with new cable ordering rules Introducing a revolution, the 9.7 GB SA600 storage array and the 4.8 GB SA600 storage array building block (SABB). Digital has made a major commitment, backed with the investment of extensive resources, to insure its high-end disk products are state-of-the-art today and in the future. The SA600 storage array reinforces that Digital world-class TP and database solutions today. With this introduction Digital has established a storage array product family. The SA600 storage array and SA600 building block are Digital's high-end storage arrays for the most demanding applications. The SA482 and SA482 building blocks are Digital's entry-level storage arrays for users who want low entry price, less than 4.8 GB of storage potential, or fast delivery. Both the SA482 and the SA600 are perfectly suited for DECtp and database applications. VAXsimPLUS Digital's new predictive maintenance tool, VAXsimPLUS, is particularly important to the storage array strategy. VAXsimPLUS gives SA600 subsystems the up-time and data availability customers expect in TP and high-end data base environments. The SA600 storage array comes with a one-year warranty which includes DECservice, Digital's premier on-site service offering. VAXsimPLUS is available to all customers whose systems are under warranty or have our standard on-site service agreement. PRODUCT DESCRIPTION The SA600 storage array is based on the 1.2 GB random access moving head RA90 component disk drive. The new RA90 drive is designed with advanced thin-film heads and thin-film 9-inch media which provides an industry- leading density of over 40 million bits per square inch of recording surface. At this density, the RA90 is technologically superior to announced minicomputer and mainframe disks, such as the IBM 3380K, and places Digital in a leadership position. Each SA600 storage array can have up to eight fully independent Digital Storage Architecture/Standard Disk Interconnect (DSA/SDI) compatible RA90 component drives mounted in it. The SA600 storage array cabinet has the same physical dimensions and is electrically similar to the SA482 storage array. SA600-JA Fully configured SA600 storage array. The SA600-JA is composed of eight RA90 1.2 GB component disk drives in a 60-inch high storage array cabinet offering 9.7 GB of storage. SA600-HA Storage array building block (SABB) configuration. The SA600-HA is partially populated with four RA90 component drives in the same 60-inch high storage array cabinet and offers 4.8 GB. The SA600-HA SABB is preconfigured with all the necessary internal cables and mounting hardware for maximum ease of service and expansion. At this time, the RA90 drive is available only as a component drive for the SA600-JA/HA. In the future, additional RA90 drives can be easily and quickly added to this SABB in the field without affecting system up-time or the operation of other existing drives in the storage array. Only RA90 component drives can be added to SA600-HA SABB configurations. FEATURES/BENEFITS The following features/benefits show what your customers need and how to win with the SA600. Benefit Associated Features Outstanding A fully configured SA600 storage array packs 9.7 GB of Footprint formatted user storage in a mere 5.5 square feet of computer room floor space. This is more than double the capacity of current mainframe products from other vendors. High The SA600 storage array extends the successful reliability Reliability record of the SA482 storage array through the use of automated manufacturing facilities and extensive qualification testing. High The RA90 component drive has an average seek time of 17.5 Performance msecs and a peak transfer rate of 2.8 MBytes/sec resulting in a typical VAXcluster subsystem I/O response time of 23.5 msecs. The RA90 provides response time parity with traditional mainframe storage solutions. With up to 8 independent RA90 actuators, the SA600 storage array offers total throughput that eclipses traditional mainframe storage solutions. Investment As illustrated by the SA600 storage array, DSA compati- Protection bility across technology generations protects your customers' total investments in hardware, software and expertise. Maximum SA600 and SA482 storage arrays in many cases can be recon- Flexibility figured, upgraded, serviced, and, in the case of SABBs, expanded without affecting system availability. Maximum Whether it is an HSC, a VAX system, or another storage Scaleability array additional functionality, performance or capacity can be dynamically added to a VAXcluster system without the need for system or cluster interruption. A 16-node VAXcluster system can now exceed 200 GB. High Data DSA, VAXcluster systems, VAXsimPLUS, and storage arrays and Availability data integrity add up to mainframe class disk subsystem up-time. HSC shadowing, dual port capability, and RA90 ease of service means maximized data availability and configurations with no single point of failure. In addition to the wealth of data availability and integrity features offered by DSA/SDI, VAXsimPLUS proactively identifies failing field replaceable units (FRUs) for replacement before unexpected applications interruption or data loss. In a VAXcluster (via HSC shadowing), VAXsimPLUS can predict and automatically fail-over the data without unexpected interruption of service. SA600 SYSTEMS SUPPORT AT THIS TIME VAXcluster Supported for connection to HSC70 and HSC50 servers in Systems VAXcluster systems with appropriate software and hardware revisions. VAXBI Supported for connection to KDB50 VAX systems with Systems appropriate software and hardware revision levels. UNIBUS VAX Supported for connection to UDA50 VAX/VMS systems with Systems appropriate software and hardware revision levels. Q-bus VAX Not offically supported for connection to KDA50s on Systems MicroVAX 3600 systems at this time. PDP-11 Not officially supported on any PDP-11 systems at this Systems time. VMS Full support with V5.0. Support as a data disk in V4.6 and V4.7. ULTRIX-32 Support in upcoming major functional release. VAXELN Support in an upcoming release. HSC70 and Must run Version 3.70 or above of HSC software. HSC50 SA600 CONFIGURATION RULES With the SA600 storage array we now offer customers the flexibility of ordering their own external SDI cables. YOU MUST ORDER BC26V-xx SDI CABLES SEPARATELY WITH SA600 ORDERS! This is a change in policy for DSA disk products. Select the desired cable length 12', 25', 50' or 80' from the U.S. Price List. Each RA90 component drive in the SA600 requires a minimum of one SDI cable (two for each drive if being dual ported). Six-foot SDI cables are not supported. SPECIFICATIONS SA600-JA SA600-HA Performance: (1) Number of RA90 actuators Eight Four Transfer rate 2.8 MBytes/second 2.8 MBytes/second Disk subsystem response time (2) 23.5 milliseconds 23.5 milliseconds Average seek time (3) 17.5 milliseconds 17.5 milliseconds Media Characteristics: Formatted capacity 9.728 GB 4.864 GB Unformatted capacity (4) 12.863 GB 6.431 GB Tracks per inch 1750 1750 Bits per inch 22,839 22,839 Rotational speed 3600 rpm 3600 rpm Number of data surfaces 104 (13/drive) 52 (13/drive) Number of heads per surface 1 1 Sectors per track 69 + 1 spare 69 + 1 spare Tracks per logical cylinder 13 13 Power Requirements: (5) Power receptacle type option -xA L21-30R L21-30R option -xD IEC 309 IEC 309 Heat dissipation 7900 Btu/h 3950 Btu/h 2314 watts 1157 watts (1)Performance specifications (and others as indicated) are provided on a per-drive basis. All others apply to a fully configured array. (2)For a typical VAX I/O environment. (3)With specification margin and subsystem overhead--19.0 milliseconds. (4)Unfomatted capacity provided for comparison purposes; only formatted capacity is user accessible in any disk drive. (5)Consult the Storage Array Family Information Sheet for additional data. ================================================================================ Note 40.49 PERIPHERAL DEVICE ANNOUNCEMENTS 49 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 47 lines 1-AUG-1988 16:55 -< TK50/TK70 clarifications >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! TK50/TK70 CLARIFICATIONS - 06-June-1988 ******************************************************************************* o This article is intended to clarify a few common misconceptions about the TK50 and TK70 tape devices A TK70 drive cannot write a cartridge in TK50 mode. During the development of the TK70 this feature was dropped because the perceived benefits were outweighed by the impact that such a feature would have had on product cost and time to market. The following matrix should clearly define the compatibilities between initialized cartridges and drives. -------------------------------------------------- | Read / Write in: | | | | TK50 | TK70 | | Drive | Drive | ------------------------------------------------- | Cartridge Initialized in: | | -------------------------------------------------| R = Read TK50 TK50 | R/W | R | R/W = Read/Write |---------------------| NC = Not Compatible TK70 | NC | R/W | --------------------- In recent months, we have had several questions on whether a TK50 or a TK70 should or can be used in realtime applications as a data logger, or in journaling types of applications. The TK50/TK70 are streaming tape devices and are best suited for continuous data transfer applications such as disk backup. They do not perform efficiently as data loggers in realtime applications because these streaming tape drives cannot start and stop within an inter-record gap and must take time to reposition themselves. Specifically, these drives may be unavailable to transfer data for periods of from four seconds (reposition time) up to four minutes (worst case--resynchronization). Customers designing applications that require realtime response should be advised of these performance factors. They may want to use disk or disk/tape combinations to maximize application performance. ================================================================================ Note 40.50 PERIPHERAL DEVICE ANNOUNCEMENTS 50 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 192 lines 1-AUG-1988 16:57 -< VAX RTA (Real Time Accelerator) H/W and S/W >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! VAX REALTIME ACCELERATOR PRODUCTS - 09-May-1988 ******************************************************************************* o VAX RTA software integrages VMS and VAXELN o VAX RTA hardware distributes realtime processing o Program announcement of VAX RTA VMS 5.0 support o DRB32 no longer restricted to selective selling PRODUCT DESCRIPTION VAX Realtime Accelerator (VAX RTA) hardware and software integrate VMS general-purpose systems and VAXELN dedicated realtime processors into a single, tightly integrated system. FORTRAN, EPASCAL, C and AdaR applications can be developed on the VMS host processor, using the standard VAXELN Toolkit, and executed on dedicated VAXELN RTA processors inside the same system. The VAX RTA hardware and software products improve overall system performance by distributing the computing power closer to the realtime event that needs it. Applications can be run on a single VAX RTA processor, distributed over multiple VAX RTA processors, or even between the VMS host and one or more VAX RTA processors. VAX RTA Hardware o One or more VAXBI-based KA800 realtime processors, expandable to 11 megabytes. o Used with a DRB32 parallel port to control high-speed I/O functions. o Available as a system add-on option. VAX RTA Software (Version 1.0) o A VMS layered product that allows VMS host processors to load, control and communicate with VAXELN realtime KA800 processors over the VAXBI bus. o VMS host processor control via program or a command language interface. o Communication among VMS and VAXELN processors occurs using the standard VAXELN message passing services. o The VAX RTA Software option includes a User Guide and Installation Guide. o Markets: - Government Prime Contractors - Department of Defense - Department of Energy - Aerospace - NASA - System Integrators - OEMs o High-performance realtime systems used in simulation and training, image processing, aerospace and defense applications. o Computer front-ends to handle data reduction for signal processing, telemetry, and radar and sonar data streams. o Applications requiring interrupt response times within a predictable time critical window. PERFORMANCE GAINS One of the most important features of VAX RTA for realtime applications is the ability of the KA800 processors to rapidly and efficiently communicate and synchronize with each other and with the host system processor(s). Preliminary performance testing results indicate: o Host and target processors achieve communication throughputs as high as 4 megabytes/sec (40 times network performance!), dependent on the system configuration, message size, and the mechanism used for performing the operation. GENERAL CONFIGURATION INFORMATION o A typical VAX RTA configuration consists of a VAXBI-based host system running VMS V4.6 or 4.7. o To this system is added the VAX RTA V1.0 layered software product and one or more VAX RTA KA800 processors running VAXELN. o Each RTA KA800 can have up to five MS800 private memory extensions. o A DRB32 high-speed parallel port can be added and controlled by the KA800 which can service interrupts on the VAXBI bus. o VAXBI memory, MS820, can be used as a memory buffer to collect high-speed data input from a user device. o One or more KA800's can then be assigned to compute on the data acquired. .__________________________. | | | VMS HOST PROCESSOR | | | | VAX RTA LAYERED SOFTWARE | |_______. .________| | | .______| |_________________________________________________. / \ \ V A X B I B U S / \____. ._____________. .___________. .________. .______/ .___| |____. .___| |____. ._| |_. .__| |__. | KA800 | | KA800 | | MS820 | | DRB32 | | VAXELN USER |_. | VAXELN USER |_. | VAXBI | | PARALLEL | | PROGRAM | | | PROGRAM | | | BUFFER | | PORT | |_____________| | |_____________| | | MEMORY | |___ ___| |MS800 PRIVATE| |MS800 PRIVATE| |________| /\ | MEMORY | | MEMORY | || |_____________| |_____________| .___ \/ ___. | | | USER | | DEVICE | |__________| SYSTEM-SPECIFIC CONFIGURATION INFORMATION o The KA800 and the MS800 modules are considered to be a single option: - Must be installed in adjacent slots within the same six-slot VAXBI card cage -- cannot cross over card cage boundaries. - A maximum of five MS800 modules can be used with one KA800 module. o The KA800 can be used to directly control the DRB32 high-speed parallel port. Use of the KA800 to control other VAXBI adapters is not supported. However, the KA800 can be physically configured with any valid VAXBI adapter in the same backplane (except the DWBUA which must be installed on a separate VAXBI backplane). o The total amount of memory visible from a VAXBI (host memory plus KA800/MS800 + MS820 memory) cannot exceed 256 Mbytes. o The number of VAX RTA subsystems that can be added to a system depends on the VAXBI slots, nodes and the amount of power available. Refer to the Systems and Options Catalog for system and DRB32 configuring information. o VAXELN applications running in a KA800 processor can be debugged remotely from the VAX/VMS host and locally from the KA800. Local debug requires use of the optional CK-KA800-LN console serial line. Kernel code in the KA800 processors can only be debugged locally. Only one console serial line is required for each KA800 which is active in a debug session. o VAX RTA Version 1.0 software can be installed on any VAXBI based host processor running VMS Version 4.6 or 4.7. o The VAXELN Version 3.1 Toolkit and VAX RTA software are required on a host system for application development and for remote debug of KA800 target processors. The actual running of the application requires only VAX RTA software and a VAXELN runtime license which is included with each KA800. o VAX RTA is not supported in VAXcluster configurations. o Refer to the Software Product Descriptions for software configuration information. POWER REQUIREMENTS Mounting dc Amps Drawn @ VAXBI I/O Panel Option Requirements +5V +5VB +12V -12V -5.2V -2V Nodes Units KA800-M 1 VAXBI Slot 7.26 2.75 0.07 0.07 0.00 0.00 1 none CK-KA800-LN none 0.00 0.00 0.00 0.00 0.00 0.00 0 2 MS800-EA 1 VAXBI Slot 1.50 1.46 0.00 0.00 0.00 0.00 0 none MS820-BA 1 VAXBI Slot 0.63 2.50 0.00 0.00 0.00 0.00 1 none MS820-CA 1 VAXBI Slot 0.63 3.10 0.00 0.00 0.00 0.00 1 none o Current calculations for +5V and +5VB must be performed to determine the maximum number of modules configurable in any system. For 82** and 83** systems, 16.35A of +5VB is available for each pair of six slot VAXBI card cages; 16.35A in Config 1 and 32.70A in Config 2. For all other VAX systems, the same power supply channel delivers current to +5V and to +5VB. o Refer to the Systems and Options Catalog for available system power and DRB32 power requirements. SUPPORTING ANNOUNCEMENTS Digital is also program announcing support under VMS V5.0 with the next major release of VAX RTA software. Your customers can: o Begin applications development with Version 1.0 of VAX RTA today, o And migrate to the VMS 5.0 environment without modifications when the next release of VAX RTA becomes available. The DRB32 high-speed VAXBI parallel port is no longer subject to selective selling; no approval process is required. The DRB32 is the preferred method for connecting complementary third-party devices to VAXBI-based systems. Ada is a registered trademark of the U.S. Government (Ada Joint Program Office) ================================================================================ Note 40.51 PERIPHERAL DEVICE ANNOUNCEMENTS 51 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 289 lines 1-AUG-1988 16:58 -< VAX RTA pricing/ordering/availability >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! PRICING/ORDERING/AVAILABILITY INFO - 10-May-1988 ******************************************************************************* VAX RTA FAMILY Hardware U.S. Discount Order No. Description List Price BMC E/U OEM KA800-M VAX RTA PROCESSOR $7,000 * Q Q CK-KA800-LN 15 FT CONSOLE LINE FOR KA800 400 N/A Q Q MS800-EA 2 MB KA800 EXTENSION MEMORY 2,000 $0** Q Q *Call Field Service **The Basic Monthly Charge (BMC) for memory is $0 (zero) when purchased and installed on a system covered by a Digital Field Service contract. The Digital Business Agreement (DBA) discounts apply with a "0" product adder; however, the MS800-EA is not available on the SVA Volume Memory Discount Schedule and cannot be sold under the terms of the MVDA. Software U.S. Order No. Description List Price Type CPU Q5ZPU-UZ VAX RTA LIC. W/WARR. PAID-UP $ 3,480 Primary 8250 Q7ZPU-UZ VAX RTA LIC. W/WARR. PAID-UP 4,524 Primary 8350 Q9ZPU-UZ VAX RTA LIC. W/WARR. PAID-UP 6,380 Primary 8530 Q2ZPU-UZ VAX RTA LIC. W/WARR. PAID-UP 7,656 Primary 8550/8700 QMZPU-UZ VAX RTA LIC. W/WARR. PAID-UP 10,440 Primary 8800 SPS - Traditional Services Model No. DPMC BSMC SMMC MUMC H-Kit HZ HZMC KZMC Inst. Q(5,7)ZPU-*M $122 $ 87 $ 52 $31 $950 $428 $24 $21 $380 Q(2,9)ZPU-*M 142 99 57 36 950 428 26 21 380 QMZPU-*M 144 103 57 36 950 428 26 21 380 ____________________________________________________________________________ rtVAX 3XXX FAMILY U.S. Discount Order No. Description List Price BMC Seg. E/U OEM DW-320R1-AA rtVAX 3200(Ped.) $18,950 TBS XX Y L DW-320R2-AA rtVAX 3200(Rack) 18,950 TBS XX Y L DW-350R1-AA rtVAX 3500(Ped.) 24,000 TBS XX Y L U.S. Discount Order No. Description List Price BMC Seg. E/U OEM DW-350R2-AA rtVAX 3500(Rack) $24,000 TBS XX Y L DW-360R1-AA rtVAX 3600(Cab.) 65,000 TBS XX Y L Availability: Q1/FY88 in limited quantities; Q1/FY89 in volume. The six levels of service available for these configurations are detailed in the FS Pricing Preface Pages of the U.S. Price List and FS Price List and are as follows: BMC DSMC BNS DNS BSS DSS DW-320R1/R2 $203 $242 $223 $265 $279 $332 DW-350R1/R2 218 260 238 283 294 350 DW-360R1 308 367 328 390 384 457 NOTE: These are the same integrated services available for MicroVAX 3XXX. Warranty: These systems are sold with a one-year system level warranty which includes both hardware and software service. See service information included in the rtVAX article for details. Levels of Service Hardware Only Service: For any system -- either timeshare systems or certain servers on LAVc or certain systems used as nodes on LAVc (or ULTRIX/NFS equivalent): BMC Basic Monthly Charge No guaranteed response time. Work performed only during coverage hours Work-through is at extra charge DSMC DECservice Monthly Guaranteed response time Charge Work started only during coverage hours Guaranteed Work-through round-the-clock at no extra charge once work is begun Integrated Hardware/Software Service: System Level Service: For any system -- either timeshare systems or certain servers on LAVc and certain systems used as nodes on LAVc (or ULTRIX/NFS equivalent): BSS Basic System Service Hardware on site - same as BMC plus 24 hour/7 day toll free telephone remedial and advisory support and call logging for both software and hardware DSS DECsystem Service Hardware on site - same as DSMC plus 24 hour/7 day toll free telephone remedial and advisory support and call logging for both software and hardware For Basic System Service and DECsystem Service, the software component includes both remedial and advisory service through telephone support with critical on-site support, access to DSIN, Right-to-use software updates. Coverage is for operating system, DECnet and LAVc or equivalent for ULTRIX/NFS or VAXELN. Node Level Service: This service offering is for LAVc client nodes only (or equivalent for ULTRIX/NFS). It is available only in conjunction with a designated customer system manager owning a server system or timeshare system under a BSS or DSS contract. BNS Basic Node Service Same as BSS except all remedial/advisory calls and access to DSIN are routed through the designated customer system manager DNS DECnode System Same as DSS except all remedial/advisory calls and access to DSIN are routed through the designated customer system manager ____________________________________________________________________________ E-SERIES INDUSTRIAL SYSTEMS U.S. Order No. Description List Price BMC DH-630R1-A2/A3 NEMA-12 CPU Enclosure, KA630-AA CPU, $30,750 $293 4 MB memory, RQDX3-SC Disk Controller; NEMA-12 Disk Enclosure, 71 MB RD53-A Disk, 1.2 MB RX33-A Floppy Disk, Documentation and Diagnostics, 120V/240V DJ-620R1-A2/A3 NEMA-12 CPU Enclosure, KA620-AA CPUT 9,950 128 DEQNA-SA Ethernet Interface, VAXELN Lic. Documentation and Diagnostics, 120/240V DH-183R1-A2/A3 NEMA-12 CPU Enclosure, KDJ11-BF 26,950 223 PDP-11/83 CPU, 2 MB MSV11-JE memory, RQDX3-SC Disk Controller, NEMA-12 Disk Enclosure, 71 MB RD53-A Disk, 1.2 MB RX33-A Floppy Disk Documentation and Diagnostics, 120V/240V U.S. Order No. Description List Price BMC BA221-BA NEMA-12 BA221-BA Disk Enclosure $12,750 $ 92 expansion of DJ-620R1 VAXELN system only. 71 MB RD53-A Winchester Disk, RX33-A, Floppy Documentation and Diagnostics, 120V/240V NOTE: You cannot add additional disk boxes systems that already have them. NOTE: The above are Segmentation Code 092; E/U Type A and OEM Type 2 discountable. NOTE: Use option variations designed for BA213, BA214. Availability: First customer ship in limited quantities scheduled for May 1988; volume shipments to begin July 1988. ____________________________________________________________________________ VAXELN VERSION 3.1 Single User Target Host Development Host Cluster Runtime Order No. Processor License License License Q8***-** VAXstation 2000 $ 1,050 Q4***-** VAXstation II 1,407 Q3***-** MicroVAX 2000 2,100 QZ***-** MicroVAX 4,200 $ 420 QC***-** VAX-11/730 7,035 420 QD***-** VAX-11/750 8,516 $ 5,103 504 QE***-** VAX-11/780 8,516 5,103 Q5***-** VAX 8200 8,453 5,072 Q7***-** VAX 8300 10,975 6,585 Q0***-** MicroVAX 3500/3600 12,660 1,200 Q9***-** VAX 8530 15,477 9,282 11,550 QK***-** VAX 8600 15,477 9,282 Q2***-** VAX 8550/VAX 8700 18,375 11,151 11,550 QM***-** VAX 8800 25,326 15,194 13,550 Availability: FCS is scheduled for Q1/FY88 ____________________________________________________________________________ rtVAX REALTIME REPRICING rtVAX 1000 Former U.S. New U.S. Order No. Description List Price List Price BMC SEG E/U OEM 620QY-A2/A3 rtVAX, BA23, $ 7,961 $ 5,347 $65 XX Y L KA620, BA23AF, 120V/240V Former U.S. New U.S. Order No. Description List Price List Price BMC SEG E/U OEM 620QZ-A2/A3 rtVAX, BA23, $ 7,961 $ 5,347 $65 XX Y L KA620, BA23AR, 120V/240/V DH-62Q1-AA/A3 rtVAX, BA23, 11,853 7,311 80 XX Y L KA620, DEQNA, BA23AF, 120V/ 240V DH-620Q1-CA/C3 rtVAX, BA23, 12,296 8,819 99 XX Y L KA620, RQDX3, RD31, RX33, BA23AF, 120V/ 240V DH-620Q2-AA/A3 rtVAX, BA23, 11,853/ 7,311 80 XX Y L KA620, DEQNA, 11,433 BA23AR, 120V/ 240V DH-620Q2-CA/C3 rtVAX, BA23, $12,296/ $ 8,819 $ 99 XX Y L KA620, RQDX3, 11,876 RD31, RX33, BA23AR, 120V/ 240V DH-620Q5-AA/A3 rtVAX, H9642, 23,478/ 14,522 99 XX Y L BA23, KA620, 23,058 RQDX3, RD31, RX33, 120V/ 240V Industrial VAX 620 Unique DJ-620H1-A2/A3 INDUS VAX 620 11,897 6,963 85 XX Y L ELN NODISK, 120V/240V DJ-620H2-A2/A3 INDUS VAX 620 25,127 17,939 179 XX Y L ELN W DISK, 120V/240V KA620 Single Board Computer KA620-BA 32-BIT VAXELN 5,245 3,145 N/A XX Y L w/Lic. RTVAX REALTIME REPRICING - DISCOUNTS Systems o The terms and conditions will be consistent across the rtVAX 1000 and the IVAX620 systems. o The end-user discount code is type "Y," which is discountable under the SVA (Standard Volume Agreement); or curve "A" of the DBA (Digital Business Agreement). o The volume discount code is type "L" of the SVA; or curve "B" plus 15% product discount of the DBA. Single Board Computer o The terms and conditions will NOT change for the KA620 SBC; the end-user discount code will remain type "Y," which is discountable under the SVA; or curve "A" of the DBA. o The volume discount code will remain type "K," discountable under the SVA; or under the end-user volume personal computer product section. ================================================================================ Note 40.52 PERIPHERAL DEVICE ANNOUNCEMENTS 52 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 14 lines 1-AUG-1988 16:59 -< VT320 is now FASTship >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! Announcing FASTship for VT320s - 5-July-1988 ******************************************************************************* The VT320 North American models were recently added to Digital's FASTship Program. You should take advantage of VT320s on FASTship when you are faced with an unexpected customer need and quick delivery is important. Any product featured in the FASTship Program ships next day and can now be purchased by calling 800-DIGITAL, through the Electronic Store or through your Digital sales representative. ================================================================================ Note 40.53 PERIPHERAL DEVICE ANNOUNCEMENTS 53 of 56 EISNER::TINKELMAN "Bob Tinkelman - CCA" 194 lines 3-OCT-1988 08:50 -< Digital's Chip Family >- -------------------------------------------------------------------------------- The attached appeared in the DEC Electronic Store, summarizing DEC's existing family of chip products.  INNOVATION IN REALTIME CHIP ARCHITECTURE AND LOGIC DESIGN Committed to excellence in integrated circuit technologies, Digital provides innovative, high-performance products for the realtime computing environment. The realtime environment requires time-critical operations synchronized to external events or processes. Digital's Chip Family offers the necessary machine-to-machine interaction and predictability of response to external events for your critical applications. Digital provides an extensive line of products that can satisfy the specific needs of your realtime applications while they remain architecturally compatible with products for your entire organization. The Digital Chip Family includes microprocessors and support devices, communication devices, and bus support devices with the following features. HIGHLIGHTS . VLSI engineering . Fabricated in high-performance CMOS (complementary MOS) chip technology or custom-designed ZMOS (double-metal MOS) and NMOS (N-channel MOS) . DIP (dual-inline package), Cerquad (quadriform ceramic chip carrier), and PGA (pin grid array) packaging . Single 5-V dc power supply MICROPROCESSORS Digital microprocessors provide a low-cost means to implement the power and versatility of PDP-11 computers into system designs. DCT11 Microprocessor -------------------- The DCT11 microprocessor is a 40-pin, dual-line package device that contains the essential elements of the PDP-11 architecture. It is available in two versions with maximum clock frequency of either 7.5 MHz or 10 MHz. Full dynamic memory support is provided for 4K/16K or 64K chip memory. This includes timing strobes, dynamic address multiplexing, automatic refresh cycles, and refresh counter. The interrupt is multilevel, using four priority levels. . Basic PDP-11 instruction set (except for the MARK instruction) . 16-or 8-bit data paths selected at initialization . Interrupts on four priority levels with 15 internally generated vectors . Option of having the interrupt device provide the vector address . DMA arbitration . Full dynamic memory support-direct dynamic addressing, RAS and CAS strobes, refresh counter, automatic refresh cycles . Programmable mode register-8- or 16-bit external data bus, start and restart address selection, static or dynamic memory support, standard or long micro- cycles, bus synchronous or constant clock output COMMUNICATION DEVICES Digital's asynchronous communication devices enhance serial-line information transfers between local and remote systems and terminals. 78808 Eight-channel Asynchronous Receiver/Transmitter (Octal ART) The 78808 is a VLSI device for new generations of asynchronous serial communication designs and for microcomputer systems. This 68-pin cerquad device performs the basic operations necessary for simultaneous transmission and reception of asynchronous messages on eight independent lines. Programmable baud rates are individually selectable for each line's transmitter/receiver (50 to 19,2000 baud). DC319 DL11 Compatible Asynchronous Receiver/Transmitter (DLART) The DC319 allows data communication between Digital's microprocessors and console terminals or communication devices. Fabricated using N-channel MOS silicon technology, the DLART is contained in a 40-pin dual-line package that can be installed on a microprocessor module or interface module. It is programmed by the CPU to operate in either 8-bit or 16-bit mode with asynchronous baud rates varying from 300 to 38.4K. The DLART accepts data characters from the CPU in parallel format and converts them into an asynchronous serial-data stream for transmissions. It can simultaneously receive serial-data streams and convert them into parallel data characters for the CPU. Q-BUS INTERFACE PRODUCTS Digital provides a series of integrated circuits (ICs) and kits used in the development of device interfaces for the Q-bus. The ICs are available separately or as part of the DCK11 series of chipkits. The chipkits minimize the number of chips necessary to develop custom program control or direct memory access (DMA) interfaces. DC003 Dual-interrupt Logic Chip ------------------------------- The DC003 is an 18-pin, DIP bipolar device that is used to perform an interrupt transaction in a computer system that uses a daisychain type of arbitration. It includes tow interrupt channels, each of which can generate an interrupt request from a device that requires service. The DC003 circuits include enable logic and provide interrupt status information to the requesting device. DC004 Register Selector Logic Chip ---------------------------------- The DC004 is a 20-pin, DIP bipolar device that operates as a register selector to control the transfer of data to and from as many as four word registers (8-bytes). The Q-bus signals directly connect to high-impedance receivers and to high-current drivers on the DC004. A resistor and capacitor circuit is included to delay the response of the peripheral interface to the data transfers. DC005 4-bit Transceiver ----------------------- The DC005 is a 20-pin, DIP bipolar device that is used in interfaces as a bidirectional buffer between the Q-bus and a peripheral device logic bus. It includes a comparison circuit for device address selection and a constant generator for interrupt-vector address generation. The DC005 provides high-impedance inputs and high-drive, open-collector outputs to allow direct connection to a computer data bus structure. The bidirectional device port includes TTL inputs and three-state driver outputs. DC006 Word Count/Bus Address Logic Chip --------------------------------------- The DC006 is a 20-pin, DIP bipolar device that is designed for use in a direct memory access (DMA) device interface. The DC006 is a low-power Schottky device that connects to the three-state outputs of the DC005 transceiver. The DC006 is controlled by the DC004 register selector, the DC010 direct memory access, and ancillary logic. DC010 Direct Memory Access Logic Chip ------------------------------------- The DC010 is a 20-pin, DIP bipolar device that provides the logic to perform the protocol operations required to request and gain control of the Q-bus. Once bus mastership has been established, the DC010 generates the required signals to perform a data-in (DATI) transfer, data-out (DATO) transfer, or a data-in/data-out (DATIO) transfer as selected by the control lines to the DC010. Four words or multiple words can be transferred before control of the bus is relinquished. UNIBUS DEVICES The UNIBUS is an asynchronous bus used with the PDP-11 processors. The UNIBUS devices facilitate the development of the bus interfaces. DC013 UNIBUS Request Logic Chip ------------------------------- The DC013 is a 16-pin, DIP device that contains the logic required to perform interrupt bus requests and nonprocessor direct memory access requests to gain control of the UNIBUS. Input signals from the UNIBUS are received by high-impedance receivers on the DC013, and signals from the DC013 to the UNIBUS are supplied by high-current, open-collector driver outputs. The signal levels between the UNIBUS and the DC013 are compatible. The input and output signals between the device and DC013 are TTL levels. The DC013 circuits include bus grant logic, bus busy logic, and slave acknowledge logic. DC021 Octal Bus Transceiver --------------------------- The DC021 is a 20-pin, DIP device compatible with both the UNIBUS and the Q-bus. The DC021 provides eight bidirectional channels that transfer information between a wired-OR bus and a user interface. It provides high impedance receiver inputs and high-current, open-collector driver outputs. The DC021 transfers TTL-level signals between the device logic and the bus. The Select and Enable inputs to the DC021 are used to control the direction of information transfer. KEF11-AA -------- THE KEF11-AA is a single- and double-precision floating-point option for use with the KDF11-AA. The KEF11-AA performs hardware operations on 32-bit and 64-bit floating-point numbers, provides up to 17 digits of precision as well as integer-to-floating-point conversions. It has 40-pin DIP IC. The KEF11-AA mounts on the KDF11 CPU board and requires a KTF11. FPJ11-AA -------- The FPJ11-AA is a floating-point coprocessor option for use with the KDJ11-AB/AC modules. This single-chip option is designed to significantly improve the performance three or four times in computation-intensive applications. The FPJ11 is compatible with other PDP-11 systems with floating point. The 40-pin package is installed on the CPU board. ORDERING INFORMATION Additional information is available in the Realtime Systems and Options Catalog. Orders may be placed: - through this Electronic Store, or - by calling 1-800-DIGITAL (1-800-344-4825) or - with your Digital Sales Representative ================================================================================ Note 40.54 PERIPHERAL DEVICE ANNOUNCEMENTS 54 of 56 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 15 lines 4-OCT-1988 23:08 -< DEC still won't sell the good chips >- -------------------------------------------------------------------------------- >> DCT11 Microprocessor >> The DCT11 microprocessor is a 40-pin, dual-line package device that >> contains the essential elements of the PDP-11 architecture. It is >> available in two versions with maximum clock frequency of either 7.5 MHz >> or 10 MHz. Lest anyone be confused, this is NOT the J-11 chip of PDP-11/73 fame. The T-11 is a far cry from it. Anyone care to guess the odds of YOU being able to buy real J-11 chips (other than on the black market)? If your answer is greater than zero, you are wrong. Alan ================================================================================ Note 40.55 PERIPHERAL DEVICE ANNOUNCEMENTS 55 of 56 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 13 lines 5-OCT-1988 01:48 -< SURE YOU CAN BUT IT - YOU HAVE MY WORD ON IT - THATS THE TICKET >- -------------------------------------------------------------------------------- > Anyone care to guess the odds of YOU being able to buy real > J-11 chips (other than on the black market)? If your answer > is greater than zero, you are wrong. Sure you can - just ask DEC for a product demand quote and they will give you the part number to order.....  KDJ11-A% 8-( ================================================================================ Note 40.56 PERIPHERAL DEVICE ANNOUNCEMENTS 56 of 56 EISNER::KENNEDY "Terry Kennedy" 10 lines 5-OCT-1988 03:55 -< Even a data book, too! >- -------------------------------------------------------------------------------- > Anyone care to guess the odds of YOU being able to buy real > J-11 chips (other than on the black market)? If your answer > is greater than zero, you are wrong. As of the last time I looked, you could even get the full data sheet for the chip (I got one). The difference is that a 3rd-party hardwarde vendor's Product Demand Quote for 1000 pieces would probably be refused, as "Digital reserves the right to limit quantities..." (as stated on the PDQ form).  ================================================================================ Note 41.0 PDP-11 SYSTEM ANNOUCEMENTS 4 replies EISNER::KILLEEN "Jeff Killeen" 2 lines 5-SEP-1987 12:06 -------------------------------------------------------------------------------- This topic will be used to discuss PDP-11 hardware system announcements. PDP-11 CPU's, system boxes, and memories. ================================================================================ Note 41.1 PDP-11 SYSTEM ANNOUCEMENTS 1 of 4 EISNER::KILLEEN "Jeff Killeen" 257 lines 5-SEP-1987 14:20 -< MICRO PDP-11 PRODUCT FAMILY >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! MicroPDP-11 HARDWARE PRODUCT POSITIONING - 09-July-1987 The MicroPDP-11 family of 16-bit multi-user systems offers a range of price and performance levels to meet customer needs. Whether it be an entry-level three-user system or one for 40 users, or whether price is a factor, there are a variety of offerings to meet your customer's requirements. The MicroPDP-11/53 is the entry-level system, for those customers that are very price sensitive. As system performance increases, so do price and functionality. The MicroPDP-11/73 offers additional performance and packaging flexibility over the MicroPDP-11/53, while the MicroPDP-11/83 offers the maximum performance and storage capability in a Q-bus PDP-11 system. A key advantage within the PDP-11 family is that the user can start out on the entry-level system and grow it as their needs increase (more users, more storage, more performance), while protecting their investment in software and hardware. There are also numerous CPU upgrade kits available. Comparison Table of Q-bus MicroPDP-11 Systems Micro- Micro- Micro- Micro- PDP- PDP- PDP- PDP- 11/53 11/53+ 11/73 11/83 Typical No. 2-6 4-8 8-12 8-24 of users Price Range SBB ($K) N/A N/A $10.8-14.3 $13.4-22.2 SS ($K) $8.6-9.6 $14.5-17.3 18.6-19.0 24.7-50.1 Performance 0.5 0.6* 0.7 1.2 (PDP-11/70 = 1) Floppies 0.8 MB RX50------------------> 1.2 MB RX33-----| RX33------------------> Disks 20 MB RD31-----| 42 MB RD32--------------------------------------------------> 31 MB RD52-------------------> 71 MB RD53------------------------------------> 159 MB RD54-------------------> 456 MB RA81-------------------> Tape 95 MB TK50------------------------------------> Enclosures 8 slots BA23--------------------------------------------------> 12 slots BA123------------------> 16 slots H9642--> * Estimated performance -- will depend on the application's use of on-board memory. NOTE: All full-height 5-1/4" devices are available and supported as external options on all four systems. MicroPDP-11/53 PRICING INFORMATION MicroPDP-11/53 PLUS Systems -- FRS scheduled for early August 1987 Order No. Description MLP Seg. DH-153Q3-BA KDJ11-DB (1.5 MB), BA23 F/T, $17,800 XXX RQDX3, RD53, TK50/TQK50, DHQ11, CK-DHQ11-AB, U.S. Info. Kit, 120 V DH-153Q3-B2/B3 KDJ11-DB (1.5 MB), BA23 F/T, 17,500 XXX RQDX3, RD53, TK50/TQK50, DHQ11, CK-DHQ11-AB, 120 V/240 V DH-153Q4-BA KDJ11-DB (1.5 MB), BA23 RM, 17,650 XXX RQDX3, RD53, TK50/TQK50, DHQ11, CK-DHQ11-AB, U.S. Info. Kit, 120 V DH-153Q4-B2/B3 KDJ11-DB (1.5 MB), BA23 RM, 17,350 XXX RQDX3, RD53, TK50/TQK50, DHQ11, CK-DHQ11-AB, 120V/240V DH-153Q7-BA KDJ11-DB (1.5 MB), BA23 F/T, 14,900 XXX RQDX3, RD32, TK50/TQK50, U.S. Info. Kit, 120 V DH-153Q7-B2/B3 KDJ11-DB (1.5 MB), BA23 F/T, 14,600 XXX RQDX3, RD32, TK50/TQK50, 120 V/240 V DH-153Q8-BA KDJ11-DB (1.5 MB), BA23 RM, 14,800 XXX RQDX3, RD32, TK50/TQK50, U.S. Info. Kit, 120 V DH-153Q8-B2/B3 KDJ11-DB (1.5 MB), BA23 RM, 14,500 XXX RQDX3, RD32, TK50/TQK50, 120 V/240 V MicroPDP-11/53 Standard Systems -- Placed on Maintenance-Only Effective Immediately Order No. Description MLP BMC DH-153Q7-AA KDJ11-DA (.5 MB), MSV11-QA (1 MB), $14,900 $146 BA23 F/T, RQDX3, RD32, TK50/TQK50, U.S. Info. Kit, 120 V DH-153Q7-A2/A3 KDJ11-DA (.5 MB), MSV11-QA (1 MB), $14,600 $146 BA23 F/T, RQDX3, RD32, TK50/TQK50, 120 V/240 V DH-153Q8-AA KDJ11-DA (.5 MB), MSV11-QA (1 MB), 14,800 146 BA23 RM, RQDX3, RD32, TK50/TQK50, U.S. Info. Kit, 120 V DH-153Q8-A2/A3 KDJ11-DA (.5 MB), MSV11-QA (1 MB), 14,500 146 BA23 RM, RQDX3, RD32, TK50/TQK50, 120 V/240 V Existing MicroPDP-11/53 Standard Systems -- Price Reductions Old New Order No. Description MLP MLP DH-153Q1-AA KDJ11-DA (.5 MB), BA23 F/T, RQDX3, $ 9,520 $ 9,020 RD31, RX33, U.S. Info. Kit, 120 V DH-153Q1-A2/A3 KDJ11-DA (.5 MB), BA23 F/T, 9,220 8,720 RQDX3, RD31, RX33, 120/240 V DH-153Q2-AA KDJ11-DA (.5 MB), BA23 RM, RQDX3, 9,390 8,870 RD31, RX33, U.S. Info. Kit, 120 V DH-153Q2-A2/A3 KDJ11-DA (.5 MB), BA23 RM, RQDX3 9,090 8,570 RD31, RX33, 120/240 V DH-153Q5-AA KDJ11-DA (.5 MB), BA23 F/T, RQDX3, 10,580 10,080 RD32, RX33, U.S. Info. Kit, 120 V DH-153Q5-A2/A3 KDJ11-DA (.5 MB), BA23 F/T, RQDX3 10,280 9,780 RD32, RX33, 120/240 V DH-153Q6-AA KDJ11-DA (.5 MB), BA23 RM, RQDX3, 10,320 9,900 RD32, RX33, U.S. Info. Kit, 120 V DH-153Q6-A2/A3 KDJ11-DA (.5 MB), BA23 RM, RQDX3, 10,120 9,600 RD32, RX33, 120/240 V MicroPDP-11/83 PRICING INFORMATION Existing MicroPDP-11/83 Systems -- Price Reductions Comp* Old New Order No. Description MLP MLP BMC DH-183Q1-BA KDJ11-BF, 2MB, BA23 F/T, RQDX3, $27,740 $25,000 $207 TK50/TQK50, RD54, DHQ11, CK-DHQ11-AB, U.S. Info. Kit, 120 V DH-183Q1-B2/B3 KDJ11-BF, 2 MB, BA23 F/T, RQDX3, 27,430 24,700 207 RD54, DHQ11, CK-DHQ11-AB, TK50/TQK50, 120/240 V DH-183Q2-CA KDJ11-BF, 2 MB, BA123, RD54, RQDX3, 33,950 31,600 232 (2) DHQ11, (2) CK-DHQ11-AA, TK50/TQK50, U.S. Info. Kit, 120 V DH-183Q2-C2/C3 KDJ11-BF, 2 MB, BA123, RD54, RQDX3 33,640 31,300 232 (2) DHQ11, (2) CK-DHQ11-AA, TK50/TQK50, 120/240 V DH-183Q3-BA DJ11-BF, 2 MB, H9642, (2) BA23, 52,700 50,100 304 RA81, KDA50, TK50/TQK50, (2) DHQ11-M, (2) CK-DHQ11-AF, U.S. Info. Kit, 120 V DH-183Q3-B2/B3 KDJ11-BF, 2 MB, H9642, (2) BA23, 52,400 49,800 304 RA81, KDA50, TK50/TQK50, (2) DHQ11-M, (2) CK-DHQ11-AF, 120/240 V 183QY-D2/D3 KDJ11-BF, 2 MB, BA23 F/T, 120/240 V 15,890 13,500 183QZ-D2/D3 KDJ11-BF, 2 MB, BA23 RM, 120/240 V 15,730 13,400 183QB-D2/D3 KDJ11-BF, 2 MB, BA123, 120/240 V 19,360 17,040 183QE-D2/D3 KDJ11-BF, 2 MB, H9642, (2) BA23, 23,390 22,200 120/240 V * Comp Old MLP = Current MLP of standard systems containing DHV11 All systems are E/U Discount Y, OEM Discount Type 1, Seg. Code XXX MicroPDP-11/83 Standard Systems -- Placed on Maintenance-Only Effective September 28, 1987 DH-183Q1-AA/A2/A3 DH-183Q2-BA/B2/B3 DH-183Q3-AA/A2/A3 KDJ11-BF, 2 MB, BA23 F/T, KDJ11-BF, 2 MB, BA123, KDJ11-BF, 2 MB, H9642, RQDX3, TK50, RD54, DHV11 RQDX3, TK50, RD54, (2) BA23, KDA50, RA81, (2) DHV11 TK50, (2) DHV11 CLASS L LICENSE PRICING FOR MicroPDP-11/83 SYSTEMS Product Description Old MLP New MLP QY950-UZ A-to-Z Base System (Class L) $1,400 $1,500 QY821-UZ DSM-11 (Class L) 800 1,300 QY829-UZ Micro/RSTS (Class L) 800 1,300 QY800-UZ Micro/RSX (Class L) 800 900 QR430-UZ RSTS/E (Class L) 800 1,300 QY628-UZ RSX-11M (Class L) 800 1,300 QY505-UZ RSX-11M-PLUS (Class L) 800 1,300 QY013-UZ RT-11 (Class L) 500 600 QJ013-UZ RT-11 (Class H) 1,000 1,200 QJV13-DZ RT-11 (Run-Time) 1,850 3,000 MicroPDP-11/73 PRICING INFORMATION MicroPDP-11/73 Standard Systems (with DHQ11 and Cabinet Kit) Model No. Description MLP DH-173Q1-CA KDJ11-BB, 1 MB, BA23 F/T, RQDX3, $18,990 RD53, TK50/TQK50, DHQ11-M, CK-DHQ11-AB, U.S. Info. Kit, 120 V DH-173Q1-C2 KDJ11-BB, 1 MB, BA23 F/T, RQDX3, $18,680 RD53, TK50/TQK50, DHQ11-M, CK-DHQ11-AB, 120 V DH-173Q1-C3 KDJ11-BB, 1 MB, BA23 F/T, RQDX3, $18,680 RD53, TK50/TQK50, DHQ11-M, CK-DHQ11-AB, 240 V TK50 Boot ROM Kit Order No. Description MLP BMC Inst. KDJ11-U1 TK50 boot ROM for KDJ11-BB $100 N/A Per call CPU module ================================================================================ Note 41.2 PDP-11 SYSTEM ANNOUCEMENTS 2 of 4 EISNER::KILLEEN "Jeff Killeen" 238 lines 5-SEP-1987 14:21 -< PDP-11/84 E SERIES >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! THE PDP-11/84 E SERIES - 06-July-1987 A little more than two years ago Digital introduced a new high-end member of the PDP-11/84 family -- the PDP-11/84 A series. Our customers responded enthusiastically and sales have continued to exceed all expectations. Today we are introducing the new, enhanced PDP-11/84 E series product set. This replacement for the existing PDP-11/84 A series consists of: o Two new OEM designs -- 5.25" and 10.5" o Two new kernel systems and SBB configurations o Upgrade Packages based on the new OEM design centers The reliability of the PDP-11/84 also allows us to offer reduced monthly service charges on both the new E series and old A series products. Our expectations are that the new PDP-11/84 product will enable you and our customers to build upon the PDP-11/84 success story. The new product set provides a unique opportunity to promote UNIBUS PDP-11 designs and investments while extending the breadth of PDP-11/84 applicability to an even wider range of installed base market opportunities. The existing A series product has been repriced, effective immediately, and will move to Maintenance-Only status, effective Q2/FY88. Current orders should be reviewed with customers and your Operations contact. New orders should be placed as soon as possible. Your Operations contact is the focal point for information on new products, pricing, availability, allocation and delivery; working with him/her will ensure a smooth transition process. PRODUCT DESCRIPTION The new, enhanced PDP-11/84 E series products are built around the PDP-11/84 board set (KDJ11-BF), its companion high-performance floating point co-processor (FPJ11), either 2 or 4 Mbytes of ECC (Error Correction Code) (PMI) memory (MSV11-J), and a UNIBUS adapter (KTJ11-B). These base system elements are now configured in a new, 9-slot processor backplane and housed in a set of new OEM box-level enclosures and system cabinets which offer an assortment of advantages over earlier UNIBUS PDP-11 products. In addition, the new PDP-11/84 systems are form, fit and function compatible with virtually any previous UNIBUS PDP-11 design. 5.25" OEM Design Center -- 11/84-D(x) o Standard 5.25" H X 19" W rack-mount enclosure o PDP-11/84 module set (2 or 4 Mbytes) o New 9-slot processor backplane o 6 System units (SU) (one 9-slot backplane capacity) o 5 Slots for system options o 650-Watt power supply (400 watts dc) 10.5" OEM Design Center -- 11/84-E(x) o Standard 10.5" H X 19" W rack-mount enclosure o PDP-11/84 module set (2 or 4 Mbytes) o New 9-slot processor backplane o 6 System units (SU) (three 9-slot backplane capacity) o Up to 23 slots for system options o DD11-DK or DD11-CK compatible o 1,100-Watt power supply (768 watts dc) NOTE: All kernel systems and system building block configurations are based on the 10.5" design center. Single Bay Kernel System and SBB Configurations -- 11X84-E(x) and SX-JX200-E(x) o PDP-11/84 module set (2 or 4 Mbytes) o 10.5" PDP-11/84 design center o 27-Slot expansion capacity (optional) o DD11-CK or DD11-DK compatible o 1,100-Watt power supply (768 watts dc) o One 10.5" H X 19" W rack-mount option space o FCC shielded I/O bulkhead Wide Body Kernel System and SBB Configurations -- 11W84-E(x) and SK-JX300-E(x) o PDP-11/84 module set (2 or 4 Mbytes) o 10.5" PDP-11/84 design center o 27-Slot expansion capacity (optional) o DD11-CK or DD11-CK compatible o 1,100-Watt power supply (768 watts dc) o Two 10.5" H X 19" W rack-mount option spaces o FCC shielded I/O bulkhead NOTE: Both of the above SBB configurations include a choice of PDP-11 operating system (-UZ) license. These new PDP-11/84 product variations are all comparably priced to the existing PDP-11/84 designs with FCS planned for Q1/FY88. All current PDP-11/84 variations are targeted for retirement (to Maintenance-Only status) at the start of Q2/FY88. In order to ensure a timely transition to the new product set, all current PDP-11/84 variations have been increased in price by approximately 10%, effective immediately. Refer to "MicroPDP-11/84 Pricing Information" for current and newly announced product prices. SYSTEM CONFIGURATIONS AND SUPPORTED OPTIONS The new PDP-11/84 E series kernel systems and system building block configurations afford customers enhancements along four primary dimensions -- improved packaging density, configuration flexibility, expansion, and growth capacity and economy. The new base systems packages are all built around the new 10.5" box product with its 27-slot backplane expansion capacity. The system packages -- 11X84, 11W84 and 11Y84* -- offer one, two or three 10.5" X 19" cavities for rack-mount system option expansion. As such, more options can be configured with greater flexibility, a smaller footprint, and for less cost than ever before. Naturally, the new PDP-11/84 supports all of the device options shipping and/or supported by the current product set. PDP-11/84 plans include the system integration of a selection of appropriate new device options, such as the SA482, as they are announced and become available in volume. At that time, we have not incorporated any of these device options into the standard systems. Currently, our intent is to maintain a streamlined product set of kernel systems and system building blocks (with software) that can be easily configured by selecting from the menu of PDP-11/84 supported options. For detailed configuration information, please consult the Q1/FY88 edition of the PDP-11 Systems and Options Catalog. UNIBUS INSTALLED BASE PDP-11/84 UPGRADE PROGRAM In addition to the new base PDP-11/84 product set, an assortment of special upgrade packages have been defined. These 11/84-U(x) variations are based on standard box and systems level products plus service and applicable software license upgrades. Refer to "PDP-11/84 Upgrade Pricing Information" for pricing details. PDP-11/84 PRICING INFORMATION Model No. Description MLP BMC Systems: 11/84-DC/DD PDP-11/84E, 2 MB, 5.25" box, 120/240 V $16,000 $121 11/84-DE/DF PDP-11/84E, 4 MB, 5.25" box, 120/240 V 18,000 153 11/84-EC/ED PDP-11/84E, 2 MB, 10.5" box, 120/240 V 20,000 136 11/84-EE/EF PDP-11/84E, 4 MB, 10.5" box, 120/240 V 22,000 168 11X84-EC/ED PDP-11/84E, 2 MB, 10.5" box, H9642 23,200 136 cabinet, 120/240 V 11X84-EE/EF PDP-11/84E, 4 MB, 10.5" box, H9642 25,200 168 cabinet, 120/240 V 11W84-EC/ED PDD-11/84E, 2 MB, 10.5" box, H9645 23,500 136 cabinet, 120/240 V 11W84-EE/EF PDP-11/84E, 4 MB, 10.5" box, H9645 25,500 168 cabinet, 120/240 V Licenses: SK-JX200-EC/ED 11X84-EC/ED, 2 MB, PDP-11 O/S $25,200 $136 license choice SK-JX200-EE/EF 11X84-EE/EF, 4 MB, PDP-11 O/S 27,200 168 license choice SK-JX300-EC/ED 11W84-EC/ED, 2 MB, PDP-11 O/S 25,500 136 license choice SK-JX300-EE/EF 11W84-EE/EF, 4 MB, PDP-11 O/S 27,500 168 license choice All systems are E/U Discount Y, OEM Discount 1 PDP-11/84 UPGRADES PRICING INFORMATION PDP-11/84 Upgrades -- FRS is scheduled for September Model No. Description MLP BMC 11/84-UD/UE 5.25" 11/84 upgrade package, box $ 5,200 $ 25 and backplane only, 120/240 V (11/84-DX/DY) 11/84-UF/UG 10.5" 11/84 upgrade package, box $ 6,800 $ 40 and backplane only, 120/240 V (11/84-EX/EY) 11/84-UH/UJ 11/84-DC/DD, 2 MB, 5.25" box upgrade 18,200 121 package, 120/240 V 11X84-UK/UL 11/84-EC/ED, 2 MB, 10.5" box upgrade 22,300 136 package, 120/240 V H9642-EX/EY H9642 single bay UNIBUS system 3,200 N/C cabinet, 120/240 V H9645-EX/EY H9645 wide UNIBUS system cabinet, 3,500 N/C 120/240 V All systems are E/U Discount Y, OEM Discount 1 REPRICING FOR EXISTING PDP-11/84 VARIATIONS Old New New Model No. Description MLP MLP BMC 11/84-AA/AB 11/84-A, 1 MB, 10.5" box, 120/240 V $18,630 $20,493 $120 11/84-BA/BB 11/84-A, 2 MB, 10.5" box, 120/240 V 20,700 22,560 136 11/84-CA/CB 11/84-A, 4 MB, 10.5" box, 120/240 V 22,700 24,560 168 11X84-AA/AB 11X84-A, 1 MB, 42" system cabinet, 23,805 26,070 160 120/240 V 11X84-BA/BB 11X84-A, 2 MB, 42" system cabinet, 25,875 28,140 176 120 V 11X84-CA/CB 11X84-A, 4 MB, 42" system cabinet, 27,945 30,140 208 120 V SK-JX100-EX/EY 11/84, 2 MB, DHU11, H9642, 120/240 V 29,498 31,760 221 SK-JXNNA-EX/EY SK-JX100-EX, dual RC25, 120/240 V 51,233 53,500 420 SK-JXEDA-EX/EY SK-JX100-EX, RA81, TU80, 120 V 78,660 67,590 458 SK-JX100-EY, SK-JX100-EY, RA81, TU81, 120 V 65,205 80,925 509 All systems are E/U Discount Y, OEM Discount 1, Seg. Code XXX ================================================================================ Note 41.3 PDP-11 SYSTEM ANNOUCEMENTS 3 of 4 EISNER::KILLEEN "Jeff Killeen" 1 line 5-SEP-1987 23:33 -< WARNING! >- -------------------------------------------------------------------------------- WARNING these PDP-11 prices are out of date. See note 37.1 ================================================================================ Note 41.4 PDP-11 SYSTEM ANNOUCEMENTS 4 of 4 EISNER::TINKELMAN "Bob Tinkelman - CCA" 61 lines 3-OCT-1988 09:00 -< Replacement part numbers for PDP-11/84-A series >- -------------------------------------------------------------------------------- The attached table appeared in the DEC Electronic Store. It summarizes replacement-part-numbers for those PDP11/84-A products moved to "maintenance only" in December, 1987.  DESCRIPTION This document was created to help you find current part numbers for PDP-11/84 products. We hope this list of Obsolete Part Numbers and Recommended Replacements will assist you. o The PDP-11/84 A series moved to Maintenance-Only status in December 1987 o Orders for the PDP-11/84 A series have NOT been accepted since 12/4/87. Moved to Maintenance-Only on Recommended PDP-1/84 E series December 24, 1987 Replacements 11/84-AA 1 MB, 10.5" box, 120V 11/84-EC, 11/84-DC 11/84-AB 1 MB, 10.5" box, 240V 11/84-ED, 11/84-DD 11/84-BA 2 MB, 10.5" box, 120V 11/84-EC, 11/84-DC 11/84-BB 2 MB, 10.5" box, 240V 11/84-ED, 11/84-DD 11/84-CA 4 MB, 10.5" box, 120V 11/84-EF, 11/84-DE 11/84-CB 4 MB, 10.5" box, 240V 11/84-EF, 11/84-DF 11X84-BB 1 MB, 42" cab, 120V 11X84-EC, 11W84-EC, 11Y84-EC 11X84-CA 1 MB, 42" cab, 240V 11X84-ED, 11W84-ED, 11Y84-ED 11X84-CB 2 MB, 42" cab, 120V 11X84-EC, 11W84-EC, 11Y84-EC 11X84-AA 2 MB, 42" cab, 240V 11X84-ED, 11W84-ED, 11Y84-EC 11X84-AB 4 MB, 42" cab, 120V 11X84-EE, 11W84-EE, 11Y84-EE 11X84-BA 4 MB, 42" cab, 240V 11X84-EF, 11W84-EF, 11Y84-EF SK-JX100-EY SBB, 120V SX-JX200-EX, SX-JX300-EX, SX-JX400-EX SK-JX100-EX SBB, 240V SX-JX200-EY, SX-JX300-EY, SX-JX400-EY SK-JXNNA-EX RC25 system, 120V SX-JX200-EX, SX-JX300-EX, SX-JX400-EX SK-JXNNA-EY RC25 system, 240V SX-JX200-EX, SX-JX300-EX, SX-JX400-EY SK-JXEDA-EX TU80 system, 120V SX-JX200-EX, SX-JX300-EX, SX-JX400-EX SK-JXEDA-EY TU80 system, 240V SX-JX200-EX, SX-JX300-EX, SX-JX400-EY SK-JXEEA-EX TU81 system, 120V SX-JX200-EX, SX-JX300-EX, SX-JX400-EX SK-JXEEA-EY TU81 system, 240V SX-JX200-EX, SX-JX300-EX, SX-JX400-EY 11/84-UA box upgrade, 120V 11/84-UK, 11/84-UH 11/84-UB box upgrade, 240V 11/84-UL, 11/84-UJ 11X84-UA cab upgrade, 120V 11/84-UK & H9642-EA or H9645-EA or H9647-EX 11X84-UB cab upgrade, 240V 11/84-UL & H9642-EA or H9645-EA or H9647-EX PDP-11/84 PRODUCT INFORMATION For additional technical information: - refer to the Rapid Lookup section on the Electronic Store - refer to Digital's PDP Systems and Options Catalog - contact your local Digital Sales Representative or Distributor Orders can be placed: - by calling 1-800-DIGITAL (1-800-344-4825) - with your Digital Sales Representative or Distributor  ================================================================================ Note 42.0 MICROVAX SYSTEM ANNOUCEMENTS 17 replies EISNER::KILLEEN "Jeff Killeen" 2 lines 5-SEP-1987 12:11 -------------------------------------------------------------------------------- This topic will be used to discuss MicroVAX hardware systems annoucements. MicroVAX CPU's, system units, and memories. ================================================================================ Note 42.1 MICROVAX SYSTEM ANNOUCEMENTS 1 of 17 EISNER::KILLEEN "Jeff Killeen" 66 lines 5-SEP-1987 12:32 -< NEW MVII PACKAGES >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! ENTRY-LEVEL MicroVAX II CONFIGURATION ADDS MORE STORAGE - 30-July-1987 o BA23/RD54-based standard system variant o Available immediately o Retirement of older MicroVAX II standard systems MSD announces a new variation of the MicroVAX II entry-level stand-alone (Configuration 2) standard system. This new standard system is an addition, versus a replacement, to the MicroVAX II family. We will continue to offer the existing DH-630Q2-F* system, along with this new, larger, DH-630Q2-H* system. PRICING/ORDERING INFORMATION The configuration menu for this new standard system will be published in the Q2 Systems and Options Catalog. With the exception of the RD53 having been replaced by the RD54, the new DH-630Q2-H* menu is identical to the existing DH-630Q2-F* menu, and may be used until the new menu is added to the SOC. Model No. Description Seg. MLP BMC DH-630Q2-HA MicroVAX II, BA23/F, MS630-BB, XXX $24,600 $273 RQDX3, RD54, TQK/TK50, DZQ11, Doc/Diag, 120 V DH-630Q2-H2 MicroVAX II, BA23/F, MS630-BB, XXX 24,300 273 RQDX3, RD54, TQK/TK50, DZQ11, 120 V Model No. Description Seg. MLP BMC DH-630Q2-H3 MicroVAX II, BA23/F, MS630-BB, XXX $24,300 $273 RQDX3, RD54, TQK/TK50, DZQ11, 240 V AVAILABILITY The DH-630Q2-H* standard systems are available immediately. RETIREMENT OF OLDER STANDARD SYSTEM VARIANTS The following standard system variations will be moved to Maintenance Only, effective August 14, 1987: Model No. Description DH-630Q1-DA MicroVAX II, 2 MB, BA23F, RQDX3, RD52, RX50, DEQNA, ZNAAA, 120 V DH-630Q1-D2 MicroVAX II, 2 MB, BA23F, RQDX3, RD52, RX50, DEQNA, 120 V DH-630Q1-D3 MicroVAX II, 2 MB, BA23F, RQDX3, RD52, RX50, DEQNA, 240 V (replaced by DH-630Q1-E* - with 5 MB and RD53 - in Q2/FY87) DH-630Q4-FA MicroVAX II, BA123, 9 MB, RQDX3, 3xRD53, TK50, DHV, DEQNA, ZNAAA, 120 V DH-630Q4-F2 MicroVAX II, BA123, 9 MB, RQDX3, 3xRD53, TK50, DHV, DEQNA, 120 V DH-630Q4-F3 MicroVAX II, BA123, 9 MB, RQDX3, 3xRD53, TK50, DHV, DEQNA, 240 V (replaced by DH-630Q4-H* - with 2xRD54 - in Q1/FY87) ================================================================================ Note 42.2 MICROVAX SYSTEM ANNOUCEMENTS 2 of 17 EISNER::KILLEEN "Jeff Killeen" 203 lines 5-SEP-1987 12:34 -< NEW MV2000 PACKAGES >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! MicroVAX 2000 CONFIGURATIONS AND PRICING - 30-July-1987 o New MicroVAX 2000 and associated VMS pricing o New MicroVAX 2000 RD54 standard system and disk options The following charts highlight examples of the price and functionality changes announced today. System Description Old MLP New MLP DH-625N1, 4 MB, RD32, RX33, BA40A, U.S. $11,100 $10,550 Plus: VMS 1-4 user license DH-625N3, 4 MB, RD53, BA40A, U.S. 18,200 14,800 Plus: TK50 and VMS 1-4 user license DH-625N6, 6 MB, RD54, BA40A, U.S. 23,050* 19,000 Plus: TK50 and VMS 1-4 user license Feature Old New MAX DISK CAPACITY: 142 MB (2 RD53s) 318 MB (2 RD54s) MAX MEMORY: 6 MB 6 MB MAX SYNCHRONOUS LINES: 1** 1** MAX VMS USERS: 16 Unlimited TERMINAL SERVER SUPPORT: Yes Yes MAX SERIAL LINES 4 4 Notes: *The RD54-based N6 configuration is new. This number is based on what it would have cost under the old pricing structure. NEW PRICING RATIONALE This new pricing announcement will make the MicroVAX 2000 more competitive in two key market areas: large project business and PC Local Area Network (LAN) server markets. The large project business typically involves sales of small quantities of VAX 8XXX systems, larger numbers of MicroVAX II systems, and still larger quantities of MicroVAX 2000 systems to large businesses with many small branch offices. In these situations, the price of the remote system, i.e., the MicroVAX 2000, is often the key factor in determining who gets the entire business. More competitive MicroVAX 2000 entry-level prices improve Digital's position in these large project opportunities. The PC LAN server market involves sales of MicroVAX 2000 systems with software to service personal computers. This market is growing at an impressive rate, and lower MicroVAX 2000 entry prices improve Digital's ability to compete more effectively in this market. The new pricing is in response to the unprecedented level of interest our customers are indicating in small, low-cost, networked, VAX/VMS systems. This pricing move is a signal that manufacturing volume shipment is underway and that Digital intends to be aggressively competitive in the low-end VAX business. The MicroVAX 2000 has captured the imagination of a variety of new customers who are solving their business problems with the sophisticated capabilities of this system. The packaging announced in February has been the impetus for many large distributed corporate buying decisions in favor of Digital. The new prices announced today are intended to make the MicroVAX 2000 the product of choice for the smallest department, remote office, or as incremental additions to a Local Area VAXcluster system. NEW PRICING The following price adjustments restore pricing consistency with both MicroVAX II VMS pricing and common VAXstation 2000 components. The new pricing results in lower total system (hardware and operating system) prices for RD32/RX33 systems with up to four users and RD53 systems with up to eight users, and higher prices for systems that can accommodate more users. The hardware prices of the RD32/RX33-based MicroVAX 2000 systems are reduced by $1,550 and the RD53 systems by $4,400. The price of the diskless Local Area VAXcluster system is reduced by $820. The expansion adapter, previously offered as a $1,200 option for the diskless and RD32/RX33 systems, is now included in all MicroVAX 2000 systems at no additional cost. The VMS software license prices have been increased $1,000 for the four-user license, $2,000 for the eight-user license, and $7,000 for the sixteen-user license. A new unlimited user license is being introduced at $18,000. STANDARD EXPANSION ADAPTER ADVANTAGES All MicroVAX 2000 offerings now include an expansion adapter as standard. The inclusion of the expansion adapter on the RD32/RX33 systems as a standard feature offers immediate advantages. For instance, the expansion adapter is ideal for customers with multiple MicroVAX 2000 systems that require periodic low-cost updates. The RX33 is adequate for low volume backup, but when higher volume backup is needed, the TK50 in the expansion box can be used. Prepackaged as a self-contained mobile unit, the TK50Z-FA is an ideal load/backup device. With the expansion adapter as a standard feature, the customer has this capability from day one. RD54 The RD54 is available in both a standard system and as a separate add-on disk drive. The standard system includes 6 MB of memory and an expansion adapter. The add-on option is in the same enclosure as the existing RD53 add-on option. STAND-ALONE VERSUS NETWORK TERMINAL EXPANSION The DECserver 200 provides terminal expansion, in multiples of eight, for networked systems. This solution requires the DESVA ThinWire Ethernet option, DECserver 200 hardware and software, and associated license, all available today. PRICING/ORDERING INFORMATION The new pricing becomes effective July 27, 1987 for all hardware components, and for the Q3001, 2 and 3 VMS options. MicroVAX 2000: Model No. Description Seg. MLP BMC DH-625N1-BA MicroVAX 2000,4MB,RD32,RX33,BA40A,US XX $ 7,550 * DH-625N1-B2 MicroVAX 2000,4MB,RD32,RX33,BA40A,120V XX 7,450 * DH-625N1-B3 MicroVAX 2000,4MB,RD32,RX33,BA40A,240V XX 7,550 * Model No. Description Seg. MLP BMC DH-625N2-BA MicroVAX 2000,6MB,RD32,RX33,BA40A,US XX $ 9,550 * DH-625N2-B2 MicroVAX 2000,6MB,RD32,RX33,BA40A,120V XX 9,450 * DH-625N2-B3 MicroVAX 2000,6MB,RD32,RX33,BA40A,240V XX 9,550 * DH-625N3-AA MicroVAX 2000,4MB,RD53,BA40A,US XX 7,300 * DH-625N3-A2 MicroVAX 2000,4MB,RD53,BA40A,120V XX 7,200 * DH-625N3-A3 MicroVAX 2000,4MB,RD53,BA40A,240V XX 7,300 * DH-625N4-AA MicroVAX 2000,6MB,RD53,BA40A,US XX 9,300 * DH-625N4-A2 MicroVAX 2000,6MB,RD53,BA40A,120V XX 9,200 * DH-625N4-A3 MicroVAX 2000,6MB,RD53,BA40A,240V XX 9,300 * DH-625N6-AA MicroVAX 2000,6MB,RD54,BA40A,US XX 11,500 * DH-625N6-A2 MicroVAX 2000,6MB,RD54,BA40A,120V XX 11,400 * DH-625N6-A3 MicroVAX 2000,6MB,RD54,BA40A,240V XX 11,500 * SV-PXXGB-EK MV2000,6MB,BA40A,NI,VMS,DECnet,LAVc XX 12,080 * SV-PXXGB-FN MV2000,6MB,BA40A,NI,VMS,DECnet,LAVc XX 12,080 * Q3001-C3 VMS 1-4 USER LICENSE RX50 W/W 076 3,000 N/A Q3001-C5 VMS 1-4 USER LICENSE TK50 W/W 076 3,000 N/A Q3002-C3 VMS 1-8 USER LICENSE RX50 W/W 076 6,000 N/A Q3002-C5 VMS 1-8 USER LICENSE TK50 W/W 076 6,000 N/A Q3003-C3 VMS 1-16 USER LICENSE RX50 W/W 076 13,000 N/A Q3003-C5 VMS 1-16 USER LICENSE TK50 W/W 076 13,000 N/A Q3004-C3 VMS 1-N USER LICENSE RX50 W/W 076 18,000 N/A Q3004-C5 VMS 1-N USER LICENSE TK50 W/W 076 18,000 N/A Q3002-CZ VMS 5-8 USER UPGRADE LIC 076 3,250 N/A Q3003-T3/T5 VMS 9-16 USER UPGRADE LIC 076 7,250 N/A Q3004-T3/T5 VMS 17-N USER UPGRADE LIC 076 5,250 N/A Previously Announced with VAXstation 2000 Systems: RD54-EA 159 MB DISK XX $ 6,900 63 RD54-FA 159 MB DISK, BOX, 120 V XX 8,150 63 RD54-F3 159 MB DISK, BOX, 240 V XX 8,150 63 Notes: *Contact Field Service for BMCs -Bx variants of DH-625N1,N2 include the BA40A expansion adapter. SV-PXXGB includes the BA40A expansion adapter. AVAILABILITY Product(s) Availability o All variants of Q3004 VMS Option o Early Q2/FY88 o New RD54 standard systems and o August 1987 add-on options o Expansion adapter for diskless o August 1987 and RD32/RX33 systems CONFIGURATION MENUS Configuration menus for the new RD54 standard systems and options will be published in the Q2 Systems and Options Catalog (SOC). They will be identical to the existing RD53-based MicroVAX 2000 systems, except that the RD54 will replace the RD53. ================================================================================ Note 42.3 MICROVAX SYSTEM ANNOUCEMENTS 3 of 17 EISNER::KILLEEN "Jeff Killeen" 407 lines 11-SEP-1987 00:29 -< MICRO VAX 3000'S >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! System Announcements - 10-September-1987 ANNOUNCING: MicroVAX 3500 AND MicroVAX 3600 TIMESHARING SYSTEMS AND VAXserver 3500, VAXserver 3600 AND VAXserver 3602 SERVERS o New CMOS microprocessor technology offers more than 3X the performance of the MicroVAX II o Extends the price/performance leadership of the VAX family o Available as timesharing and server systems in both cabinets and pedestals o Also features the new RA70 and TK70 5.25" mass storage devices, and the RA82 Micro Systems Development announces three new standard system packages for general-purpose timesharing use and three new server configurations for operation as a boot node on a Local Area VAXcluster or ULTRIX-32 Local Area system. These Q-bus systems are positioned above the current MicroVAX II systems in both price and functionality and therefore will be sold as the higher end complement of the MicroVAX family. Applications written for MicroVAX II will now run more than three times faster with no software modifications required. Individual applications performance will vary in a range from 2.6 to 4.2 times previous MicroVAX II performance. Check related applications marketing articles and Competitive Update articles for more information on performance. These new high-end complements to the MicroVAX II-based systems offer attractive price/performance which will open new sales opportunities for more rapid distribution of applications which formerly could be run only on expensive computer room systems. FEATURES BENEFITS More than 3X performance of Supports a new class of compute- MicroVAX II intensive applications at low-end Supports 32 MB memory -- Many applications run faster, less 2X more than MicroVAX II dependent on disk accessing ECC main memory Greater data integrity BA213 enclosure Acoustically quiet, rugged, simplified cable management FEATURES BENEFITS New RA70 5.25" disk drive and Offer significant mass storage TK70 cartridge tape -- plus capacity and performance enhancements 622 MB RA82 available PCSA/PC software VMS configurations are instantly compatible with most networks of IBM PCs NEW TECHNOLOGY AND PACKAGING New CPU These new systems all use a new CPU based on Digital's new proprietary CMOS microprocessor, the 78034, and the CVAX 78134 floating point unit. The performance improvements over MicroVAX II result from the increased operating speed of the CVAX microprocessor (90 ns) plus a two-level, write-through caching architecture. In addition to a 1 KB cache on the CVAX chip, there is a second-level 64 KB cache (operating at 180 ns) on the KA650-AA CPU board. The CPU board also includes custom VLSI CMOS chips for Q-bus interface, memory control, and auxiliary functions (such as console support). It is important to note that within the CVAX microprocessor, the distribution of the VAX instruction set is nearly identical to that of the MicroVAX CPU -- except six important commercial string instructions have been moved to microcode: CMPC3, CMPC5, LOCC, SKPC, SCANC and SPANC. Nine infrequently used floating point instructions were moved to software: EMODf, AACBf, and POLYf. This minor change in the instruction mix combined with the increased speed of the CVAX microprocessor will enhance the system-level performance of many commercial applications relative to MicroVAX II. The "simple" performance message is that the MicroVAX 3xxx-based systems offer more than three times the performance of MicroVAX II. Of course, it is impossible to fully characterize the performance of any machine with a single number. Over 200 different benchmarks and application programs, written in FORTRAN, C, COBOL, Macro, LISP, and OPS5 and spread broadly across all technical and commercial application areas have shown that the MicroVAX 3xxx-based systems provide 2.6 to 4.2 times the performance of MicroVAX II (with "outlying" benchmarks as low as 2.2X and as high as 5.7X). New Memory Unlike MicroVAX II, there is no main memory on the MicroVAX 3xxx-based system's CPU board. These new systems use the MicroVAX II local memory architecture, and support up to four 8 MB memory modules for a total of 32 MB ECC main memory. The 8 MB MS650 memory module uses 256K ZIP dynamic RAMs. All memory controlling operations are on a special VLSI memory controller resident on the KA650 CPU module. The memory module contains special custom data transceivers (XMX chips) that provide for communication between the array and the controller. The MS650-Ax is offered as the main memory for the MicroVAX 3500/3600, the VAXstation 3200/3500, and the VAXserver 3500/3600/3602 models. Maximum capacity is 32 MB (except the VAXstation 3200, which has 16 MB capacity). MEMORY SIZE Field-Installed System Memory Content Max. Capacity Add-On Memory Option MicroVAX 3600 32 MBytes 32 MBytes N/A MicroVAX 3500 16 MBytes 32 MBytes MS650-AF VAXserver 3500/3600/3602 16 MBytes 32 MBytes MS650-AF VAXstation 3500 16 MBytes 32 MBytes MS650-AF VAXstation 3200 8 MBytes 16 MBytes MS650-AA All Digital memory has a one-year warranty. Additionally, during succeeding years, the Basic Monthly Charge (BMC) for memory is zero ($0) when purchased and installed on a system covered by a Digital Field Service contract. BA213 Enclosure All of the systems being announced in this issue, except the VAXstation 3200, use the BA213 enclosure, which was first announced with the Industrial VAX and Industrial PDP-11 systems (Sales Update Special Issue dated March 24, 1987). The enclosure features 12 module slots and a mass storage cage that supports three 5.25-inch mass storage devices: the TK70 cartridge tape and up to two RA70 disk drives. The BA213 modules feature connections through a module handle rather than through internal cabling to a back insert panel, as in the BA23 and BA123. The result is easier cable management and system maintainability. (For more details, see the article on system packaging in this issue). All MicroVAX 3500 systems are based on the BA213 enclosure. All MicroVAX 3600 configurations come in one corporate cabinet, 40-inch high with a rack-mounted BA213 and room for one additional 14-inch Winchester (RA82). Additional 40-inch high cabinets are available as expander cabinets to house additional combinations of TU81-Plus and RA82s. A special cabinet housing three RA82s may be ordered for maximum disk storage capacity. New Disks The two primary disks supported in standard system configurations are the RA70, a 5 1/4" 280 MB Winchester and the 14" RA82 622 MB Winchester. Both disks are supported by the KDA50 disk controller, offered as standard on all systems. The KDA50 supports a maximum of four RA disks. Support for RA81 and RA60 disks is also available. Users will discover significant I/O throughput improvements over previous MicroVAX II RD disks (see following chart). For larger storage needs, the MicroVAX 3600 can be configured with up to four RA82s for a maximum disk storage capacity of 2.5 GBytes. DISK PERFORMANCE COMPARISON Access Transfer Disk Capacity Time Rate RD53/54 71/159 MB 38 ms 625 KB/sec RA70 280 MB 27 ms 1.4 MB/sec RA81 456 MB 36 ms 2.2 MB/sec RA82 622 MB 32 ms 2.4 MB/sec Improved Backup Backup tape subsystems are incorporated as standard in every system. The TK70 offers three times the storage capacity (296 MB), better than twice the transfer rate (125 KB/sec), and has improved operator handling features over its predecessor, the TK50. The TK70 reads TK50 media for compatibility of data and applications media across the MicroVAX family. In addition, the TSV05 is supported as an option, and the TU81-Plus is offered both as an option and standard in the MicroVAX 3600 Extended Cabinet configuration. New Packaging The standard systems offered make configuring and ordering easy. Each system comes complete with system software, either VMS or ULTRIX-32. VAXELN support for these systems will be included in the next major release of VAXELN. Timesharing system configurations also include networking software licenses and VMS services for MS-DOS. VAXserver Configurations Each VAXserver configuration has been designed to offer maximum opportunity in the workstation environment for applications that do not require timesharing functionality. Single user operating system licenses and appropriate networking software licenses are included. If your customer requires a mix of timesharing and workstation services, sell a general- purpose MicroVAX 3500/3600 system. PRODUCT DESCRIPTIONS For office environments, the MicroVAX 3500 is a 27-inch high pedestal timesharing system. The VAXserver 3500 pedestal server, like the VAXserver 3600 cabinet, is optimally configured to sell as a boot member in Local Area VAXcluster environments. MicroVAX 3500 VAXserver 3500 Timesharing System Server System DV-350T1-Ax -- VMS DV-350S1-Ax -- VMS DU-350T1-Ax -- ULTRIX-32 DU-350S1-Ax -- ULTRIX-32 Includes: Includes: o KA650-AA CPU board o KA650-AA CPU o 16 MB main memory o 16 MB main memory o BA213 pedestal o BA213 pedestal o TK70 296 MB cartridge tape o TK70 tape o KDA50 disk controller o KDA50 disk controller o RA70 280 MB 5.25" fixed disk o RA70 o DELQA-SA Ethernet controller o DELQA-SA o 1-Year system warranty o 1-Year system warranty DV-350T1-Ax includes: DV-350S1-Ax includes: o 1-20 user VMS license o VMS File and Application Services license o DECnet End-Node license o DECnet Full-Function license o VMS services for MS-DOS (PCSA/PC) o Local Area VAXcluster license DU-350T1-Ax includes: DU-350S1-Ax includes: o 1-2 user ULTRIX-32 license o 1-2 user ULTRIX-32* license WITH unlimited user upgrade o DECnet End-Node license o DECnet End-Node license *ULTRIX-32 includes NFS networking software NOTE: In each system: - The BA213 pedestal supports a second RA70, for a total of 560 MB of disk storage in the pedestal. - One or two additional MS650-Ax 8 MB memory boards may be added, for a maximum total of 32 MB. MicroVAX 3600 MicroVAX 3600 Standard Cabinet Timesharing System Expanded Cabinet Timesharing System DV-360T1-Ax -- VMS DV-360T2-Ax -- VMS DU-360T1-Ax -- ULTRIX-32 DU-360T2-Ax -- ULTRIX-32 Includes: Includes: o KA650 CPU module o KA650 CPU o 32 MB main memory o 32 MB main memory o H9644 40" high cabinet o H9644 40" high cabinet with BA213 chassis with BA213 chassis o TK70 296 MB cartridge tape o TK70 o KDA50 disk controller o KDA50 o RA82 622 MB disk --> Two RA82 622 MB disks --> TU81-Plus 1600/6250 tape drive o DELQA-SA Ethernet controller o DELQA o 1-Year system warranty o 1-Year system warranty DV-360T1-Ax includes: DV-360T2-Ax includes: o 1-20 user VMS license o 1-40 user VMS license o DECnet End-Node license o DECnet End-Node license o VMS services for MS-DOS (PSCA/PC) o VMS services for MS-DOS (PCSA/PC) DU-360T1-Ax includes: DU-360T2-Ax includes: o 1-2 user ULTRIX-32 license o 1-2 user ULTRIX-32 license WITH unlimited user upgrade WITH unlimited user upgrade o DECnet End-Node license o DECnet End-Node license NOTE: The CPU cabinet in both the MicroVAX 3600 Standard and Expanded Cabinet systems supports two optional RA70 disk drives within the enclosure. A total of four RA disks can be supported, for a possible mass storage maximum of 2.5 GB (four RA82 disks, requiring additional cabinet). For distributed workstation environments that require a cost-effective compute or disk server, the VAXserver 3600 is an attractively priced cabinet configuration. The VAXserver 3602 is a configuration of TWO VAXserver CPUs, sharing a dual-ported RA82 disk. Each CPU can support up to three more RA disks, either shared or "private." The VMS configuration includes a DECnet Full-Function license on one CPU, and DECnet End-Node license on the second, and LAVc licenses on both. The ULTRIX-32 configuration includes DECnet End-Node and NFS licenses. Typically, the first CPU acts as the boot node and file server; and the second CPU acts as a compute resource for the rest of the LAVc or LAN system, while providing a "standby" for the boot node. Together, they provide up to eight times a MicroVAX II in compute power applied to shared services. VAXserver 3600 VAXserver 3602 Cabinet Server System Dual-CPU Cabinet Server System DV-360S1-Ax -- VMS DV-360S2-Ax -- VMS DU-360S1-Ax -- ULTRIX-32 DU-360S2-Ax -- ULTRIX-32 o KA650 CPU module o 2 KA650 CPUs o 16 MB main memory o Each with 16 MB main memory o H9644 40" high cabinet o 2 H9644 cabinets with with BA213 chassis BA213 chassis o TK70 296 MB cartridge tape o 1 TK70 o KDA50 disk controller o 2 KDA50s o RA82 622 MB disk o 1 RA82 (dual-ported) o DELQA Ethernet controller o 2 DELQAs o 1-Year system warranty o 1-Year system warranty DV-360S1-Ax includes: DV-360S2-Ax includes: o VMS File and Application o 2 VMS File and Application Services license Services licenses o DECnet Full-Function license o 1 DECnet Full-Function license o Local Area VAXcluster license o 1 DECnet End-Node license o Local Area VAXcluster software DU-360S1-Ax includes: DU-360S2-Ax includes: o 1-2 user ULTRIX-32 license o Two 1-2 user ULTRIX-32 licenses o DECnet End-Node license o 2 DECnet End-Node licenses NOTE: The VAXserver 3600 supports two optional RA70 disk drives within the rack enclosure. A total of four RA disks can be supported. OPTIONAL HARDWARE SUPPORTED There are many Q-bus options that will be immediately supported with the new systems. For a complete list of supported options see the configuration worksheets. VAX PRODUCT POSITIONING The MicroVAX 3500/3600 and VAXserver 3500/3600/3602 products feature mid-range performance, and they are clearly distinguished from the VAX 8xxx systems on the dimension of total system functionality. The MicroVAX 3500/3600 are price/performance leaders but do not support multiprocessing, VAXBI options, or CI-based VAXcluster systems. VAXclusters offer high bandwidth disk I/O along with more sophisticated capabilities, such as automatic failover of disks, volume shadowing, and therefore high availability with no single point of failure. Understanding these differences is of primary importance when determining when to sell a MicroVAX 3500/3600 and when to sell a VAX 8xxx system. The chart below, which points out functional differences, should aid you in proposing the right solution. These differences must be carefully considered before recommending an appropriate system. No longer can CPU performance alone be an adequate comparison point. COMPARISONS FUNCTIONALITY: MicroVAX Family <-- --> VAX 8xxx o Single CPU o Single/Multiple CPUs (8350) o Q-bus (3 MBytes/sec) o VAXBI (13.3 MBytes/sec) o Ethernet Cluster (10 Mbits/sec) o CI Cluster (70 Mbits/sec) o Disk Capacity (up to 2.4 GBytes) o Disk Capacity (6-60 GBytes with HSC70) o Memory Size (32 MBytes) o Memory Size (128 to 256 MBytes) APPLICATION ENVIRONMENT: DEVELOPMENT <-- --> PRODUCTION o Entry Vehicle o Installed Base Vehicle o Office Environment o Computer Room Environment o Departmental, Work Group, o Centralized Computing Small Departments o Distributed Functions o MIS Functions o Low Entry Level Price o Win at High End o Boot Node on LAVc o Node on CI Cluster o Application Development o Data Processing Production ONE-YEAR SYSTEM WARRANTY The one-year system warranty on these systems includes integrated services for both the system hardware and software, including the operating system. Hardware warranty services include on-site repair at the DECservice level, which features four-hour response time and guaranteed workthroughs. Software Product Services include telephone support through a 24-hour (800) number, right-to-use updates, and access to the Digital Software Information Network (DSIN). ================================================================================ Note 42.4 MICROVAX SYSTEM ANNOUCEMENTS 4 of 17 EISNER::KILLEEN "Jeff Killeen" 168 lines 11-SEP-1987 00:30 -< MICRO VAX 3000 WORK STATIONS >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! THE VAXstation 3200 AND VAXstation 3500 - 10-September-1987 o First workstations using the next generation, high-performance CPU o More than 3X performance of VAXstation II/GPX o High-performance VAXstation for low-end 3D applications o Competitive new prices and configurations for VAXstation II/GPX The VAXstation 3200 and VAXstation 3500 -- the newest members of the VAX family of workstations -- provide high-performance graphics with the new increased performance of the CMOS VAX microprocessor. PRODUCT DESCRIPTION The VAXstation 3200 and VAXstation 3500 are single-user VAXstation systems based on the new CVAX CPU and FPU chip set. The VAXstation 3200 is housed in the familiar BA23 enclosure. The VAXstation 3500 is housed in the newer BA213 pedestal enclosure, which increases memory and disk expansion capacity over that of existing VAXstations. The new BA213 enclosure also offers easier access to modules, and simple, unique I/O connection design. The increased performance of these products enables the customer to meet the needs for faster turn-around time with the ability to handle larger models which require more intensive analysis. With these products, the customer will have a competitive workstation with the power to design, model and analyze complex problems at his/her desk. VAXstation II/GPX Price Reductions Improvements in technology and lower manufacturing costs have made possible the announcement of 15-20% price reductions on selected VAXstation II/GPX configurations. VAXstation II/GPX continues to be a cost-effective alternative for customers who do not require the higher CPU performance of the VAXstation 3xxx workstations. See the Appendix at the end of this special issue for VAXstation II/GPX price details. FEATURES BENEFITS CMOS Technology Higher CPU Performance Increased Performance Power for more sophisticated applications and better user response FEATURES BENEFITS Software Compatibility Immediate access to over 300 VAX applications and over 100 VAXstation graphics applications Price/Performance More productivity per dollar than the workstation competition POSITIONING WITH OTHER DIGITAL PRODUCTS The following table summarizes the features and differences of the VAXstation family: =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- | | VS2000 | VSII/GPX | VS3200 | VS3500 | =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- | | | | | | | CPU/FPU | MicroVAX II | MicroVAX II | CVAX | CVAX | | | | | | | | # PLanes | 1 or 4 | 4 or 8 | 4 or 8 | 4 or 8 | | | | | | | | Monitors | 15+19"/M+C | 19"/M+C | 19"/M+C | 19"/M+C | | | | | | | | Resolution | 1024 x 864 | 1024 x 864 | 1024 x 864 | 1024 x 864 | | | | | | | | Backplane | None | Q-bus | Q-bus | Q-bus | | # Slots | | 8/12 | 8 | 12 | | | | | | | | Max. Memory | 6 MB | 16 MB | 16 MB | 32 MB | | | | | | | | Min. Disk | 42 MB | 71 MB | 71 MB | 280 MB | | Max. Disk | 318 MB | 477 MB | 318 MB | 560 MB | | | | 1,824* | | | | | | | | | | CPU Perf. | 1 | 1 | 3 | 3 | | | | | | | =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- *Disk capacity is achieved by adding the GPX graphics subsystem to a Configuration 5 MicroVAX II. CONFIGURATION INFORMATION Basic Configurations VMS-based systems include the following licenses: o VMS Operating System o VWS o DECnet End-Node License o Local Area VAXcluster License ULTRIX-based systems include the following licenses: o ULTRIX Operating System o NFSTM o ULTRIX C, VAX C o ULTRIX FORTRAN o ULTRIX Pascal o X-Windows VAXstation 3200: o BA23 enclosure o CMOS CPU and FPU o 8 MBytes of memory, expandable to 16 MB o Ethernet controller o Graphics subsystem: - 4-plane monochrome with 19-inch monitor, OR - 8-plane color with 19-inch monitor o One-year hardware and software warranty VS300 BA23,4-PL MONO,8MB,DELQA,DISKLESS VS305 BA23,8-PL COLOR,8MB,DELQA,DISKLESS VS310 BA23,4-PL MONO,8MB,DELQA,RD54 VS315 BA23,8-PL COLOR,8MB,DELQA,RD54 VAXstation 3500: o BA213 enclosure o CMOS CPU and FPU o 16 MBytes of memory, expandable to 32 MB o 280 MByte disk subsystem (RA70) o 296 MByte tape subsystem (TK70) o Ethernet controller o Graphics subsystem: - 4-plane monochrome with 19-inch monitor, OR - 8-plane color with 19-inch monitor o One-year hardware and software warranty VS350 BA213,4-PL MONO,16MB,DELQA,RA70,TK70 VS355 BA213,8-PL COLOR,16MB,DELQA,RA70,TK70 SERVICES The VAXstation 3200 and VAXstation 3500 include a one-year hardware and software warranty which includes: o One-year Basic on-site hardware service o One-year helpline for hardware and software (operating system and all layered software products) o Right-to-Use software updates (operating system) o One-year access to Digital Software Information Network NFS is a trademark of Sun Microsystems, Inc. ================================================================================ Note 42.5 MICROVAX SYSTEM ANNOUCEMENTS 5 of 17 EISNER::KILLEEN "Jeff Killeen" 73 lines 11-SEP-1987 00:31 -< MV3000 PACKAGING (BA200 BOXES) >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! MicroVAX 3xxx-BASED SYSTEMS - 10-September-1987 ADOPT NEW STANDARD IN SYSTEMS PACKAGING o Introduced initially for factory floor applications -- equally at home in an office environment o Enhanced reliability o The enclosure allows easier access/maintenance In March, 1987, Digital introduced a new concept in Q-bus electro-mechanical packaging - the new BA200 Series Q-bus chassis -- on its Industrial VAX (IVAX) and Industrial PDP families of factory floor computers (Sales Update Special Issue dated March 24, 1987). The many innovative features, originally included to minimize the effects of environmental hazards, will prove beneficial in more hospitable environments as well. Quiet operation and enhanced reliablitity and operating characteristics are features of all systems that use the BA200 family of enclosures and plug-in options. For these reasons, Digital has chosen to debut its new MicroVAX 3500/3600, VAXserver 3500/3600/3602, and VAXstation 3500 systems in this new "industrial strength" packaging. WHAT'S NEW ABOUT THIS CHASSIS? In order to improve access, indicator visibility, and cabling from the front of the system, space has been allowed for the attachment of an integral I/O bulkhead to provide filtering and placement of the I/O connectors. The bulkhead also improves module rigidity and tolerance to vibration and shock. For this design consideration, the space between the backplane rows (and therefore, the modules) has been increased from the 0.5 inches of previous enclosures, to 0.95 inches. The New BA213 Chassis The heart of the new chassis is the bulkhead, which Digital manufactures to exacting specifications. Bulkhead Design Kits, including an Evaluation Kit, will be available to assist customers in evaluating the adaptation of options, and to provide a source of finished and semi-finished parts for production use. NOTE: The Q-bus has not been changed in any way. Also, the printed circuit board form factor is unaltered. HANDLES Because all cabling and I/O connections in the new BA200 Series enclosures are now accessible from the front of the system, some modules will need to be modified. To fit into the new bulkhead, modules must be outfitted with an appropriate "handle." Digital will offer a standard handle, which is recommended for most designs. It has cutouts that accommodate a range of D-type connectors, from 10-pin to 50-pin. It comes completely assembled (including springs, levers, fasteners, etc.). Digital is prepared to assist with designs that require other types of handles as well. Some general guidelines: o Modules whose sole connection is to the backplane need not be changed. This includes devices such as array processors and transform engines. o For modules with external connections, such as communications controllers and line printer controllers, the customer will need to select and adapt an appropriate handle. ================================================================================ Note 42.6 MICROVAX SYSTEM ANNOUCEMENTS 6 of 17 EISNER::KILLEEN "Jeff Killeen" 294 lines 11-SEP-1987 00:32 -< MV3000 / GPX PRICING >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! New Product Pricing Information - 10-September-1987 Pricing is effective as of September 9, 1987. MicroVAX 3500/3600 SYSTEMS VAXserver 3500/3600/3602 SYSTEMS Order No. Description U. S. List Price BMC DV-350T1-AA MV3500, 16MB, TK70, RA70, 1-20 VMS USERS, U.S. $74,800 * DV-350T1-A2 MV3500, 16MB, TK70, RA70, 1-20 VMS USERS, 120V ** * DV-350T1-A3 MV3500, 16MB, TK70, RA70, 1-20 VMS USERS, 240V ** * DU-350T1-AA MV3500, 16MB, TK70, RA70, 65+USER ULTRIX, U.S. $74,800 * DU-350T1-A2 MV3500,16MB, TK70, RA70, 65+USER ULTRIX, 120V ** * DU-350T1-A3 MV3500, 16MB, TK70, RA70, 65+USER ULTRIX, 240V ** * DV-350S1-AA VXSRV3500, 16MB, TK70, RA70, SINGL/USER VMS, U.S. $49,900 * DV-350S1-A2 VXSRV3500, 16MB, TK70, RA70, SINGL/USER VMS, 120V ** * DV-350S1-A3 VXSRV3500, 16MB, TK70, RA70, SINGL/USER VMS, 240V ** * DU-350S1-AA VXSRV3500, 16MB, TK70, RA70, 1-2 USER ULTRIX, U.S. $49,900 * DU-350S1-A2 VXSRV3500, 16MB, TK70, RA70, 1-2 USER ULTRIX, 120V ** * DU-350S1-A3 VXSRV3500, 16MB, TK70, RA70, 1-2 USER ULTRIX, 240V ** * DV-360T1-AA MV3600, 32MB, TK70, RA82, 1-20VMS USERS, U.S. $99,800 * DV-360T1-A2 MV3600, 32MB, TK70, RA82, 1-20VMS USERS, 120V ** * DV-360T1-A3 MV3600, 32MB, TK70, RA82, 1-20VMS USERS, 240V ** * DU-360T1-AA MV3600, 32MB, TK70, RA82, 65+ USER ULTRIX, U.S. $99,800 * DU-360T1-A2 MV3600, 32MB, TK70, RA82, 65+ USER ULTRIX, 120V ** * DU-360T1-A3 MV3600, 32MB, TK70, RA82, 65+ USER ULTRIX, 240V ** * *Call Field Service **Consult AQS Order No. Description U. S. List Price BMC DV-360T2-AA MV3600, 32MB, TK70, 2-RA82, TU81-E, 1-40 VMS USERS, U.S. $169,800 * DV-360T2-A2 MV3600, 32MB, TK70, 2-RA82, TU81-E, 1-40 VMS USERS, 120V ** * DV-360T2-A3 MV3600, 32MB, TK70, 2-RA82, TU81-E, 1-40 VMS USERS, 240V ** * DU-360T2-AA MV3600, 32MB, TK70, 2-RA82, TU81-E, 65+ USER ULTRIX, U.S. $149,800 * DU-360T2-A2 MV3600, 32MB, TK70, 2-RA82, TU81-E, 65+ USER ULTRIX, 120V ** * DU-360T2-A3 MV3600, 32MB, TK70, 2-RA82, TU81-E, 65+ USER ULTRIX, 240V ** * DV-360S1-AA VXSRV3600, 16MB, TK70, RA82, SINGL/USER VMS, U.S. $66,900 * DV-360S1-A2 VXSRV3600, 16MB, TK70, RA82, SINGL/USER VMS, 120V ** * DV-360S1-A3 VXSRV3600, 16MB, TK70, RA82, SINGL/USER VMS, 240V ** * DU-360S1-AA VXSRV3600, 16MB, TK70, RA82, 1-2 USER ULTRIX, U.S. $66,900 * DU-360S1-A2 VXSRV3600, 16MB, TK70, RA82, 1-2 USER ULTRIX, 120V ** * DU-360S1-A3 VXSRV3600, 16MB, TK70, RA82, 1-2 USER ULTRIX, 240V ** * DV-360S2-AA VXSRV3602, 32MB, TK70, RA82, SINGL/USER VMS, U.S. $105,500 * DV-360S2-A2 VXSRV3602, 32MB, TK70, RA82, SINGL/USER VMS, 120V ** * DV-360S2-A3 VXSRV3602, 32MB, TK70, RA82, SINGL/USER VMS, 240V ** * DU-360S2-AA VXSRV3602, 32MB, TK70, RA82, 1-2 USER ULTRIX, U.S. $105,500 * DU-360S2-A2 VXSRV3602, 32MB, TK70, RA82, 1-2 USER ULTRIX, 120V ** * DU-360S2-A3 VXSRV3602, 32MB, TK70, RA82, 1-2 USER ULTRIX, 240V ** * VAXstation 3200 SYSTEMS Order No. Description U. S. List Price BMC VS300-AA VS3200,4-PL MONO,8MB,DELQA,DISKLESS, SINGL/USER VMS,VWS,LAVC,DECNET E/N,120V $19,525 * VS300-A3 VS3200,4-PL MONO,8MB,DELQA,DISKLESS, SINGL/USER VMS,VWS,LAVC,DECNET E/N,240V ** * Order No. Description U. S. List Price BMC VS305-AA VS3200,8-PL COLOR,8MB,DELQA,DISKLESS, SINGL/USER VMS,VWS,LAVC,DECNET E/N,120V $27,525 * VS305-A3 VS3200,8-PL COLOR,8MB,DELQA,DISKLESS, SINGL/USER VMS,VWS,LAVC,DECNET E/N,240V ** * VS305-A4 VS3200,8-PL COLOR,8MB,DELQA,DISKLESS SINGL/USER VMS,VWS,LAVC,DECNET E/N, 240V,SH ** * VS310-AA VS3200,4-PL MONO,8MB,DELQA,RD54, SINGL/USER VMS,VWS,LAVC,DECNET E/N,120V $26,575 * VS310-A3 VS3200,4-PL MONO,8MB,DELQA,RD54, SINGL/USER VMS,VWS,LAVC,DECNET E/N,240V ** * VS315-AA VS3200,8-PL COLOR,8MB,DELQA,RD54, SINGL/USER VMS,VWS,LAVC,DECNET E/N,120V $34,575 * VS315-A3 VS3200,8-PL COLOR,8MB,DELQA,RD54, SINGL/USER VMS,VWS,LAVC,DECNET E/N,240V ** * VS315-A4 VS3200,8-PL COLOR,8MB,DELQA,RD54, SINGL/USER VMS,VWS,LAVC,DECNET E/N, 240V,SH ** * VS300-BA VS3200,4-PL MONO,8MB,DELQA,DISKLESS, 1-2 ULTRIX WS,120V $19,525 * VS300-B3 VS3200,4-PL MONO,8MB,DELQA,DISKLESS, 1-2 ULTRIX WS,240V ** * VS305-BA VS3200,8-PL COLOR,8MB,DELQA,DISKLESS, 1-2 ULTRIX WS,120V $27,525 * VS305-B3 VS3200,8-PL COLOR,8MB,DELQA,DISKLESS, 1-2 ULTRIX WS,240V ** * VS305-B4 VS3200,8-PL COLOR,8MB,DELQA,DISKLESS, 1-2 ULTRIX WS,240V,SH ** * VS310-BA VS3200,4-PL MONO,8MB,DELQA,RD54, 1-2 ULTRIX WS,120V $26,575 * VS310-B3 VS3200,4-PL MONO,8MB,DELQA,RD54, 1-2 ULTRIX WS,240V ** * VS315-BA VS3200,8-PL COLOR,8MB,DELQA,RD54, 1-2 ULTRIX WS,120V $34,575 * VS315-B3 VS3200,8-PL COLOR,8MB,DELQA,RD54, 1-2 ULTRIX WS,240V ** * VS315-B4 VS3200,8-PL COLOR,8MB,DELQA,RD54, 1-2 ULTRIX WS,240V,SH ** * VAXstation 3500 SYSTEMS Order No. Description U. S. List Price BMC VS350-AA VS3500,4-PL MONO,16MB,DELQA,RA70,TK70, SINGL/USER VMS,VWS,LAVC,DECNET E/N,120V $50,005 * VS350-A3 VS3500,4-PL MONO,16MB,DELQA,RA70,TK70, SINGL/USER VMS,VWS,LAVC,DECNET E/N,240V ** * VS355-AA VS3500,8-PL COLOR,16MB,DELQA,RA70,TK70, SINGL/USER VMS,VWS,LAVC,DECNET E/N,120V $58,005 * VS355-A3 VS3500,8-PL COLOR,16MB,DELQA,RA70,TK70, SINGL/USER VMS,VWS,LAVC,DECNET E/N,240V ** * VS355-A4 VS3500,8-PL COLOR,16MB,DELQA,RA70,TK70, SINGL/USER VMS,VWS,LAVC,DECNET E/N, 240V,SH ** * VS350-BA VS3500,4-PL MONO,16MB,DELQA,RA70, TK70,1-2 ULTRIX WS,120V $50,005 * VS350-B3 VS3500,4-PL MONO,16MB,DELQA,RA70, TK70,1-2 ULTRIX WS,240V ** * VS355-BA VS3500,8-PL COLOR,16MB,DELQA,RA70, TK70,1-2 ULTRIX WS,120V $58,005 * VS355-B3 VS3500,8-PL COLOR,16MB,DELQA,RA70, TK70,1-2 ULTRIX WS,240V ** * VS355-B4 VS3500,8-PL COLOR,16MB,DELQA,RA70, TK70,1-2 ULTRIX WS,240V, SH ** * VAXstation II/GPX SYSTEMS VS230-FA BA23,8-PL COLOR,5MB,DELQA,DISKLESS SINGL/USER VMS,VWS,LAVC,DECNET E/N,120V $21,525 * VS230-F3 BA23,8-PL COLOR,5MB,DELQA,DISKLESS SINGL/USER VMS,VWS,LAVC,DECNET E/N,240V ** * VS230-F4 BA23,8-PL COLOR,5MB,DELQA,DISKLESS SINGL/USER VMS,VWS,LAVC,DECNET E/N, 240V-SH ** * VS230-HA BA23,8-PL COLOR,5MB,DELQA,RD54 SINGL/USER VMS,VWS,LAVC,DECNET E/N,120V $28,575 * VS230-H3 BA23,8-PL COLOR,5MB,DELQA,RD54 SINGL/USER VMS,VWS,LAVC,DECNET E/N,240V ** * VS230-H4 BA23,8-PL COLOR,5MB,DELQA,RD54 SINGL/USER VMS,VWS,LAVC,DECNET E/N, 240V-SH ** * VS230-GA BA23,8-PL COLOR,5MB,DELQA,DISKLESS 1-2 ULTRIX WS,120V $21,525 * VS230-G3 BA23,8-PL COLOR,5MB,DELQA,DISKLESS 1-2 ULTRIX WS,240V ** * VS230-G4 BA23,8-PL COLOR,5MB,DELQA,DISKLESS 1-2 ULTRIX WS,240V-SH ** * Order No. Description U. S. List Price BMC VS230-JA BA23,8-PL COLOR,5MB,DELQA,RD54 1-2 ULTRIX WS,120V $28,575 * VS230-J3 BA23,8-PL COLOR,5MB,DELQA,RD54 1-2 ULTRIX WS,240V ** * VS230-J4 BA23,8-PL COLOR,5MB,DELQA,RD54 1-2 ULTRIX WS,240V-SH ** * VS235-FA BA123,8-PL COLOR,5MB,DELQA,RD54,TK50 SINGL/USER VMS,VWS,LAVC,DECNET E/N,120V $36,475 * VS235-F3 BA123,8-PL COLOR,5MB,DELQA,RD54,TK50 SINGL/USER VMS,VWS,LAVC,DECNET E/N,240V ** * VS235-F4 BA123,8-PL COLOR,5MB,DELQA,RD54,TK50 SINGL/USER VMS,VWS,LAVC,DECNET E/N, 240V-SH ** * VS235-GA BA123,8-PL COLOR,5MB,DELQA,RD54,TK50 1-2 ULTRIX WS,120V $36,475 * VS235-G3 BA123,8-PL COLOR,5MB,DELQA,RD54,TK50 1-2 ULTRIX WS,240V ** * VS235-G4 BA123,8-PL COLOR,5MB,DELQA,RD54,TK50 1-2 ULTRIX WS,240V-SH ** * *Call Field Service **Consult AQS MEMORY * Order No. Description U. S. List Price BMC Instal. MS650-AA 8MB ECC MicroVAX 3xxx-Based Memory, Factory Installed in BA213 and BA23 Enclosures; Field Installed in BA23 Enclosure $4,000 0 TBD MS650-AF 8MB ECC MicroVAX 3xxx-Based Memory, Field Installed in BA213 Enclosure $4,000 0 TBD RA70 and RA80 Add-on Order No. Description U. S. List Price BMC DSMC Instal. RA70E-SA 280 MByte fixed disk drive (Factory Installed) $ 9,000 $ 45 $ 54 N/A RA70E-SF 280 MByte fixed disk drive (Field Add-on) 90 days after FRS $ 9,000 $ 45 $ 54 $458 RA82-EA 3 RA82 disk drives in a 3-high cabinet 120V/60Hz $65,500 $177 $210 $950* RA82-ED 3 RA82 disk drives in a 3-high cabinet 240V/50Hz ** ** ** ** *No installation charge when included with the system. **Consult AQS TK70 Order No. Description MLP BMC DSMC Instal. TK52-K 296 MB Cartridge $45 N/A N/A N/A LAVc MLP Q0ZCE-UZ LAVc license $4,350 Please note that the new VAXstations and VAXserver systems include LAVc licenses. For more Local Area VAXcluster licensing information, see SPD #27.65.xx. ================================================================================ Note 42.7 MICROVAX SYSTEM ANNOUCEMENTS 7 of 17 EISNER::FRISBIE "Alan E. Frisbie" 37 lines 14-SEP-1987 12:56 -< BA123 vs BA213 backplane slots >- -------------------------------------------------------------------------------- One important point has been overlooked in the announcement of DEC's new BA213 cabinet (used with the mV3xxx series). That is the configuration of the backplane slots. Both cabinets have 12 Quad slots, but the new BA213 backplane only has the Q22 bus on the A/B fingers. This reduces the number of dual-width Q22 slots by 8! To illustrate: BA123 (old) BA213 (new) slot +-------+-------+ +-------+-------+ 1 | Q22 | C/D | | Q22 | C/D | +-------+-------+ +-------+-------+ 2 | Q22 | C/D | | Q22 | C/D | +-------+-------+ +-------+-------+ 3 | Q22 | C/D | | Q22 | C/D | +-------+-------+ +-------+-------+ 4 | Q22 | C/D | | Q22 | C/D | +-------+-------+ +-------+-------+ 5 | Q22 | Q22 | | Q22 | C/D | +-------+-------+ +-------+-------+ . . . . . . (same as slot 5) . . . +-------+-------+ +-------+-------+ 12 | Q22 | Q22 | | Q22 | C/D | +-------+-------+ +-------+-------+ Those users who already have a shortage of Q-bus backplane slots in the BA123 boxes will have a *real* problem if they try to use the new BA213 boxes. You have been warned. As a side note, would anyone care to make a bet on how many orders get screwed up because someone confuses BA123 with BA213? Alan Frisbie ================================================================================ Note 42.8 MICROVAX SYSTEM ANNOUCEMENTS 8 of 17 EISNER::KASPER "Beverly T. Kasper" 7 lines 23-SEP-1987 11:51 -< You mean you want *software* for it, too?! >- -------------------------------------------------------------------------------- Does *ANYBODY* know what the designator code for these beasts are? We're trying to get a price quote from the E-Store for some layered products; of course, we need to know what the ? in Q?015-QZ should be. Our software services person tells us to call our sales rep. She's not in the office today, nor are any of the other sales reps. No one else we've talked to at DEC seems to know this little tidbit . . . ================================================================================ Note 42.9 MICROVAX SYSTEM ANNOUCEMENTS 9 of 17 EISNER::KILLEEN "Jeff Killeen" 7 lines 23-SEP-1987 18:35 -< HOW ABOUT Q0 ??? >- -------------------------------------------------------------------------------- This is a guess... Q0???-?? If you type Q?001-UZ into the electonic store that is the only VMS product that comes up that does not have an official CPU code. ================================================================================ Note 42.10 MICROVAX SYSTEM ANNOUCEMENTS 10 of 17 EISNER::KILLEEN "Jeff Killeen" 193 lines 10-OCT-1987 15:00 -< MicroVAX 3xxx GROWTH >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! MicroVAX 3xxx GROWTH PLANNING MESSAGES - 29-September-1987 o Help your customer answer questions about growth planning strategies o MicroVAX II and PDP-11 customers have multiple growth options o "In-cabinet" upgrades from the MicroVAX II to the MicroVAX 3xxx systems are not currently planned With Digital's latest major product announcement of the MicroVAX 3xxx series of systems and workstations, our installed base of PDP-11, MicroVAX II and smaller VAX (11/725, 11/730 and 11/750) systems now have more growth options than ever before! Naturally, your customers will also want to understand the upgrade strategy to the MicroVAX 3xxx. Presented here is Digital's current position on upgrades. We recommend these responses in answer to customer questions and for assisting customers with direction in system growth planning. Specifically, the Questions and Answers address the following: o Upgrades from a MicroVAX II to the MicroVAX 3xxx o Other growth options for the MicroVAX II user o PDP-11 migration to the MicroVAX 3xxx o Trade-Ins to the MicroVAX 3xxx for complete system replacements o Future migration from MicroVAX 3xxx to VAXBI systems PRODUCT POSITIONING The MicroVAX 3xxx systems offer very attractive price/performance as potential upgrade paths from smaller PDP-11 or VAX systems. Both End-User and Volume customers are encouraged to evaluate the performance enhancements of the MicroVAX 3xxx and purchase the appropriate level system, for purposes of adding to the existing installation or replacing the current installation. The relative CPU performance increases of the MicroVAX 3xxx over existing installed systems will prompt many users to: o Replace multiple smaller systems with one MicroVAX 3xxx o Purchase the VAXserver 3xxx as a boot node for an existing Local Area VAXcluster o Or buy a MicroVAX 3xxx as a new high-performance satellite member in an existing LAVc o Create a new Local Area VAXcluster, using existing MicroVAX IIs as satellite nodes and a new VAXserver 3xxx as the boot node o Or use new MicroVAX 3xxx systems as initial systems in a new LAVc o Replace an existing workstation with a higher performance VAXstation 3xxx/GPX o Or add a new MicroVAX 3xxx for separate, demanding applications that require more performance. We can now grow Local Area VAXcluster compute power in increments of either one times (MicroVAX II addition) or three times (MicroVAX 3xxx addition) a MicroVAX II system. Adding a second system to a nonclustered installation not only offers higher performance for existing applications, but also room for future growth and capacity to add new applications. Some examples of estimated CPU relative performance comparisons (relative to the MicroVAX 3xxx) are: --------------------------------------------------------- MicroVAX 3xxx | ***************************************************** | | | MicroVAX II | ***************** | | | MicroPDP-11/83 | *************** | | | VAX-11/750 | *********** | | | MicroPDP-11/73, | ********* | PDP-11/44 | ********* | | | MicroPDP-11/53, | ****** | VAX-11/730, | ****** | MicroVAX-11/725 | ****** | | | MicroVAX I | ****** | | | PDP-11/24, | **** | PDP-11/34 | **** | | | MicroPDP-11/23 | ** | --------------------------------------------------------- Relative CPU Performance** **NOTE: All relative CPU performance ratings are application dependent. We recommended that specified benchmark tests be done on comparison systems with the customer's unique application and configuration requirements. QUESTIONS AND ANSWERS Q. Does Digital intend to offer upgrades from MicroVAX II to MicroVAX 3xxx systems? Upgrading the performance of the MicroVAX II to that of the MicroVAX 3xxx system would be "technically feasible." However, Digital has no specific plans to offer a MicroVAX II to MicroVAX 3xxx upgrade at this time. Q. When could the upgrades become available? Upgrades are not currently planned. Digital does not comment on future product considerations. We recommend that existing MicroVAX II users consider alternate growth paths to address current performance improvement requirements. The preferred upgrade path is to create a Local Area VAXcluster or to grow an existing Local Area VAXcluster by adding computing resources as needed. MicroVAX II users can easily move their existing MicroVAX II application to a MicroVAX 3xxx, retaining a high degree of peripheral compatibility and avoiding conversion costs. Q. If the upgrade becomes available, how much will it cost? We cannot comment on the possible pricing of unplanned products at this time. Q. Will upgrades to the MicroVAX 3xxx be offered from any other systems, like the MicroPDP-11? An upgrade from the MicroPDP-11 is not currently being planned. We recommend that existing MicroPDP-11 users who are considering migration to VAX systems purchase the MicroVAX 3500 entry-level system. The MicroPDP-11 and MicroVAX 3500 can then be networked together during the software conversion phase, and can continue to be used in a coexistence environment. MicroPDP-11 customers who are looking for an "in-cabinet" upgrade to a VAX system are encouraged to purchase the MicroVAX II Upgrade Kit, which is economically priced as a total packaged solution, while offering up to twice the relative CPU performance of the MicroPDP-11/73 and almost five times the performance of a MicroPDP-11/23. All PDP-11 users who migrate to VAX will find that Digital's VAX/VMS computer systems are the most cost-effective offerings in the industry today. With networking, VAX/VMS systems span the range from small desktop computers to large mainframe systems, all sharing one common architecture. Q. Will Digital offer trade-ins to MicroVAX 3xxx system purchases? Trade-ins to the MicroVAX 3xxx for system replacements are not available at this time, but may be considered in the future, if appropriate. Q. Will Digital offer future performance upgrades from the MicroVAX 3xxx series to its larger VAXBI systems? New MicroVAX 3xxx customers will be able to grow their systems much the same way that existing MicroVAX II customers do: clustering (within the Local Area VAXcluster environment) and adding new stand-alone or networked CPUs to the current installation. Customers who anticipate the need for a VAXBI system in the near term should consider purchasing the appropriate VAXBI system and familiarize themselves with the built-in upgradeability of many of those systems. Should MicroVAX 3xxx users require a VAXBI system later, the VAX/VMS architecture offers an attractive migration path. Although trade-in allowances for MicroVAX 3xxx systems toward the purchase of larger VAXBI systems are currently not offered, software applications and user programs can be easily moved between MicroVAX 3xxx and VAXBI systems. OVERALL GROWTH PLANNING MESSAGES Digital offers the widest range of growth alternatives in the industry. Specific in-cabinet upgrades and trade-ins to newly announced systems, however, are typically offered during different phases of the product life cycle. These offerings depend on customer demand, technology constraints and available components. We strongly recommend that considerable planning of long-term computing requirements be done in order to carefully evaluate the growth alternatives from Digital. This will help your customers make more predictable growth decisions about their information resources. Digital's customers are encouraged to implement proactive performance management techniques in order to achieve optimum productivity from existing systems, while planning future computing needs to meet their business growth expansion. ================================================================================ Note 42.11 MICROVAX SYSTEM ANNOUCEMENTS 11 of 17 EISNER::KILLEEN "Jeff Killeen" 305 lines 10-OCT-1987 15:02 -< NEW MV-II SYSTEMS >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! ENHANCED MicroVAX II SYSTEMS - 29-September-1987 o New Standard Systems with DELQA and DHQ11, replacing DEQNA and DHV11 o Configurations 4 and 5 now include TK70 o New RD54-based Configuration 3 Micro Systems Development announces new MicroVAX II standard system packages which now include the DELQA, DHQ11 and TK70. In addition, a new BA123-based single RD54 configuration has been created. The LAVc Diskless Compute Server (DJ-630PI-Bx) now includes the DELQA. The Full Boot Node (DJ-630P5-Bx) now includes the DHQ11 and DELQA. The TK70 is now available as an option on all MicroVAX II System Building Blocks. All of these systems, as do all MicroVAX II systems, include a one-year, on-site hardware warranty. There will be an immediate price increase on Standard Systems which include the DEQNA and DHV11. All of these systems will be retired on December 26, 1987. PRODUCT OVERVIEW The MicroVAX II is the mid-range member of the MicroVAX family. All of the systems being introduced in this article have all of the features of the current MicroVAX II Q-bus systems -- MicroVAX chip, floating point accelerator chip, and 1 MB onboard memory. All these systems can support up to 16 MB of memory. The following changes to the MicroVAX II standard systems are: Standard System Part No. New Old Variation Ethernet Node DH-630Q1-Fx DELQA DH-630Q1-Ex with DEQNA BA123-based DH-630Q3-Hx DHQ11, DELQA DH-630Q3-Fx with DHV11, (71 MB) DEQNA BA123-based DH-630Q3-Jx RD54, DHQ11, DELQA N/A* (159 MB) BA123-based DH-630Q4-Jx TQK/TK70, DHQ11, DH-630Q4-Hx with (318 MB) DELQA TQK/TK50,DHV11,DEQNA High-Capacity DH-630Q5-Hx TQK/TK70, DHQ11, DH-630Q5-Fx with (Cabinet) DELQA TQK/TK50,DHV11,DEQNA LAVc Diskless Compute Server DJ-630P1-Bx DELQA DJ-630P1-Ax with DEQNA Full Boot Node DJ-630P5-Bx DHQ11, DELQA DJ-630P5-Ax with DHV11, DEQNA *New as of September 21, 1987 The addition of the DHQ11 to the Standard Systems combines improved reliability with reduced power requirements. It provides backward compatibility with the DHV11. The DELQA is a Q-bus network adapter used for connecting MicroVAX II to both Ethernet V2.0 and IEEE 802.3 Local Area Networks. It is backward compatible with the latest releases of DEQNA software drivers. The TK70 tape subsystem offers three times the storage capacity (296 MB), twice the transfer rate (125 KB/sec), and has improved operator handling features over the TK50. Configurations 4 and 5, which now include the TK70, also include a five pack of TK70 cartridges. Configuration 1 is the Ethernet Node Standard System. It is suitable for use as a client node in a Local Area VAXcluster or in a network of systems maintained by the Remote System Manager software. Enhancements to Configuration 2 were announced on July 27. The most popular Standard System has been Configuration 3, the entry-level BA123 office system. In addition to enhancing the current Configuration 3 (which contains an RD53) with DHQ11 and DELQA, a new variant of Configuration 3 has been created. This new version includes an RD54 disk as well as DHQ11 and DELQA. These configurations offer customers an excellent growth path using the BA123 enclosure. The RD54 variation will give customers the ability to have the maximum storage capacity (expansion to 477 MB) in a BA123 enclosure. Configuration 4 is the larger office system, also in a BA123 enclosure. It will now include two RD54s, a TK70 tape drive, DHQ11, DELQA and 9 MB of memory. This is the more appropriate system for applications that require a great deal of storage, memory or many users to start. With the TK70, users can achieve a 1:1 backup ratio in this system. Configuration 5 is the Standard System for customers needing maximum storage and maximum number of users. For customers that can take the MicroVAX II into a computer room environment, the 40-inch cabinet allows RA disks to be configured with almost 2 GB of storage. The Diskless Compute Server (DJ-630P1-Bx) contains no disk or load device, relying on the Boot Node disks for data files and the VMS system image. The Full Boot Node (DJ-630P5-Bx) is a cabinet-based MicroVAX, which includes an RA81, TU81E, and all the software necessary for a VAXcluster Boot Node. SOFTWARE SUPPORTED The following versions of Digital's operating systems will support these new Standard Systems: MicroVMS V4.6 ULTRIX-32 V2.2 VAXELN V2.4 PRICING/ORDERING INFORMATION NEW MicroVAX II STANDARD SYSTEMS: All of the systems below have the same codes: Seg. Code Effective Date XXX September 21, 1987 U.S. Order No. Description List Price BMC DH-630Q1-FA MicroVAX II, BA23/F, MS630-BB, RQDX3, RD53, $20,900 $222 DELQA & CK-DELQA-YB, Doc/Diag, 120 V DH-630Q1-F2/F3 MicroVAX II, BA23/F, MS630-BB, RQDX3, RD53, 20,800 222 DELQA & CK-DELQA-YB, 120 V/240 V* DH-630Q3-HA MicroVAX II, BA123, MS630-BB, RQDX3, RD53, 30,900 277 DELQA & CK-DELQA-YA, TK50/TQK50, DHQ11 & CK-DHQ11-AA, Doc/Diag, 120 V DH-630Q3-H2/H3 MicroVAX II, BA123, MS630-BB, RQDX3, RD53, 30,600 277 DELQA & CK-DELQA-YA, TK50/TQK50, DHQ11 & CK-DHQ11-AA, 120 V/240 V* DH-630Q3-JA MicroVAX II, BA123, MS630-BB, RQDX3, RD54, 33,900 302 DELQA & CK-DELQA-YA, TK50/TQK50, DHQ11 & CK-DHQ11-AA, Doc/Diag, 120 V DH-630Q3-J2/J3 MicroVAX II, BA123, MS630-BB, RQDX3, RD54, 33,600 302 DELQA & CK-DELQA-YA, TK50/TQK50, DHQ11 & CK-DHQ11-AA, 120 V/240 V* DH-630Q4-JA MicroVAX II, BA123, MS630-CA, TK70/TQK70, 44,500 375 TK52K-05 (2)RD54, DHQ11 & CK-DHQ11-AA, DELQA & CK-DELQA-YA, RQDX3, Doc/Diag, 120 V DH-630Q4-J2/J3 MicroVAX II, BA123, MS630-CA, TK70/TQK70, 44,200 375 TK52K-05 (2)RD54, DHQ11 & CK-DHQ11-AA, DELQA & CK-DELQA-YA, RQDX3, 120 V/240 V* U.S. Model No. Description List Price BMC DH-630Q5-HA MicroVAX II, (2)BA23, H9642-JA, (2)MS630-CA, $44,600 $301 TK70, TQK70, TK52K-05, KDA50, DHQ11 & CK-DHQ11-AF, DELQA & CK-DELQA-YF, Doc/Diag, 120 V DH-630Q5-H2/H3 MicroVAX II, (2)BA23, H9642-JA, (2)MS630-CA, 44,300 301 TK70, TQK70, TK52K-05, KDA50, DHQ11 & CK-DHQ11-AF, DELQA & CK-DELQA-YF, 120 V/240 V* DJ-630P1-BA MicroVAX II, BA23, (2)MS630-CA, DELQA & 25,600 168 CK-DELQA-YB, QZ001-CZ (2 USER VMS), QZZCE-UZ (LAVc), QZD04-UZ (DECnet E/N), Doc, 120 V DJ-630P1-B3 MicroVAX II, BA23, (2)MS630-CA, DELQA & * 168 CK-DELQA-YB, QZ001-CZ (2 USER VMS), QZZCE-UZ (LAVc), QZD04-UZ (DECnet E/N), 240 V DJ-630P5-BA MicroVAX II, H9642, (2)BA23, (2) MS630-CA, 99,600 514 TK50/TQK50, KDA50, DELQA & CK-DELQA-YF, DHQ11 & CK-DHQ11-AF, TU81E, RA81, QZ002-C5 (VMS 8 USER), QZZCE-UZ (LAVc), QZD05-UZ (DECnet FF), Doc, 120 V DJ-630P5-B3 MicroVAX II, H9642, (2)BA23, (2) MS630-CA, * 514 TK50/TQK50, KDA50, DELQA & CK-DELQA-YF, DHQ11 & CK-DHQ11-AF, TU81E, RA81, QZ002-C5 (VMS 8 USER), QZZCE-UZ (LAVc), QZD05-UZ (DECnet FF), 240 V CURRENT MicroVAX II Standard Systems (with DEQNA/DHV11), with price increase effective September 21, 1987: NOTE: These Standard Systems will be retired on December 26, 1987. U.S. Model No. Description List Price DH-630Q1-EA MicroVAX II, BA23, MS630-BB, RQDX3, RD53, DEQNA $ 25,900 DH-630Q1-E2/E3 MicroVAX II, BA23, MS630-BB, RQDX3, RD53, DEQNA, 25,800 120 V/240 V* DH-630Q3-FA MicroVAX II, BA123, MS630-BB, RQDX3, RD53, TK50, 35,800 TQK50, DEQNA & CK KIT, DHV11 & CK KIT DH-630Q3-F2/F3 MicroVAX II, BA123, MS630-BB, RQDX3, RD53, TK50, 35,500 TQK50, DEQNA & CK KIT, DHV11 & CK KIT, 120 V/240 V* DH-630Q4-HA MicroVAX II, BA123, MS630-CA, RQDX3, (2)RD54, 47,700 TK50, TQK50, DEQNA & CK KIT, DHV11 & CK KIT DH-630Q4-H2/H3 MicroVAX II, BA123, MS630-CA, RQDX3, (2)RD54, 47,400 TK50, TQK50, DEQNA & CK KIT, DHV11 & CK KIT, 120 V/240 V* U.S. Model No. Description List Price DH-630Q5-FA MicroVAX II, H9642, (2)MS630-CA, (2)BA23, TK50, $45,000 TQK50, KDA50, DEQNA & CK KIT, DHV11 & CK KIT DH-630Q5-F2/F3 MicroVAX II, H9642, (2)MS630-CA, (2)BA23, 44,700 TK50/TQK50, KDA50, DEQNA & CK KIT, DHV11 & CK KIT, 120 V/240 V* DJ-630P1-AA MicroVAX II, BA23, (2)MS630-CA, DEQNA & CK KIT, 28,700 QZ001-CZ (2 USER VMS), QZZCE-UZ (LAVc), QZD04-UZ (DECnet E/N), Doc, 120 V DJ-630P1-A3 MicroVAX II, BA23, (2)MS630-CA, DEQNA & CK KIT, * QZ001-CZ (2 USER VMS), QZZCE-UZ (LAVc), QZD04-UZ (DECnet E/N), 240 V DJ-630P5-AA MicroVAX II, H9642, (2)BA23, (2) MS630-CA, $100,600 TK50/TQK50, KDA50, DEQNA & CK KIT, DHV11 & CK KIT, TU81E, RA81, QZ002-C5 (VMS 8 USER), QZZCE-UZ (LAVc), QZD05-UZ (DECnet FF), Doc, 120 V DJ-630P5-A3 MicroVAX II, H9642, (2)BA23, (2) MS630-CA, * TK50/TQK50, KDA50, DEQNA & CK KIT, DHV11 & CK KIT, TU81E, RA81, QZ002-C5 (VMS 8 USER), QZZCE-UZ (LAVc), QZD05-UZ (DECnet FF), 240 V AVAILABILITY Orders can be placed now for the new systems, with mid-October availability. SERVICE Digital's Field Service organization will continue to offer a wide variety of standard Hardware and Software Product Service maintenance agreements. This will allow you to meet your customer's needs for maintenance and support for both MicroVAX II hardware and software. MicroVAX II FAMILY PRODUCT COMPARISON Standard Systems Config.1 Config.2 Config.2. Config.3 Config.3 Config.4 Config.5 (RD53) (RD53) (RD54) (RD53) (RD54) (2xRD54) (RA81) U.S. List(1) Price ($K) 20.8 22.3 24.3 30.6 33.6 44.2 60.7 PERFORM. (X780) 0.9 0.9 0.9 0.9 0.9 0.9 0.9 I/O BUS Q22 Q22 Q22 Q22 Q22 Q22 Q22 O/S MicroVMS/ ----------------------------------------------------------> VMS ULTRIX-32 ----------------------------------------------------------> VAXELN ----------------------------------------------------------> ENCLOSURE BA23 BA23 BA23 BA123 BA123 BA123 H9642 MEMORY/MIN 5 MB 5 MB 5 MB 5 MB 5 MB 9 MB 16 MB MEMORY/MAX 16 MB 16 MB 16 MB 16 MB 16 MB 16 MB 16 MB No. of Lines(2) (modm cntrl) MIN 1 5 5 9 9 9 9 MAX 13 13 13 21 21 21 49 DISKS OFFERED(3) RD53 RD53 RD54 RD53/54 RD53/54 RD53/54 RA81,RA60 RD53/54 MIN. DISK CAPACITY 71 MB 71 MB 159 MB 71 MB 159 MB 318 MB 456 MB MAX. DISK CAPACITY 71 MB 71 MB 159 MB 389 MB 477 MB 477 MB 2 GB(4) BACKUP LAVc,RSM TK50 TK50 TK50 TK50 TK70 TK70,TU81+ TSV05 UNIQUE Entry- | Entry- | Entry- | Large | High- MESSAGES Level |------Level-------|------Level-------| Office | Capacity ENET | 4-User | Office | System | Standard Node | Stand-Alone | System | | System | System | | | 25+ | | | | Users (1) Base system, hardware only (2) Directly connected (3) Internal disks (4) Can support a total of four RA81s, with a second cabinet ================================================================================ Note 42.12 MICROVAX SYSTEM ANNOUCEMENTS 12 of 17 EISNER::TINKELMAN "Bob Tinkelman - CCA" 37 lines 2-AUG-1988 22:59 -< 04/25/88 Revised LAVC configurations >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! REVISED LOCAL AREA VAXcluster CONFIGURATIONS - 25-April-1988 ******************************************************************************* o LAVc Software V1.3 configurations have been updated to include several previously unsupported options o Available on line in early May These new configurations will help to make our offering more competitive in the low-end market. Since none of the changes requires actual code enhancements, but rather are based on testing and experience, there will not be a new release of LAVc. The revisions will be made to the SPD (27.65.04) only. The following lists the changes in the new revision: o The number of MicroVAX 3xxx-series processors has been increased to 10. This restriction is based on the number of MicroVAX 3xxx systems tested in an LAVc environment. Testing is continuing and the numbers will increase upon successful testing. o The VAX-11/750 has been added as a boot member for a maximum of two (2) MicroVAX 3xxx satellites. Note that an alternative configuration is to utilize both a VAX-11/750 and a MicroVAX 3xxx processor as (Class B) boot members. Performance may be improved by booting satellite members through the MicroVAX 3xxx only. The VAX-11/750 is considered a "boot member" by the presence of a local VMS system disk. o The VAXstation 2000 has been added as a boot member for a maximum of three VAXstation 2000 satellites. o Support for the DELQA Ethernet adapter has been added. ================================================================================ Note 42.13 MICROVAX SYSTEM ANNOUCEMENTS 13 of 17 EISNER::TINKELMAN "Bob Tinkelman - CCA" 107 lines 2-AUG-1988 22:59 -< 06/07/88 mVAX II cost of ownership reduced >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! mVAX II COST OF OWNERSHIP REDUCED - 07-June-1988 ******************************************************************************* BY UP TO $27,000 WITH NEW ONE-YEAR INTEGRATED HARDWARE AND SOFTWARE WARRANTY o Integrated hardware and software services for a single price o Customer realizes as much as 50% service cost reduction with integrated services o Integrated hardware and software services now uniform across the MicroVAX family Single-price integrated hardware and software service for the MicroVAX II, means Digital's customers now realize more than 50% service savings. Also included is an integrated one-year warranty covering both hardware and software as well as all Digital licensed software purchased during the first year following the initial system purchase. This emphasizes Digital's commitment to total solutions. PRODUCT DESCRIPTION The new service offering includes DECservice on-site hardware support during coverage hours, right to use/copy software updates, access to Digital Software Information Network (DSIN), and toll-free 24-hour, seven days a week telephone support for remedial and advisory service on both hardware and software. Media and documentation update services may be purchased separately once and used on all systems. There will be a modest additional charge for software layered product service in post-warranty years. These prices will be available shortly. Who Do You Sell To? There are two typical customer profiles for which this new offering will have significant importance. The first profile typifies the customer who normally wants hardware AND software services. This customer would realize a substantial savings in services. The second profile is for customers who only want hardware OR software services. Now this customer can purchase integrated (hardware and software) services at a price less than software services by itself. Refer to the "Five Year Cost of Ownership Table" for examples. No one can match the products full breadth of features: o A single point of contact for hardware and software service. o An advisory hotline available 24 hours a day, seven days a week, for both software and hardware. o The integrated one-year warranty including all Digital-licensed software and hardware. The new service solution for the MicroVAX II is consistent with the integrated services available for the MicroVAX 2000 and the MicroVAX 3500/3600, yet still allows for a la carte hardware and software service for those customers with unique needs. Five-Year Cost of Ownership Table (Example) Old U.S. New U.S. List Price List Price 8-User MicroVAX II: DH-630Q2-HA in BA23 configured w/hdw for 8 users. Hardware $ 29,488 $ 29,488 Software 8,962 8,962 (1) 5-year Hardware Service 14,352 -- (2) 5-year Software Service 21,717 -- (3) Integrated Service -- 17,232 ________ ________ TOTAL $ 74,519 $ 55,682 16-User MicroVAX II: DH-630Q4-JA in BA123 configured w/hdw for 16 users. Hardware $ 47,514 $ 47,514 Software 16,312 16,312 (1) 5-year Hardware Service 18,720 -- (2) 5-year Software Service 26,277 -- (3) Integrated Service -- 21,600 ________ ________ TOTAL $108,823 $ 83,426 32-User MicroVAX II: DH-630Q5-HA in H9642 configured w/hdw for 32 users. Hardware $ 66,319 $ 66,319 Software 21,562 21,562 (1) 5-year Hardware Service Price 19,440 -- (2) 5-year Software Service Price 30,267 -- (3) Integrated Service -- 22,320 ________ ________ TOTAL $137,588 $110,201 (1) First year hardware warranty and four-year hardware service price. (2) Five-year software service price. (3) First year hardware and software warranty and four-year integrated service price. Software for Digital systems is VMS, DECnet to support # users. Media and Documentation update is not included in integrated price, but it can be purchased separately once, and used on all systems. ================================================================================ Note 42.14 MICROVAX SYSTEM ANNOUCEMENTS 14 of 17 EISNER::TINKELMAN "Bob Tinkelman - CCA" 102 lines 2-AUG-1988 23:01 -< 05/09/88 VAXELN packaged rtVAX 3200/3500/3600 systems >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! rtVAX 3200/3500/3600 - 09-May-1988 ******************************************************************************* VAXELN PACKAGED SYSTEMS Three new rtVAX 3XXX realtime run-time only target systems for high-performance applications: o rtVAX 3200 BA23 Pedestal and Rack VAXELN target system o rtVAX 3500 BA213 Pedestal and Rack VAXELN target system o rtVAX 3600 BA213 Cabinet/VAXELN target system NEW FEATURES o CMOS technology has expanded the realtime MicroVAX family to three times MicroVAX II performance. o The new systems are one line item orderables that include both the basic system and the VAXELN runtime licenses. o These packages provide improvements in performance, packaging and storage options for realtime applications not presently available in the mid-range. o The complete rtVAX 3000 series meets the needs of customers who require stand-alone or embedded realtime applications. o The rack configurations can be embedded in customer enclosures; the cabinet configuration can accommodate a customer's own 19-inch rack- mountable equipment in an optional add-on stand-alone cabinet. PRODUCT DESCRIPTION rtVAX System Hardware rtVAX 3200: o The lowest entry price of the series; based on the BA23 packaging. o Includes the KA650-BA CPU board, the MS650 8 MB memory board, the DELQA Ethernet adapter and a VAXELN runtime license. o Reference rtVAX 1000 menu of Realtime Systems and Options Catalog for add-on information rtVAX 3500: o Is based upon the BA213 form factor. o Includes the KA650-BA CPU board, MS650, the DELQA Ethernet adapter and VAXELN runtime license. o Tape or disk may be added by order TQK70-SA/TK70E-SA or KDA50-SA/ RA70E-SA. Reference VAX 3500 menu in VAX Systems and Options Catalog for additional add-on information. rtVAX 3600: o Provides the cabinet configuration and rack mounting space in an optional add-on cabinet. o Includes the KA650-BA CPU board, 8 MB MS650, DELQA Ethernet adapter, the KDA50, RA82, TQK70/TK70 and VAXELN runtime license. POSITIONING o The rtVAX 3XXX series complements the rtVAX 1000 and rtVAX 8550/8700 product series. o VAXELN complements VMS in the realtime marketplace. o VAXELN promotes the rich VMS development environment for use in developing efficient, tailored, dedicated realtime applications. o The high degree of integration via networking and common file structure between VMS and VAXELN, combined with superior realtime performance, uniquely suit VAXELN along with the rtVAX 3XXX series for 32-bit realtime applications. o Markets: - Electronics - Oil and Gas - Federal Agencies - Food and Beverage - Defense Contractors - Chemical - Aerospace - Medical - Metals o Applications: - Computer integrated manufacturing (CIM) - Radar equipment - Work cell controllers - Strategic military programs - Continuous process controls - Satellite tracking - Batch processing - Space craft launching - Machine automation - Nuclear power plant simulators - Quality control systems - Medical imaging/diagnostics - Aircraft flight simulators - Jet engine testing - Wind tunnels - Ease of ordering - Superior price/performance - Efficient packaging - Superior functionality - Networking - Large complement of options ================================================================================ Note 42.15 MICROVAX SYSTEM ANNOUCEMENTS 15 of 17 EISNER::TINKELMAN "Bob Tinkelman - CCA" 32 lines 2-AUG-1988 23:02 -< 05/10/88 rtVAX price reductions >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! AGGRESSIVELY PRICED rtVAX REALTIME PRODUCTS ******************************************************************************* o Aggressive pricing at constant functionality increases business opportunities for: - rtVAX 1000, IVAX 620--for entry-level realtime applications - KA620 SBC--for embedded realtime applications o Unified rtVAX Family Discount Structure NEW PRICING To reinforce our strong commitment to the realtime market, Digital announces price reductions of approximately 30% to 40% and discount changes on the above products. CONFIGURATION MENUS o Configuration menus for the rtVAX 1000 and Industrial VAX 620 VAXELN target systems are published in the Realtime Systems and Options Catalog (RSOC). SUPPORT There is no change to the one-year hardware warranty for these systems. ================================================================================ Note 42.16 MICROVAX SYSTEM ANNOUCEMENTS 16 of 17 EISNER::TINKELMAN "Bob Tinkelman - CCA" 126 lines 2-AUG-1988 23:03 -< 06/13/88 New VAXstation 2000 configurations >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! NEW VAXstation 2000 CONFIGURATIONS - 13-June-1988 ******************************************************************************* o New 8-plane color VAXstation 2000 configurations o New 14 MB VAXstation 2000 configurations o 8-plane graphics accelerator option Digital's commitment to the workstation market is stronger than ever. To position Digital to become the leading vendor of workstation and workgroup solutions, we are pleased to announce the following: o 8-plane color VAXstation 2000 configuration with a 15" or 19" monitor. These configurations are available with 6 MB or 14 MB of memory. o 4-plane color 14 MB VAXstation 2000 configuration with a 15" or 19" monitor. PRODUCT DESCRIPTION Basic Configurations The base configuration for all Vaxstation 2000 products described in this announcement is as follows: o MicroVAX II CPU and FPU processing chip set, o 6 or 14 Megabytes of memory, o Disk controller, o Interface to keyboard, mouse, printer and modem, o ThinWire/Thickwire Ethernet controller, o T Connector and two terminators, o English hardware documentation, o One-Year hardware and software warranty. VMS-based systems include the following components: o VMS operating system, o VWS (See "Announcing VWS Workstation Software V3.3" in this issue.) o DECnet-VAX End-Node License, o Local Area VAXcluster License. ULTRIX-based systems include the following components: o ULTRIX operating system, o NFS, o ULTRIX C, VAX C, o ULTRIX FORTRAN, o ULTRIX PASCAL. Color VAXstation 2000 The color VAXstation 2000 uses a four-plane (16 colors) or eight-plane (256 colors) graphics subsystem built on the same graphics chip set used in the VAXstation II/GPX. The system supports a 15" or a 19" monitor. The resolution of both monitors is 1024 X 864. Using the same graphics subsystem as the VAXstation II/GPX ensures the same high performance graphics capabilities and total software compatibility across the VAXstation family of products. PRICING/ORDERING INFORMATION U.S. Discount Order No. Description List Price BMC EU OEM VS460-JA/KA VAXstation 2000, 19", 4PL color $17,025 $134 Y K 14 MB, diskless VS461-JA/KA VAXstation 2000, 15", 4PL color 13,025 113 Y K 14 MB, diskless VS470-EA/GA VAXstation 2000, 19", 8PL color 16,525 144 Y K 6 MB, diskless VS470-JA/KA VAXstation 2000, 19", 8PL color 20,025 144 Y K 14 MB, diskless VS471-EA/GA VAXstation 2000, 15", 8PL color 12,525 123 Y K 6 MB, diskless VS471-JA/KA VAXstation 2000, 15", 8PL color 16,025 123 Y K 14 MB, diskless VS40X-PA 8-Plane Graphics Accel. 5,000 N/A Y K E,J = VMS G,K = ULTRIX AVAILABILITY The availability of each product is summarized below. Model Availability 14 MB 8-plane color VAXstation 2000 July 1988 6 MB 8-plane color VAXstation 2000 July 1988 Model Availability 14 MB 4-plane color VAXstation 2000 July 1988 8-plane graphics ACC. September 1988 NOTES: 1. The eight-plane graphics accelerator (VS40X-PA) is available for the monochrome versions of the VAXstation 2000 for an additional charge of $5,000. Adding this option allows customers to display up to 256 shades of gray and provides significant graphics performance improvement. 2. These prices do not include the keyboard and pointing device (mouse or tablet) which must be ordered with each model. 3. For disk-based systems order a diskless system and the appropriate disks. 4. Actual diagonal measurements for the monitors are as follows: - VR160 14.3" (15" color monitor) - VR290 19.0" (19" color monitor) ================================================================================ Note 42.17 MICROVAX SYSTEM ANNOUCEMENTS 17 of 17 EISNER::TINKELMAN "Bob Tinkelman - CCA" 53 lines 4-AUG-1988 23:42 -< 05/09/88 DEC news: E-series industrial VAX/PDPs >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! THE E-SERIES FAMILY - 09-May-1988 ******************************************************************************* OF INDUSTRIAL VAX AND PDP-11 SYSTEMS o IVAX and IPDP packaged for the factory floor o Extended shock and vibration protection o Unique design cools passively without air conditioners or filters PRODUCT DESCRIPTION The E-Series are small industrial MicroVAX II rtVAX or industrial PDP-11 systems based on the BA214 6-slot chassis; they are packaged in two totally sealed enclosures -- CPU and mass storage -- and mount on wall, rack or floor. Designed to survive in dirty factory environments, these enclosures protect the systems from vibration and rough handling while their unique, passive cooling design permits them to withstand higher operating temperatures than most systems. The packaging, classified as NEMA-12, provides protection against hazards such as dirt, oil seepage, falling or splashing liquids, and metal chips. o Manufacturers who have applications in: - machine shops - automotive production lines - textile plants - pharmaceutical and other manufacturing industries What Do They Need? o High reliability for their manufacturing applications o Access to data from factory floor devices o Ease of maintenance What Do You Sell? o Sell compatibility -- existing VAX and PDP-11 applications can run without costly reprogramming. o Sell all MicroVAX and PDP-11 architecture and networking strengths. o Sell availability of application software. How Do You Win? o Position Digital as the only major computer vendor offering this level of industrial protection for its standard products.  ================================================================================ Note 43.0 VAX SYSTEM ANNOUNCEMENTS 25 replies EISNER::KILLEEN "Jeff Killeen" 2 lines 5-SEP-1987 12:15 -------------------------------------------------------------------------------- This topic will be used to discuss VAX hardward systems annoucements. VAX (non-MicroVAX) CPU's, system units, and memories. ================================================================================ Note 43.1 VAX SYSTEM ANNOUNCEMENTS 1 of 25 EISNER::KILLEEN "Jeff Killeen" 78 lines 5-SEP-1987 12:15 -< 8000 MEMORY CHANGES >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! NEW MEMORY PRICES FOR VAX 8800 AND VAX 8600 FAMILY - 20-August-1987 o Prices reduced up to 54% for VAX 8800 and VAX 8600 memory products o Memory options available now o New low prices for MS88-HA and MS85-HA upgrades benefit early VAX 8800 and VAX 8500 customers Announcing new, low prices for memory options for the VAX 8800 and VAX 8600 systems families. New price reductions apply to MS88-HA, MS85-HA upgrade packages and MS86-DA and MS86-CA memory arrays. Now you can offer cost-effective solutions to performance problems frequently incurred after customers add new applications to existing systems or include the system in a VAXcluster. PRODUCT DESCRIPTION The MS88-HA and MS85-HA, which include two 16-MByte memory modules plus a wider backplane, offer early VAX 8800 and VAX 8500 system owners an opportunity to upgrade their systems' backplanes and add more memory most cost effectively. These memory upgrades expand system capacities to 256 MBytes running VMS V4.6 and to 128 MBytes running ULTRIX-32 V2.0. The wider backplane accommodates the new 64-MByte module (MS88-DA) and the 16-MByte module (MS88-CA) in single slots. The MS86-DA and MS86-CA are 64-MByte and 16-MByte memory options for the VAX 8650/8600 systems. VAX 8800/8700/85xx customers can take advantage of the power of their systems by increasing memory at a savings of up to 54% compared to previous prices. MLPs for the MS88-HA and MS85-HA have been reduced from $48,000 to $22,000. And the wider backplane offered as part of the upgrade package allows customers to add the newly announced 64-MByte module (MS88-DA) in a single slot on the new backplane (see article in this issue entitled "Announcing New High-Density Memory Array for VAX 8800, VAX 8700 and VAX 85xx Series"). VAX 8600/8650 customers can order the 64-MByte and 16-MByte modules at a 30% savings. MLPs for the MS86-DA and MS86-CA have been reduced from $36,000 to $25,000 and from $10,500 to $7000 respectively. These new prices, combined with the one-year, on-site service offered with all memory products, give your customers an opportunity to decrease their overall cost-of-ownership and enhance their systems' performance. HARDWARE SERVICE All Digital memory products have a one-year warranty that includes on-site service. Additionally, during subsequent years, the Basic Monthly Charge (BMC) for memory is zero ($0) when the memory is installed on a system covered by a Digital Field Service contract. PRICING/ORDERING INFORMATION Option No. Description MLP BMC Inst. MS88-HA 32 MB ECC/256K W/BP VAX 8800/8700 $22,000 $0 $758 MS85-HA 32 MB ECC/256K W/BP VAX 85xx 22,000 0 675 MS86-DA 64 MB ECC/1MBIT VAX 8600/8650 25,000 0 578 MS86-CA 16 MB ECC/256K VAX 8600/8650 7,000 0 575 AVAILABILITY All memory options for the VAX 8800/8700/85xx and VAX 8650/8600 systems are available now. ================================================================================ Note 43.2 VAX SYSTEM ANNOUNCEMENTS 2 of 25 EISNER::KILLEEN "Jeff Killeen" 239 lines 5-SEP-1987 12:21 -< DEBNA >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! ANNOUNCING THE DEBNA - 27-August-1987 o VAXBI Ethernet adapter now available as an option o DEBNA also included with all VAXBI systems o Supports multiple Ethernets for VAXBI VMS systems o VAXBI LAVc system boot and IEEE 802.3 supported The DEBNA is a high-performance network adapter that connects VAXBI systems to both Ethernet and IEEE 802.3 local area networks. This adapter allows your customers to add multiple Ethernets to their VAXBI systems and to connect VAXBI systems to Local Area VAXclusters or IEEE 802.3 local area networks. It can also be used to upgrade existing UNIBUS-based Ethernets to direct VAXBI connections for better performance and reliability. One DEBNA is now included with each new VAXBI system. PRODUCT DESCRIPTION The DEBNA is a single VAXBI module that supports one Ethernet port, providing the physical and data-link communication layers. It provides up to 5 megabits per second peak hardware throughput capability (achieved throughput will depend on system configuration, packet sizes and application). The DEBNA implements the complete Ethernet V2.0 protocol and is compatible with the IEEE 802.3 specification. FEATURES/BENEFITS Feature Benefit o Single VAXBI Module o Requires only one VAXBI backplane slot o 5 Mbit/Second peak o Increased system throughput hardware performance o Multiple Ethernets per o Greater system flexibility VAXBI VMS system supported o LAVc Support o VAXBI VMS System can be a boot member in LAVc; no UNIBUS is required o Ethernet V2.0 o Multivendor communications 802.3 IEEE Compatible on VAXBI sytems with VAX OSI Transport installed o Downline loading o 82** and 83** VMS systems can be downline loaded by other VAX systems via the LAN GENERAL CONFIGURATION INFORMATION The DEBNA-M option consists of the base VAXBI module only, which must be connected to one of the CK-DEBNA-L* Cabinet Kit options. Each DEBNA-M module option requires ordering a separate CK-DEBNA-L* Cabinet Kit option, which consists of an internal Ethernet/802.3 compatible cable, an I/O connector panel, and a boot-enable jumper for Ethernet. This physically and electrically connects to the Ethernet via transceiver cable and one of the following separately orderable Ethernet options: - H4000 baseband Ethernet transceiver - DECOM broadband Ethernet transceiver - DELNI local network interconnect Transceiver cable (BNE3X or BNE4X series) can be a maximum of 40 meters (131 feet) and must also be ordered separately. Each Ethernet controller must be connected to a separate Ethernet, with one exception. Two Ethernet controllers can be connected to the same wire if only one is active. The inactive Ethernet controller is on standby in case the first device fails. This is useful in applications where uptime is critical. Two Ethernet controllers cannot dynamically load share the activities of one Ethernet. SYSTEM-SPECIFIC CONFIGURATION INFORMATION VAX 8700/8800 systems may be configured to a maximum of three DEBNA-M per VAXBI channel and a maximum of four DEBNA-M per system (including the DEBNA shipped with the system). The maximum of four connections may be made to an H4000, DELNI or DECOM from internal VAXBI channels. External VAXBI channels are limited to one H4000 connection per channel; additional connections require a port on a DELNI or DECOM. Order the CK-DEBNA-LJ Cabinet Kit when the DEBNA-M is placed on an internal VAXBI channel; order the CK-DEBNA-LN Cabinet Kit when the DEBNA-M is placed on an external VAXBI channel. VAX 8500/8530/8550 systems may be configured to a maximum of three DEBNA-M per VAXBI channel and a maximum of four DEBNA-M per system (including the DEBNA shipped with the system). A maximum of two connections may be made to an H4000 from internal VAXBI channels; external VAXBI channels are limited to one H4000 connection per channel. Additional connections require a port on a DELNI or DECOM and cannot be connected directly to an H4000. Order the CK-DEBNA-LJ Cabinet Kit when the DEBNA-M is placed on an internal VAXBI channel; order the CK-DEBNA-LN Cabinet Kit when the DEBNA-M is placed on an external VAXBI channel. VAX 8250/8350 Configuration 2 systems may be configured to a maximum of two DEBNA-M per system (including the DEBNA shipped with the system). Both DEBNA options may be connected to either H4000, DELNI or DECOM. Order the CK-DEBNA-LM Cabinet Kit for all VAX 8250/8350 systems. VAX 8250/8350 Configuration 1 systems may be configured to a maximum of two DEBNA-M per system (including the DEBNA shipped with the system). One connection may be made to an H4000; the additional connection must be made to a DELNI or DECOM. Order the CK-DEBNA-LM Cabinet Kit for all VAX 8250/8350 systems. The maximum number of DEBNAs may be less than listed above depending on DC power availability and/or number of empty VAXBI slots in the system. For systems which include a UNIBUS, the maximum number of combined DEUNA/DELUA/DEBNA's per system cannot exceed the maximum number of DEBNA's listed above. POWER REQUIREMENTS Option Mounting dc Amps Drawn @ VAXBI I/O Panel Requirements Nodes Units ------------------------------------------------------------------------ 5V 12V -12V -2V -5.2V ------------------------------------------------------------------------ DEBNA-M 1 VAXBI Slot 6.72 0.00 0.00 0.00 0.00 1 none CK-DEBNA-LJ none 0.00 0.50 0.00 0.00 0.00 0 1 CK-DEBNA-LM none 0.00 0.00 0.00 0.00 0.00 0 2 CK-DEBNA-LN none 0.00 0.00 0.00 0.00 0.00 0 2 SOFTWARE SUPPORT O / S Version Functions Supported VMS V4.5 DECnet, LAT V4.6 DECnet, LAT, Multiples, LAVc, IEEE 802.3, VOTS Maximum: See System-Specific Configuration Information ULTRIX V2.0 TCP/IP, NFS, DECnet (End Node only), LAT Maximum: One Ethernet per system. VAXELN V2.4 DECnet Maximum: One Ethernet per system. The DEBNA is supported under DECnet Phase IV and 802.3 Software. The new functions supported by the DEBNA (Multiples, 802.3 and LAVc) require VMS Version 4.6. Refer to the Software Product Descriptions for VMS, LAT, DECnet/VAX, LAVc, VAXELN and ULTRIX for software configuration information. LOCAL AREA VAXcluster PERFORMANCE The DEBNA allows all VAXBI systems to be used as boot members in Local Area VAXcluster systems. Both VAXBI systems and MicroVAX systems are now supported as LAVc boot members; however, the number of satellite nodes which can be supported will vary by system. In a fully loaded LAVc, replacement of a MicroVAX II boot member by a VAX 85XX or VAX 8700/8800 system can add significant compute power to the LAVc. The VAX 85XX or VAX 8700/8800 system can enhance the overall LAVc performance and the number of possible satellites. If a MicroVAX II boot member is replaced by a VAX 82XX or VAX 83XX system, 30-40% fewer satellite nodes may be supported with an equivalent level of performance. However, VAX 82XX/83XX systems offer benefits of larger system expandability including support of large numbers of local disk and tape drives, as well as support for multiple Ethernets. NOTE: LAVc will operate over a single Ethernet only. Use these guidelines to ensure that you select the system which will best meet your customer's requirements for LAVc configurations. UPGRADES The DEBNT (T1032) is not compatible with the DEBNA (T1034). When a DEBNA add-on option is installed by Digital Field Service, and the system already has a DEBNT, Field Service will replace the DEBNT with a DEBNA at no charge. Customers with systems that contain the DEBNT will be upgraded to the DEBNA at no charge beginning in Q2. These upgrades will be completed by the end of Q4/FY88. Customers needing to upgrade to the 802.3 IEEE Electrical Standard will need to purchase the cabinet kits listed in this article. Cables shipped with the DEBNT do not comply electrically with the 802.3 Standard, and are not included in the DEBNA "no charge" upgrade. Systems with UNIBUS-based DELUAs or DEUNAs can be upgraded by having the customer purchase a DEBNA, but will be charged for the upgrade. FIELD SERVICE SUPPORT The DEBNA includes one-year on-site warranty. Field Service will maintain the DEBNA for contract customers through the normal on-site Basic and DECservice. Off-site service for self-maintenance customers will be provided by "Per-Call" and the DECmailer program. Remote Services will be provided for both warranty and contract customers. PRICING/ORDERING INFORMATION Order No. Description MLP BMC DEBNA-M VAXBI Ethernet Adapter $3,900 $ 30 Requires one CK-DEBNA-L* CK-DEBNA-LJ Cabinet Kit W/5 Ft Cables for 300 N/C VAX 8500, 8530, 8550, 8700, 8800 (without H9652-EC/ED Cabinets) CK-DEBNA-LM Cabinet Kit W/8 Ft Cables for 400 N/C VAX 8200, 8250, 8300, 8350 CK-DEBNA-LN Cabinet Kit W/15 Ft Cables for 500 N/C VAX 8500, 8530, 8550, 8700, 8800 H9652-EC/ED Expansion Cabinet AVAILABILITY The DEBNA add-on option is available in limited quantities in this quarter, with volume availability beginning in Q2. Each VAXBI system contains a VAXBI Ethernet adapter. The initial version of that adapter was the T1032 module, also known as the DEBNT. The DEBNT was never shipped as an add-on option, and has been phased out of systems. Systems shipped after the dates listed below have the DEBNA (T1034): VAX 8700, 8800 shipments after June 1, 1987. VAX 8250, 8350, 8530, 8550 shipments after July 1, 1987. ================================================================================ Note 43.3 VAX SYSTEM ANNOUNCEMENTS 3 of 25 EISNER::KILLEEN "Jeff Killeen" 160 lines 5-SEP-1987 14:24 -< 8250/8350 SYSTEMS >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! ANNOUNCING ENHANCED VAX 8250 AND VAX 8350 SYSTEMS - 04-SEPTEMBER-1987 o New configurations with increased memory o Increase in maximum memory supported to 128 MB o Systems can be ordered now for availability within 30 days o Previous configurations placed in Maintenance-Only New VAX 8250/8350 system configurations with increased memory are now available. With these new configurations, your customers can run their applications more efficiently, and have the capability to greatly expand the memory in their systems to support applications requiring very large memory support. The VAX 8350 standard memory has been increased to 32 MB. This is especially significant in that it enhances the VAX 8350 positioning as an extremely cost-effective VAXBI and VAXcluster system. You should attempt to convert all VAX 8250 customers to the VAX 8350 system to obtain the higher performance for a small incremental cost. The VAX 8350 is the entry-level VAXBI system of choice. PRODUCT DESCRIPTION The following improvements have been made to the entry-level VAXBI systems: o New VAX 8250/8350 configurations are now available which incorporate new 16 MB memory modules. Each of the new VAX 8350 configurations provide 32 MB as standard memory content and the VAX 8250 configurations provide 16 MB. The increased memory is provided with minimal change in price, making these systems even more attractive. System Imbedded Memory Increase VAX 8350 Systems/Rack-mount Boxes 12 MB --> 32 MB VAX 8250 Systems 8 MB --> 16 MB VAX 8250 Rack-mount Boxes 4 MB --> 16 MB o Maximum memory for VAX 8250/8350 systems is increased four-fold up to 128 MB based on the new 16 MB memory modules. This amount of memory is unprecedented for systems in this price and performance range. System Maximum Memory Increase VAX 8250/8350 Configuration 1 24 MB --> 96 MB VAX 8250/8350 Configuration 2, without Battery Backup 32 MB --> 128 MB VAX 8250/8350 Configuration 2, with Battery Backup 24 MB --> 96 MB o With the recent announcement of the DEBNA Ethernet interface, VAX 8250/8350 systems can support up to two separate Ethernets attached directly to the VAXBI. Examples of using two Ethernets might be as secure and non-secure networks, computer and terminal networks, or broadband and baseband networks. CONFIGURATION INFORMATION Complete menus for these new VAX 8250/8350 configurations will be provided in the Q2 VAX Systems and Options Catalog. In the interim, contact your digital Representative for help with specific configurations. SOFTWARE SUPPORT VAX 8250/8350 systems with the new 16 MB memory modules are supported by VMS V4.6 and ULTRIX V2.0. PRICING/ORDERING INFORMATION The enhanced VAX 8250/8350 system configurations directly replace the previous configurations. However, orders for the previous configurations will not be automatically converted to the new part numbers. The previous configurations are being placed in Maintenance-Only and should no longer be ordered. VAX 8350 Systems New Model No. Description U.S. List Price | Old Model No. SV-8BEEB-GL/GM VAX 8350/1 PRECONF SYS PPL $148,050 | SV-8BEEA-GL/GM SV-8BEEB-GK/GN VAX 8350/1 PRECONF SYS PDUP 173,250 | SV-8BEEA-GK/GN | 834CC-DP/DT VAX 8350/1 CLUSTER SYS PPL 102,900 | 834CB-DP/DT 834CC-AP/AT VAX 8350/1 CLUSTER SYS PDUP 128,100 | 834CB-AP/AT | 834BB-DE/DJ VAX 8350/1 VMS SYS PPL 98,700 | 834BA-DE/DJ 834BB-AE/AJ VAX 8350/1 VMS SYS PDUP 123,900 | 834BA-AE/AJ | 834BB-EE/EJ VAX 8350/1 ULT SYS PPL 98,700 | 834BA-EE/EJ 834BB-BE/BJ VAX 8350/1 ULT SYS PDUP 123,900 | 834BA-BE/BJ | SV-8BEES-GL/GM VAX 8350/2 PRECONF SYS PPL 158,550 | SV-8BEEN-GL/GM SV-8BEES-GK/GN VAX 8350/2 PRECONF SYS PDUP 183,750 | SV-8BEEN-GK/GN | 835CC-DP/DT VAX 8350/2 CLUSTER SYS PPL 108,150 | 835CB-DP/DT 835CC-AP/AT VAX 8350/2 CLUSTER SYS PDUP 133,350 | 835CB-AP/AT | 835BB-DE/DJ VAX 8350/2 VMS SYS PPL 103,950 | 835BA-DE/DJ 835BB-AE/AJ VAX 8350/2 VMS SYS PDUP 129,150 | 835BA-AE/AJ | 835BB-EE/EJ VAX 8350/2 ULT SYS PPL 103,950 | 835BA-EE/EJ 835BB-BE/BJ VAX 8350/2 ULT SYS PDUP 129,150 | 835BA-BE/BJ New Model No. Description U.S. List Price | Old Model No. 8BXBB-DE/DJ VAX 8350 VMS 12-SLOT BOX PPL * | 8BXBA-DE/DJ 8BXBB-AE/AJ VAX 8350 VMS 12-SLOT BOX PDUP * | 8BXBA-AE/AJ | 8BXBB-EE/EJ VAX 8350 ULT 12-SLOT BOX PPL * | 8BXBA-EE/EJ 8BXBB-BE/BJ VAX 8350 ULT 12-SLOT BOX PDUP * | 8BXBA-BE/BJ VAX 8250 Systems SV-8AEEB-GL/GM VAX 8250/1 PRECONF SYS PPL $121,800 | SV-8AEEA-GL/GM SV-8AEEB-GK/GN VAX 8250/1 PRECONF SYS PDUP 141,750 | SV-8AEEA-GK/GN | 824CC-DP/DT VAX 8250/1 CLUSTER SYS PPL 76,650 | 824CB-DP/DT 824CC-AP/AT VAX 8250/1 CLUSTER SYS PDUP 96,600 | 824CB-AP/AT | 824BC-DE/DJ VAX 8250/1 VMS SYS PPL 72,450 | 824BB-DE/DJ 824BC-AE/AJ VAX 8250/1 VMS SYS PDUP 92,400 | 824BB-AE/AJ | 824BC-EE/EJ VAX 8250/1 ULT SYS PPL 72,450 | 824BB-EE/EJ 824BC-BE/BJ VAX 8250/1 ULT SYS PDUP 92,400 | 824BB-BE/BJ | SV-8AEEP-GL/GM VAX 8250/2 PRECONF SYS PPL 132,300 | SV-8AEEN-GL/GM SV-8AEEP-GK/GN VAX 8250/2 PRECONF SYS PDUP 152,250 | SV-8AEEN-GK/GN | 825CD-DP/DT VAX 8250/2 CLUSTER SYS PPL 81,900 | 825CB-DP/DT 825CD-AP/AT VAX 8250/2 CLUSTER SYS PDUP 101,850 | 825CB-AP/AT | 825BB-DE/DJ VAX 8250/2 VMS SYS PPL 77,700 | 825BA-DE/DJ 825BB-AE/AJ VAX 8250/2 VMS SYS PDUP 97,650 | 825BA-AE/AJ | 825BB-EE/EJ VAX 8250/2 ULT SYS PPL 77,700 | 825BA-EE/EJ 825BB-BE/BJ VAX 8250/2 ULT SYS PDUP 97,650 | 825BA-BE/BJ | 8AXBD-DE/DJ VAX 8250 VMS 12-SLOT BOX PPL * | 8AXBB-DE/DJ 8AXBD-AE/AJ VAX 8250 VMS 12-SLOT BOX PDUP * | 8AXBB-AE/AJ | 8AXBD-EE/EJ VAX 8250 ULT 12-SLOT BOX PPL * | 8AXBB-EE/EJ 8AXBD-BE/BJ VAX 8250 ULT 12-SLOT BOX PDUP * | 8AXBB-BE/BJ | 8AXBC-DE/DJ VAX 8250 VMS 6-SLOT BOX PPL * | 8AXBA-DE/DJ 8AXBC-AE/AJ VAX 8250 VMS 6-SLOT BOX PDUP * | 8AXBA-AE/AJ | 8AXBC-EE/EJ VAX 8250 ULT 6-SLOT BOX PPL * | 8AXBA-EE/EJ 8AXBC-BE/BJ VAX 8250 ULT 6-SLOT BOX PDUP * | 8AXBA-BE/BJ ================================================================================ Note 43.4 VAX SYSTEM ANNOUNCEMENTS 4 of 25 EISNER::KILLEEN "Jeff Killeen" 140 lines 5-SEP-1987 14:25 -< 8530,8550,8700,8800 SYSTEMS >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! ENHANCED VAX 8530, 8550, 8700 AND 8800 SYSTEMS - 04-September-1987 o New configurations with increased memory o Smaller footprint for VAX 8530, 8550, 8700 and 8800 o Systems can be ordered now for availability within 30 days o Previous configurations placed in Maintenance-Only VAX 8530 systems are now significantly improved with increased memory content. Also, VAXcluster configurations of the VAX 8530 incorporate the CIBCA interface, allowing for smaller footprint systems. These changes solidly position the VAX 8530 as a high-performance system providing large system expandability in a small footprint. At the same time, new VAX 8550/8700/8800 system configurations are now being announced which provide increased memory content. With these new configurations, your customers obtain more memory to run their applications more efficiently. VAX 8550 systems also benefit from VAXcluster improvements of a smaller footprint. PRODUCT DESCRIPTION The following improvements have been made to these high-end VAX systems: o Each of the high-end VAX system configurations are being improved by the inclusion of an additional 16 MB memory module. System Imbedded Memory Increase VAX 8530 16 MB --> 32 MB VAX 8550 32 MB --> 48 MB VAX 8700 32 MB --> 48 MB VAX 8800 48 MB --> 64 MB o Maximum memory for VAX 8530/8550/8700/8800 systems has been increased substantially to 256 MB based on the announcement of the new 64 MB memory modules. This amount of memory is especially impressive when compared to IBM systems in this performance range. System Maximum Memory Increase VAX 8530 80 MB --> 256 MB VAX 8550 80 MB --> 256 MB VAX 8700 128 MB --> 256 MB VAX 8800 128 MB --> 256 MB o VAX 8530/8550 VAXcluster systems have been recently improved with the inclusion of the CIBCA VAXcluster interface. This new interface consists of only two VAXBI boards and does not require the additional cabinet included with the original VAXcluster interface. Thus, the VAXcluster system footprint is reduced by 50%, an especially important consideration for your customers with limited computer room space. VAX 8530/8550 systems provide exceptional VAXcluster engines by delivering high performance in a very small footprint. o Using the recently announced DEBNA interface, it is now possible to attach up to four Ethernets directly to the VAXBI to obtain increased performance and maximum connectivity. Also, the DEBNA can be used to connect the systems directly to a Local Area VAXcluster in a very cost-effective manner. This especially enhances the positioning of the VAX 8530/8550 systems as high-performance LAVc boot nodes. CONFIGURATION INFORMATION Complete menus for these new VAX 8530/8550/8700/8800 configurations will be provided in the Q2 VAX Systems and Options Catalog. In the interim contact your Digital Representative for help with specific configurations. PRICING/ORDERING INFORMATION The enhanced VAX 8530/8550/8700/8800 system configurations directly replace the previous configurations. However, orders for the previous configurations will not be automatically converted to the new part numbers. The previous configurations are being placed in Maintenance-Only and should no longer be ordered. VAX 8800 Systems New Model No. Description U.S. List Price | Old Model No. 882CC-DP/DT VAX 8800 CLUSTER SYS PPL $833,700 | 882CB-DP/DT 882CC-AP/AT VAX 8800 CLUSTER SYS PDUP 885,150 | 882CB-AP/AT | 882BD-EE/EJ VAX 8800 ULTRIX SYS PPL 821,100 | 882BC-EE/EJ 882BD-BE/BJ VAX 8800 ULTRIX SYS PDUP 872,550 | 882BC-BE/BJ VAX 8700 Systems SV-87EEC-EL/EM VAX 8700 PRECONF SYS PPL $593,250 | SV-87EEB-EL/EM SV-87EEC-EK/EN VAX 8700 PRECONF SYS PDUP 641,550 | SV-87EEB-EK/EN | 871CC-DP/DT VAX 8700 CLUSTER SYS PPL 562,800 | 871CB-DP/DT 871CC-AP/AT VAX 8700 CLUSTER SYS PDUP 615,300 | 871CB-AP/AT | 871BD-DE/DJ VAX 8700 VMS SYS PPL 543,900 | 871BC-DE/DJ 871BD-AE/AJ VAX 8700 VMS SYS PDUP 592,200 | 871BC-AE/AJ | 871BD-EE/EJ VAX 8700 ULTRIX SYS PPL 543,900 | 871BC-EE/EJ 871BD-BE/BJ VAX 8700 ULTRIX SYS PDUP 592,200 | 871BC-BE/BJ VAX 8550 Systems SV-8EEED-EL/EM VAX 8550 PRECONF SYS PPL $506,100 | SV-8EEEB-EL/EM SV-8EEED-EK/EN VAX 8550 PRECONF SYS PDUP 554,400 | SV-8EEEB-EK/EN | 855CD-DP/DT VAX 8550 CLUSTER SYS PPL 474,600 | 855CC-DP/DT 855CD-AP/AT VAX 8550 CLUSTER SYS PDUP 528,150 | 855CC-AP/AT 855BC-DE/DJ VAX 8550 VMS SYS PPL $456,750 | 855BB-DE/DJ 855BC-AE/AJ VAX 8550 VMS SYS PDUP 506,100 | 855BB-AE/AJ | 855BC-EE/EJ VAX 8550 ULTRIX SYS PPL 456,750 | 855BB-EE/EJ 855BC-BE/BJ VAX 8550 ULTRIX SYS PDUP 506,100 | 855BB-BE/BJ VAX 8530 Systems New Model No. Description U.S. List Price | Old Model No. SV-85EEC-EL/EM VAX 8530 PRECONF SYS PPL $349,650 | SV-85EEB-EL/EM SV-85EEC-EK/EN VAX 8530 PRECONF SYS PDUP 390,600 | SV-85EEB-EK/EN | 851CD-DP/DT VAX 8530 CLUSTER SYS PPL 323,400 | 851CC-DP/DT 851CD-AP/AT VAX 8530 CLUSTER SYS PDUP 363,300 | 851CC-AP/AT | 851BC-DE/DJ VAX 8530 VMS SYS PPL 301,350 | 851BB-DE/DJ 851BC-AE/AJ VAX 8530 VMS SYS PDUP 342,300 | 851BB-AE/AJ | 851BC-EE/EJ VAX 8530 ULTRIX SYS PPL 301,350 | 851BB-EE/EJ 851BC-BE/BJ VAX 8530 ULTRIX SYS PDUP 342,300 | 851BB-BE/BJ ================================================================================ Note 43.5 VAX SYSTEM ANNOUNCEMENTS 5 of 25 EISNER::KILLEEN "Jeff Killeen" 97 lines 5-SEP-1987 14:27 -< 82XX/8300 MEMORY ARRAYS >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! ANNOUNCING NEW HIGH-DENSITY 16-MBYTE MEMORY ARRAY - 04-September-1987 FOR VAX 82xx/83xx SYSTEMS o Increases VAX 82xx/83xx system capacities from 32 MBytes to 128 MBytes o Supports battery backup up to 96 MBytes o On-board control logic ensures high system availability o Price reduced on the 4-MByte MS820-BA Now your customers can increase the memory content of their VAX 82xx/83xx systems and achieve optimum system performance. Digital's newest memory array, the MS820-CA, offers four times the density of its predecessor and increases system capacity on systems with 24 slots to 128 MBytes running VMS V4.6 or ULTRIX-32 V2.0. Capacities of systems with 12 slots increase to 96 MBytes. Each module requires only one slot on the backplane. On-board control logic provides redundancy and, therefore, a higher degree of reliability when more than one MS820-CA is installed. Battery backup is supported as an option up to 96 MBytes. This feature ensures that no data in memory is lost as a result of a power failure. Coincident with this new product announcement is the repricing of the MS820-BA, which represents a 33% decrease. This new price benefits customers who require less than 8 MBytes of additional memory. PRODUCT DESCRIPTION The MS820-CA, a 16-MByte VAXBI memory array, uses 1-Mbit DRAMS with an on-board memory controller. The memory interfaces to the KA820 CPU via the VAXBI interconnect. The 16-MByte module is similar to the existing 4-MByte VAXBI memory module (MS820-BA), provides the same system performance, and requires only one slot on the backplane. The 16-MByte VAXBI memory array is compatible with the 4-MByte and 2-MByte memory modules and requires no hardware changes. The MS820-BA is a 4-MByte, 256K DRAMS-based VAXBI memory with an on-board controller. This module is currently shipping on the VAX 8250 and VAX 8350 systems. CONFIGURATION GUIDELINES Battery backup supports up to six slots on the system backplane. These slots can be filled with any combination of 16-MByte, 4-MByte or 2-MByte modules, but is limited to 96 MBytes of memory. Without battery backup, up to eight slots can be filled with a mix of 16-MByte, 4-MByte or 2-MByte boards, up to a maximum of 128 MBytes on a system with 24 slots (Configuration 2) and 96 MBytes, or six slots, on a 12-slot system (Configuration 1). Power Requirements Current Amps Used for the VAXBI Voltages: Option +5 +5B +12 -12 -5.2 -2 MS820-CA 0.66 2.50* 0.00 0.00 0.00 0.00 * During self-test, all MS820-CA memory modules simultaneously draw 2.50 amps of +5B. During normal operation, only one MS820-CA can be active and all other MS820-CA modules are in standby mode. When active, an MS820-CA draws 3.80 amps of +5B. In standby mode, an MS820-CA draws 1.90 amps of +5B. HARDWARE SERVICE In addition to the twelve months of on-site service offered through the product warranty, customers with a Digital Field Service agreement for VAX 82xx/83xx systems continue to receive free on-site service for their memory products after year one. Installation for add-on memory is available for a nominal fee. PRICING/ORDERING INFORMATION The MS820-CA, 16-MByte memory module is priced at a lower price per megabyte than the MS820-BA 4-MByte array. Option No. Description U.S. List Price BMC Inst. MS820-CA 16MB ECC 1MBIT VAXBI Memory $8,000 $0 $465 MS820-BA 4MB ECC 256K VAXBI Memory 4,000 0 495 AVAILABILITY The MS820-CA and MS820-BA memory arrays are available now. ================================================================================ Note 43.6 VAX SYSTEM ANNOUNCEMENTS 6 of 25 EISNER::KILLEEN "Jeff Killeen" 241 lines 5-SEP-1987 14:28 -< TA79 AND TU79 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! TA79 AND TU79 MAGNETIC TAPE SUBSYSTEM - 02-September-1987 o A shared resource with enhanced performance to complement VAXcluster systems o Significantly improved read/write data reliability o Higher subsystem reliability and availability o Direct replacement for the TA78 and TU78 product set Effective immediately, the TA/TU78 product set is being retired and replaced by the TA/TU79 product set. The TA/TU79, a direct replacement in form and function for the TA/TU78, will be substituted for all open orders of the TA/TU78 product. PRODUCT DESCRIPTION The TA/TU79 magnetic tape sub- system is a significantly redesigned and repackaged successor to the popular TA/TU78 tape subsystem. The TA/TU79 is engineered for high reliability and offers dramatically enhanced reliability and data integrity as compared to its predecessor. The TA/TU79 features ANSI Stan- dard 9-track, dual-density Phase Encoded (PE, 1600 bits/in) and Group Code Recording (GCR, 6250 bits/in) on 1/2-inch tape at 125 in/s, Start/Stop performance. The TA/TU79, the high-performance tape subsystem offering from Digital for customers with tape-intensive applications, contains one formatter, one TU79 tape transport and its own power supply in a single cabinet. It is also the primary tape companion for the RA81 and SA482 disk drives on the HSC Mass Storage Server. Digital's HSC-based tape subsystem, the TA79 offers significant advantages to VAXcluster system users. This high-performance, top-of-the-line subsystem has incorporated major enhancements for reliability and operator convenience. The subsystem acts as a shared resource, available to users anywhere in the VAXcluster. Assuming that a VAXcluster is configured with multiple tape drives, this unique accessibility provides redundancy for increased availability and increases the likelihood of finding a tape drive free when it is needed. There are performance benefits as well. By attaching the tape subsystems to an HSC Storage Server, you can take advantage of the controllers' Local Backup Utility, which shortens backup time significantly. Both TA/TU79 products offer the following benefits: o Outstanding data integrity with read-after-write verification and automatic error detection and correction during operation. o Easy to use because they provide automatic tape threading and loading and conveniently located operator controls. o Conform to the ANSI standard for Group Code Recording (GCR) (6,250 bits/in) and for Phase Encoded (PE) recording (1,600 bits/in) on half-inch nine-track tape. o Investment protection; upgrading TU79 Master to TA79 tape subsystems can be done on site in one day or less. o Include a one-year warranty. As a dual-density drive, both subsystems can read and write industry- standard tapes for interchange with other tape subsystems -- both Digital's and those of other manufacturers. When recording in GCR format, these subsystems have a capacity of 145 megabytes per reel, using 8-kilobyte records; maximum capacity is 167 megabytes, using 64-kilobyte records. The sophisticated start/stop design of the TA/TU79 provides the flexibility to handle a wide variety of applications with ease. Because of the start/ stop design, it is well suited for journaling and transaction processing applications and provides maximum performance in backup operation as well. SUMMARY OF IMPROVEMENTS The TA/TU79 product is the culmination of a long-term reliability program on the TA/TU78 product. The TA/TU79 has significantly improved read/write data reliability, higher subsystem reliability and availability, and reduced maintenance and service down-time requirements to increase drive availability. The most important improvement made in the new TA/TU79 product is to the drive's read/write electronics which now EXCEED ANSI Standards for variation in signal amplitude when reading tapes. What this means to your customer is: o Fewer "soft" or correctable errors will be reported by the subsystem to the VAX host. o The drive will be more tolerant when attempting to read tapes that have marginal signal amplitude due to poor recording, extended shelf storage, or contamination. o The need for tape head adjustments and cleaning will be lessened. These changes may be transparent to many customers, but will be of value to accounts that have tape-intensive applications. The plastic front door has been replaced by a heavy-duty metal front door, which incorporates a sliding front window. This modification will provide the following benefits: o Operators will find the drive easier to use o Wear-out problems associated with the plastic assembly will be eliminated o Reduced contamination due to improved tape path sealing The card cage has been repositioned within the cabinet to improve air flow and reduce internal temperatures. This increases the drive's tolerance to temporary variations in computer room temperature. Again, this change will be unnoticed by most customers but is representative of Digital's continuing commitment to improve quality and reliability. CONFIGURATIONS Masters and Slaves -- A TA/TU79 master tape subsystem contains one formatter and one TU79 tape transport, packaged in a single cabinet. A maximum of three additional TU79 slave transports can be attached to each TA/TU79 master, connecting to the formatter in a radial fashion. As many as three TU79 slave tape transports can be attached to each TA79 or TU79 master tape subsystem. These add-on drives are functionally identical to the TU78 slave tape drives now in the field. This protects a customer's investment by allowing them to connect any TU78 slave they now have to the newer TA/TU79 master tape subsystems and also allows them to attach the new TU79 slave onto an existing TA/TU78 string. Dual Porting -- The TA79 system includes two ports and is supplied with two cables for connection to HSC Servers. Dual ports are provided to facilitate rapid reconfiguration should an HSC malfunction. However, only one port should be enabled at any one time. The two ports are not intended for functions such as simultaneous access through multiple HSCs or for automatic failover. If customers already have an HSC Server but do not yet have clustered tape units, they must order an HSC5X-CA tape channel module with their first TA79 tape subsystem. Although each module can support four tape units, they may prefer to add additional modules as their tape farm grows, in order to maximize throughput and increase availability. TA79's can be attached to the same HSC5X-CA module. At altitudes above 610 meters (2,000 feet), the TA79 subsystem must have a high-altitude kit installed. Contact your local Field Service office to arrange for installation. OPERATING SYSTEM SUPPORT The TA79 Master is fully supported in VMS Version 4.6 and HSC Microcode level 350. The TU79 Massbus Master is supported under current ULTRIX and VMS releases as a TU78. Full support for the TU79 device nomenclature is planned in future releases of VMS and ULTRIX. Please refer to the respective Operating System Release Notes for more details. The TA79 and TU79 tape subsystems support all standard VMS tape utilities. Several VMS switches can also be used to optimize the performance of VMS Backup without compromising data integrity, when operating at 6250 bits/in density (GCR mode). The recommended settings used to obtain the maximum performance are: Buffer = 5/NOCRC/Block Size = 16 kilobytes. TU79 TO TA79 UPGRADES TA79 upgrade kits protect storage system investments. Customers can buy a stand-alone VAX system now with a TU79 Master, upgrade to a VAXcluster system later, and retain and use the original tape subsystem peripherals. All that needs to be done is to install an upgrade kit to the TU79 master subsystems. The kit contains all that is needed to upgrade a TU to a TA79 configuration quickly and easily.* It takes less than one working day to install a kit. * If the present system configuration has no HSC-based tape subsystems but does have an HSC Server, the HSC5X-CA tape interface module must be ordered for the first TA subsystem. PRICING/ORDERING INFORMATION Option No. Description U.S. List Price BMC Add-on Inst. TA79-BF Dual-Access Master $56,600 $357 $1,182 STI Bus, 208V, 60Hz TA79-BJ Dual-Access Master * 357 1,182 STI Bus, 240V, 50Hz TU79-AB Single Access Master 52,100 322 748 Massbus, 208V, 60Hz TU79-AD Single Access Master * 322 748 Massbus, 240V, 50Hz TU79-AF Add-On Slave Drive 28,000 196 624 208V, 60Hz TU79-AJ Add-On Slave Drive * 196 624 240V, 50Hz TA79-UG Upgrade Kit, Converts 15,000 --- 1,800 TU79 Master to TA79 TM79-C Dual Access Kit for 5,150 21 1,032 TU79-AB/AD TU79K-EA End-Panel/Mounting Hardware 450 --- Kit for TU79 on Tx78 string TU79K-EB End-Panel/Mounting Hardware 450 --- Kit for TU78 on Tx79 string * Consult AQS QUESTIONS AND ANSWERS HAVE THE TA/TU79 PRODUCTS REPLACED THE TA/TU78 PRODUCTS? Yes. ARE TEU79 AND TGU79 SUBSYSTEMS AVAILABLE TO REPLACE THE TEU78 AND TGU78? No. The TEU78 and TGU78 are being obsoleted. If you have a customer who needs one of these packaged configurations, please talk to your Digital Representative. CAN TA/TU79 AND TA/TU78 DRIVES BE INTERMIXED ON THE SAME TAPE SUBSYSTEM STRING? Yes. Special end-panel/mounting hardware kits are available to allow a TU78 to connect to a TA79 string or vice versa. CAN A TA/TU78 SUBSYSTEM BE UPGRADED INTO A TA/TU79? No. The TA/TU79 incorporates significant mechanical and cabinetry changes from the TA/TU78. However, the electronic changes incorporated in the TA/TU79 are field upgradeable into the TA/TU78 via an FCO kit. Contact your Field Service representative to check availability of the FCO kit. ================================================================================ Note 43.7 VAX SYSTEM ANNOUNCEMENTS 7 of 25 EISNER::HANRAHAN "Jamie Hanrahan" 2 lines 15-SEP-1987 16:22 -< what about upgrades? >- -------------------------------------------------------------------------------- Any info on 8200 to 8250 or 8350 upgrades? With pricing and order numbers? ================================================================================ Note 43.8 VAX SYSTEM ANNOUNCEMENTS 8 of 25 EISNER::KILLEEN "Jeff Killeen" 78 lines 10-OCT-1987 13:46 -< TEMPEST 8xxx SYSTEMS >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! ANNOUNCING THE TEMPEST VAX 8700 AND VAX 8800 - 2-October-1987 o Aimed at classified high-end processing o VAXcluster support o Preconfigured systems available PRODUCT DESCRIPTION The Tempest VAX 8700/8800 is a commercial VAX 8700/8800 placed inside standard Tempest cabinets (H9662 and H9661). Power and communication lines are modified versions of previously developed filters. The Tempest Ethernet link is the standard fiber optic interface to the Digital Tempest Ethernet; the CI communication links are shielded flexible conduit. The console is a commercial VAX 8700/8800 console unit housed in a Tempest cabinet with Tempest monitor and keyboard (RF-PC38N-AA). The communication link between the VAX 8700/8800 and the console is a shielded flexible conduit. CONFIGURATION INFORMATION There are four configurations offered. All configurations include one-year on-site hardware warranty, VMS and DECnet with either one-year or paid-up licenses, and the Tempest Console subsystem. o Tempest VAX 8800 VAXcluster System includes VAX 8800 CPU, 48 MB 256K MEM, FP, BBU, VAXBI channels (two), CI Port + BNC1A-20 cables (one set), and Ethernet communications interface o Tempest VAX 8700 Preconfigured System includes VAX 8700 CPU, 32 MB 256K MEM, FP, BBU, VAXBI channel (one), KDB50 disk controller, Ethernet communications interface, Tempest DMB32 communications controller, RA60 210 MByte removable disk drive, and Tempest TU81E tape drive o Tempest VAX 8700 VAXcluster System includes VAX 8700 CPU, 32MB 256K MEM, FP, BBU VAXBI channel (one), CI port + BNC1A-20 cables (one set), and Ethernet communications interface o Tempest VAX 8700 VMS System includes VAX 8700 CPU, 32 MB 256K MEM, FP, BBU, VAXBI channel (one), KDB50 disk controller, and Ethernet communications interface PRICING/ORDERING INFORMATION Model No. Description U.S. List Price BMC/DSMC RF-882CB-AP VAX 8800 VAXcluster Paid-Up $915,000 * RF-882CB-DP VAX 8800 VAXcluster One-Year 863,900 * RF-87EEB-EK VAX 8700 Preconfigured Paid-Up 687,800 * RF-87EEB-EL VAX 8700 Preconfigured One-Year 636,300 * RF-871CB-AP VAX 8700 VAXcluster Paid-Up 636,300 * RF-871CB-DP VAX 8700 VAXcluster One-Year 584,300 * RF-871BC-AE VAX 8700 VMS Paid-Up 645,400 * RF-871BC-DE VAX 8700 VMS One-Year 594,800 * *For the latest pricing information, consult the Corporate Price File. AVAILABILITY TEMPEST VAX 8700 and 8800 systems will only be offered in the United States and Canada. Listed configurations will be shipped within 90 days of acceptance of orders. Additional options may increase delivery time. ================================================================================ Note 43.9 VAX SYSTEM ANNOUNCEMENTS 9 of 25 EISNER::KILLEEN "Jeff Killeen" 100 lines 10-OCT-1987 14:59 -< PHASE DOWN OF 11/785 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! PHASE DOWN OF VAX-11/785 MODULE UPGRADE - 29-September-1987 o Last order date for VAX-11/785 Module Upgrade is July 1, 1988 o Limited availability: Orders filled on "First Come, First Served" basis Consistent with the strategy to phase down the VAX-11/780 and VAX-11/785 systems, the VAX-11/785 Module Upgrade (that converts the VAX-11/780 into an 11/785) will be phased down this fiscal year. This is the LAST CHANCE customers will have to order the VAX-11/785 Module Upgrade. Orders may be placed until July 1, 1988, or while supplies last. Your customers should be advised that there is a limited supply and orders will be filled on a "First Come, First Served" basis. Advise your customers to order as soon as possible. Because many orders are expected after this announcement, this is a great opportunity to increase short-term certs. PRODUCT MESSAGES More Performance The VAX-11/785 operates more than 50% faster than the VAX-11/780 processor, while using the same power and packaging. MASSBUS/UNIBUS Support The VAX-11/785 supports MASSBUS devices and multiple UNIBUS channels. Customers can use the same standard VAX interfaces and peripherals as on the VAX-11/780. Low Risk Solution Proven in more than 3,000 successful installations, the VAX-11/785 "in-cabinet" upgrade option is still a viable alternative for VAX-11/780 users. POSITIONING THE UPGRADE VAX-11/780 users have multiple growth options available to them. Position the VAX-11/785 Module Upgrade as just one of those options. For example: | Problem Solution | |-------------------------------------------------| | Needs short-term Upgrade in cabinet to | | performance VAX-11/785 with the | | enhancement intention of later | | upgrading to VAXBI | | or VAXcluster systems | |-------------------------------------------------| | Wants more system Create a VAXcluster | | availability or LAVc and distribute | | shared applications to | | multiple systems | |-------------------------------------------------| | Needs to consolidate Replace with VAX 8700, | | multiple CPUs to 8550 or 8530 | | conserve floor space | | or needs a long-term, | | higher performance | | system | | | Naturally, these growth options are dependent on the specific customer requirements, applications, growth rates, service level objectives and the user environment. To determine the appropriate solution and insure the customer achieves the expected performance gain, we recommend that a proactive performance monitoring system is established. ORDERING INFORMATION U.S. Model No. Description List Price* 780UP-XA VAX-11/785 Module Upgrade Kit $96,600 Includes VAX-11/785 CPU and FP785 board set. (VAX-11/780 and FP780 board sets must be returned to Digital.) 780UP-XB Same as 780UP-XA except without 90,300 floating point. *Installation included CUSTOMER NOTIFICATION This phase-down message will be communicated to customers in the Fall Edition of the DECdirect PLUS Catalog and in a special direct mail letter. Since orders cannot be handled by DECdirect, all orders will be processed by local Sales. ================================================================================ Note 43.10 VAX SYSTEM ANNOUNCEMENTS 10 of 25 EISNER::DEC_NEWS_1 "News about DEC from DEC" 380 lines 8-MAR-1988 20:55 -< DEC ANNOUNCES A REPLACEMENT FOR THE PDP-11/03 >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! DIGITAL EQUIPMENT CORPORATION INTRODUCES - 08-March-1988 ******************************************************************************* MOST POWERFUL VAX, SYMMETRIC MULTIPROCESSING MARLBORO, Mass., March 8, 1988 -- Digital Equipment Corporation today introduced its most powerful VAX computers -- the new VAX 8800 Series. The new series provides an easy, continuous growth path and delivers from 1 to 3.7 times the processing power of Digital's VAX 8700 computer. The new VAX 8800 Series pricing begins at $543,900 and ranges to over $1.7 million, depending on configuration. The VAX 8800 Series is available immediately. With the new VAX 8800 Series, Digital: o introduced VMS symmetric multiprocessing for the VAX 8800 Series. This enables users to increase throughput by exploiting the new series' multiprocessor capabilities; o announced a support and service program for the VAX 8800 Series that is the most comprehensive ever offered with any Digital system; o presented a straightforward growth path through the series, as well as the ability for users to upgrade existing VAX 8700 and VAX 8800 systems to the new series. The upgrade is done on-site and typically takes less than a day. New VAX 8800 Series Supports Large Computer Resource Centers "The versatility and expandability of the new VAX 8800 Series is ideal for large computer resource centers," said Peter J. Smith, Digital's Vice President of Product Marketing. "The new VAX 8800 Series is most effective when serving the varying needs of a wide range of users and applications, providing consistent high performance and overall throughput in environments such as engineering departments, giving technical professionals access to files and databases, technical computations, documentation and electronic mail. Other areas include shop floor information systems, tracking critical shop floor status to be integrated with planning and control, and banks, for analyzing and sharing data, electronic mail and data exchange with external services," Smith explained. Smith also noted that adding these systems to their existing VAXcluster system gives customers a more productive, cost effective alternative for new applications than an equivalent mainframe solution. The new VAX 8800 Series is a fully compatible extension to Digital's VAX family, and runs the VMS operating system, protecting users' software investments. The new VAX 8800 Series also can run over 3,500 application packages, available from Digital, from participants in Digital's Cooperative Marketing Program, and from other software suppliers. The ULTRIX operating system and VAX ELN software are supported on the VAX 8810 system today; both will be available for the other configurations in the future. VAX 8800 Series Includes Five Configurations The new VAX 8800 Series is available in five system configurations -- the VAX 8810, the VAX 8820, the VAX 8830, the VAX 8840 and the VAX 8842. The VAX 8800 Series incorporates from one to four processors, and uses VMS with symmetric multiprocessing to deliver balanced performance across all applications. According to William J. Demmer, Vice President of Mid-Range Systems, "The VAX 8800 Series extends high-end VAX performance up to 3.7 times the range of the very powerful VAX 8700 processor." The VAX 8842 combines Digital's VAXcluster system technology for high availability and access to a common database, with the versatility and flexibility of symmetric multiprocessing in VMS. "This system provides an ideal solution for customers with high availability requirements," Demmer noted. The VAX 8840 and VAX 8842 are also complemented with two optional solution packages. These packages, tailored for new VAXcluster system installations, provide extensive startup and support services to insure easy installation and implementation of large VAXcluster system sites (see fact sheet). Symmetric Multiprocessing Increases Power The VAX 8800 Series supports VMS symmetric multiprocessing, which increases the throughput and efficiency of multiprocessor VAX systems. The symmetric multiprocessing software does this by treating the processors as "peers." Digital's symmetric multiprocessing software also increases system throughput by dynamically balancing the workload across the VAX 8800 Series' multiple processors. The combination of simultaneous processing of multiple applications and dynamic load balancing enables the systems to run more jobs, support more users, and run larger workloads than any other VAX system. I/O Enhancements Demmer noted that the new series includes multiple Ethernet connections and a VAXcluster system port. "When computers reach this level of performance, a systems approach is necessary. In order to deliver high performance, a system needs more than CPU cycles -- it needs to balance performance with larger memories, more I/O capacity, and efficient software. We improved I/O throughput by supporting up to six VAXBI buses and two Ethernet adapters. This enhances the capabilities of both the system and the VAXcluster system," Demmer said. Upgrades Made Easy The new VAX 8800 Series provides a smooth and easy upgrade path as applications grow or the number of users increases. "Users upgrade a VAX 8810, VAX 8820, or VAX 8830 by adding one or more processors," said Demmer. "Upgrading is a straightforward process that is done on-site and typically takes less than a day," he added. Upgrades are priced at $280,000 per configuration. Most Comprehensive Service And Support The price of each multiprocessor VAX 8800 configuration includes the most comprehensive service and support program Digital has yet offered. The program includes installation, 24-hour, 7-day hardware support, and Performance Reporting Service for one year. This enables data processing departments to concentrate on integrating the system into their company's business operations. This one-year warranty covers all Digital hardware, software, and Digital-licensed software purchased with the system. Digital also assigns each customer an account manager at Digital's Customer Support Center, who helps promote high system efficiency by coordinating service and support and administering the Performance Reporting Service. Support service for the entry-level VAX 8810 configuration includes a full one-year DECservice hardware warranty that provides hardware installation, 24-hour hardware telephone support, and on-site hardware support during business hours. The warranty programs for the first year are included in the system price, and can be extended at the time of purchase to a total of up to three years. Digital Equipment Corporation, headquartered in Maynard, Massachusetts, is the world's leading manufacturer of networked computer systems and associated peripheral equipment, and is the leader in systems integration with its networks, communications, software, and service products. ### FACT SHEET Product Name: VAX 8800 Series Function: The VAX 8800 Series are the most powerful VAX systems ever offered, extending Digital's leadership in distributed computing beyond the mid-range to the highest performance range yet. These new systems are designed to meet the demanding needs of large resource centers running a combination of interactive applications such as electronic mail, data base inquiry, status checking and document editing, with multiple streams of computational tasks such as finite element analysis, economic modeling or process simulation. The new VAX 8800 Series is available in five system configurations. Two solution packages are available for VAX 8840 and VAX 8842 systems. One solution package provides a basic VAXcluster system configuration for new VAXcluster system sites. The other package provides high data availability. The new series provides an easy upgrade path through the series, including upgrades for existing VAX 8700 and VAX 8800 systems. Each VAX 8800 Series multiprocessor configuration includes a comprehensive one-year hardware and software warranty and a service and support program that features 24-hour, seven-day on-site hardware service, an account support manager for every customer, 24-hour telephone advisory support, and Digital's Performance Reporting Service. Key Features and Benefits: o Available in one-, two-, three-, and four-processor configurations, providing an easy, continuous growth path o Compatible with existing VAX application software and hardware, protecting existing user investments o Supports symmetric multiprocessing to fully exploit multiprocessor capabilities o Supports up to 512 Mbytes of main memory, allowing large programs to remain memory resident for fast execution o Supports up to six VAXBI buses for high I/O throughput o Customer support for multiprocessor systems includes one-year warranty on all Digital hardware and software, 24-hour, seven-day on-site hardware service, a dedicated account support manager, and Performance Reporting Service Sample Configurations: VAX 8810 o 60"h x 74"w x 30"d cabinet occupies 15.5 square feet (1.44 square meters) of floor space o One processor o 48 Mbytes of main memory o MicroPDP-11 console subsystem o VMS operating system software license o DECnet-VAX software end node license o 1 Ethernet adapter o 1 VAXBI bus o 1 VAXcluster system adapter or 1 KDB50 local disk controller o One year DECservice hardware warranty including installation, 24-hour hardware telephone support and on-site support during business hours o Performance: Equivalent to VAX 8700 o Prices begin at US$543,900 VAX 8820 o 60"h x 106"w x 30"d cabinet occupies 22.2 square feet (2.0 square meters) of floor space o Two processors o 128 Mbytes of main memory o MicroVAX II console subsystem with VT320 terminal, LA75 printer, 71 Mbyte RD53 fixed disk drive, and 95 Mbyte TK50 tape drive o VMS operating system software license o DECnet-VAX software end node license o VAXcluster system software license o 1 Ethernet adapter o 2 VAXBI buses o 1 VAXcluster system adapter o One-year hardware and software warranty, 24-hour, 7-day on-site hardware service, account support manager, Performance Reporting Service o Performance: up to 1.9 times the performance of VAX 8810 o Prices begin at US$833,700 VAX 8830 o 60"h x 106"w x 30"d cabinet occupies 22.2 square feet (2.0 square meters) of floor space o Three processors o 128 Mbytes of main memory o MicroVAX II console subsystem with VT320 terminal and LA75 printer, 71 Mbyte RD53 fixed disk drive, and 95 Mbyte TK50 tape drive o VMS operating system software license o DECnet-VAX software end node license o VAXcluster system software license o 2 Ethernet adapters o 2 VAXBI buses o 1 VAXcluster system adapter o One-year hardware and software warranty, 24-hour, 7-day on-site hardware service, account support manager, Performance Reporting Service, and 6-month resident software engineer o Performance: up to 2.8 times the performance of VAX 8100 o Prices begin at US$1,062,000 VAX 8840 o 60"h x 106"w x 30"d cabinet occupies 22.2 square feet (2.0 square meters) feet of floor space o Four processors o 128 Mbytes of main memory o 2.5 Gbytes of disk storage o Intelligent storage controller o MicroVAX II console subsystem with VT320 terminal and LA75 printer, 71 Mbyte RD53 fixed disk drive, and 95 Mbyte TK50 tape drive o VMS operating system software license o DECnet-VAX software end node license o VAXcluster system software license o 2 Ethernet adapters o 2 VAXBI buses o 1 VAXcluster system adapter o One-year hardware and software warranty, 24-hour, 7-day on-site hardware service, account support manager, Performance Reporting Service, and 6-month resident software engineer o Performance: up to 3.7 times the performance of VAX 8810 o Prices begin at US$1,472,000 VAX 8842 o Two 8820 systems o 2.5 Gbytes of disk storage o Intelligent storage controller o One-year hardware and software warranty, 24-hour, 7-day on-site hardware service, account support manager, Performance Reporting Service, and 6-month resident software engineer o Prices begin at: US$1,577,000 Solution Start-Up and Service Package: o Additional 2.5 Gbytes of disk storage o Dual density tape drive, Star Coupler o VAXcluster system console system and VAX Performance Advisor system management tools o Pre-site planning, DECstart Plus, software installation o DECplan educational credits o Additional 6-month software engineering resident o Prices begin at US$510,900 High Data Availability Package: o Additional intelligent storage controller, Dual-porting capability o VAX Volume Shadowing software o Price: US$125,100 ULTRIX, VAX, VAXBI, VAXcluster, VAX ELN, VAX/VMS, and VMS are trademarks of Digital Equipment Corporation. ================================================================================ Note 43.11 VAX SYSTEM ANNOUNCEMENTS 11 of 25 EISNER::SHANNON "Terry C. Shannon (R.I.P. Chas. M.)" 7 lines 9-MAR-1988 11:25 -< Field Test Two? >- -------------------------------------------------------------------------------- > -< DEC ANNOUNCES A REPLACEMENT FOR THE PDP-11/03 >- Conspicuously absent from the Polarstar announcement was a VMS version number. On the other hand, the announcement DID indicate that SMP Ultrix is waiting in the wings . . . ================================================================================ Note 43.12 VAX SYSTEM ANNOUNCEMENTS 12 of 25 EISNER::KILGALLEN "Larry Kilgallen" 1 line 9-MAR-1988 16:59 -< You DO remember MicroVMS V1.0, don't you? >- -------------------------------------------------------------------------------- -< Field Test Two? >- ================================================================================ Note 43.13 VAX SYSTEM ANNOUNCEMENTS 13 of 25 EISNER::CETRON 7 lines 10-MAR-1988 12:24 -< curiouser and curioser >- -------------------------------------------------------------------------------- Even more important is the fact the SMP Ultrix may actually start shipping (despite foregoing announcement) FIRST! Apparently, it is ready, and (Future unnamed major release) of VMS is not. ed ================================================================================ Note 43.14 VAX SYSTEM ANNOUNCEMENTS 14 of 25 EISNER::DEC_NEWS_1 "News about DEC from DEC" 322 lines 21-APR-1988 23:00 -< VAX 62xx >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. 20-Apr-88 04:01 PM Page: 1 ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! THE VAX 6200 SERIES - 19-April-1988 ******************************************************************************* VAX 6210, VAX 6220, VAX 6230, VAX 6240 o VAX 6200 series sets a new standard for mid-range price/performance and system growth capabilities o Systems span the range from 2.8 up to 11 times the performance of the VAX-11/780 o The VAX 6200 series is available in volume NOW! The new VAX 6200 series offers impressive system expansion capa bilities and sets a new standard for performance and cost of ownership of a mid-range VAX system. Each member of the VAX 6200 series -- the VAX 6210, VAX 6220, VAX 6230 and VAX 6240 -- features high-end system capabilities in a compact package. The broad range of performance achieved within a single cabinet allows customers to match performance to a specific application, and increase system capacity and performance as needs change and grow. The VAX 6240 offers the highest aggregate throughput of the series: up to 11 times the performance of the VAX-11/780. The VAX 6230 is a triprocessor system, the VAX 6220 is a dual-processor system, and the VAX 6210 is a uniprocessor system. Since all four models share the same system architecture and packaging, the VAX 6210, VAX 6220 and VAX 6230 can grow to match the user's needs. Performance Enhancement Packages, described elsewhere in this issue, provide the upgrade path. FEATURES/BENEFITS Feature Benefit CMOS Technology - 2.8 times the VAX-11/780 per processor - Small system footprint, high reliability, low cost of ownership Multiprocessor Architecture - High aggregate throughput, greater application performance - Seamless growth from 2.8 up to 11 times the VAX-11/780 performance Symmetric Multiprocessing - Dynamic load balancing Up to 256 MB Memory Capacity - Reduces disk I/O requirements in many applications Supports up to Six VAXBI - High aggregate I/O bandwidth Channels - Supports existing VAXBI options High throughput makes the VAX 6200 series systems ideal solutions for a wide variety of applications at both the department and enterprise level. It is particularly appropriate for multi-user or multistream and interactive applications in which multiple users perform different applications or multiple copies of the same application. These applications include computer-aided software development, timesharing, MIS, lab and office automation, computer-integrated manufacturing, circuit and mechanical simulation, and data analysis. Department or enterprise-wide solutions must provide an environment for well-balanced processing and I/O performance. These customers typically share computer resources among many users and need more data storage and back-up capabilities than found in microcomputer solutions. The key to success is compatibility with existing applications, along with the availability of a full range of software compatible hardware products. PRODUCT DESCRIPTION VAX 6200 series systems employ a multiple bus architecture. A high- bandwidth 64-bit internal system interconnect is used to manage the high- speed traffic between the CPU(s), main memory and multiple VAXBI I/O channels. Processors, memories and VAXBI adapters are options on the system interconnect, and each consists of a single module. The interconnect supports up to four processors, up to eight memory modules, and up to six VAXBI adapters. Central Processor The processor used in the VAX 6200 series is based upon the CVAX processor chipset. This allows the entire processor, including Cache memory, ROM and support logic, to be contained on a single low-power module. The VAX 6210 is a uniprocessor configuration, while the VAX 6220, VAX 6230 and VAX 6240 are dual-, tri- and quad-processor configurations, respectively. Each processor module in the configuration contains a floating point accelerator, and Cache memory (1 KB on the CVAX chipset and 256 KB on the module). In addition, each processor module includes a bootstrap/diagnostic ROM, an electrically erasable ROM (used by the console and diagnostics), a time-of-year clock, and a full set of console logic. The system console is typically an LA100 terminal, since the console logic is contained within the processor(s). In multiprocessor systems, such as VAX 6220, VAX 6230 or VAX 6240, one of the processors will control the console, while the others will have this logic disabled. Each system includes a TK50 and controller for software and diagnostic distribution. Two VAXBI channels are included with the processor. Four more VAXBI channels can be added in a separate cabinet. The UNIBUS must be configured on a separate VAXBI. UNIBUS supports DR11-W, DUP11 and LP11. Performance and Multistream Throughput VAX 6200 systems gain their performance from the CVAX chipset and the high-aggregate performance provided by the multiprocessor architecture and VMS Symmetric Multiprocessing (SMP). Performance of each processor in a VAX 6200 system is 2.8 times a VAX-11/780. Extensive benchmarking has shown that multistream applications achieve a total system throughput nearly equal to the total aggregate performance of the processors. System performance is detailed elsewhere in this issue. For general positioning purposes, the following guidelines apply: Performance x VAX-11/780 Number of Model Processors Single-Stream Multistream VAX 6210 1 2.8 2.8 VAX 6220 2 2.8 Up to 5.5 VAX 6230 3 2.8 Up to 8.3 VAX 6240 4 2.8 Up to 11 Performance: Parallel Processing In general, no modification of application software is required for use on VAX 6200 systems. However, some applications, such as a single-user simulation environment, might be sensitive to single-stream performance. In these cases, users can elect to "decompose" their applications using tools contained in the VMS V5.0 Run-Time Library; or, for third-party applications, they may obtain a "decomposed" version from the application supplier. When decomposed, a single copy of an application can be made to run on all processors simultaneously at a greater speed than the single-stream performance. Performance Enhancement Packages VAX 6200 systems are designed for expansion in all dimensions: mass storage, VAXBI options, main memory and processor performance, allowing the system to grow as the requirements of the customer dictate. A key part of this design is the modularity of processors and memory components which allows VAX 6210, VAX 6220 and VAX 6230 systems to be field upgraded to higher system performance by the addition of the appropriate modules. VAX 6200 series Performance Enhancement Packages can be ordered now and will be available in Q1/FY89. System Memory Memory in the VAX 6200 systems can be configured up to 256 MB in 32 MB arrays. This large capacity allows large programs and data structures to fit within memory, reducing disk I/O requirements and increasing application performance. Each memory module is protected by an 8-bit ECC code with its own controller. The memory controllers feature command queuing and automatic interleaving (up to 8-way) for high performance. The VAX 6200 systems support battery back-up. The battery back-up option will be available in Q1/FY89 and can be ordered as H7231-N. VAXBI I/O Subsystems VAX 6200 systems employ the VAXBI for system I/O, and the connection to the system interconnect is made via a high-bandwidth adapter. The VAXBI offers high bandwidth and compatibility with a rich set of options. Each VAXBI adapter has a throughput of approximately 10 MB/sec, so many VAX 6200 systems will require no more than two VAXBI adapters. However, VAX 6200 systems can be configured with up to six VAXBI adapters for the most demanding applications. Each VAXBI adapter consists of a board set that connects the system interconnect with a 5-slot VAXBI channel. Within the cabinet, there is dedicated space for two VAXBI adapters and two 5-slot VAXBI channels. If more than two VAXBI channels or 10 VAXBI slots are required for an application, the VAXBI capability can be expanded by adding channels in a VAX BI Expander Cabinet. This 60-inch cabinet has up to four additional 5-slot backplanes, for a system total of six channels and a total of 30 VAXBI slots. The VAXBI Expander Cabinet will be available in Q1/FY89 and can be ordered as DWMBA-BA/BB (208/416V). It contains one 5-slot VAXBI backplane with a VAX 6200 interface module set. Up to three additional 5-slot VAXBI backplanes (each with its own interface modules) can be ordered as DWMBA-CA. VAXBI Options The following VAXBI options can be configured in the VAX 6200 systems subject to rules published in the VAX 6200 Systems and Options Catalog Supplement. Supported Supported Option Per VAXBI Per System KDB50 2 8 DEBNA 2 4 CIBCA 1 1 TU81-Plus 2 4 DMB32/DHB32* 4 8 DRB32-M 2 8 DRB32-E 2 6 DRB32-W 2 6 *A maximum of two DMB32s, or one DMB32 and one DHB32, is supported in the CPU cabinet. UNIBUS Support VAX 6200 systems have limited support for UNIBUS devices, based on the following guidelines: o UNIBUS is supported only on the VAX 6210 and VAX 6220 systems, due to bus latency considerations. When a UNIBUS is no longer required by the customer, these systems can be upgraded to VAX 6230 and VAX 6240. o Only one UNIBUS adapter is available per system, and that adapter must be configured on a separate external VAXBI channel. o Systems with a UNIBUS adapter cannot have additional external VAXBI channels. Thus, these systems are limited to two internal VAXBI channels and the dedicated UNIBUS channel. o Only the DR11-W, DUP11 and LP11 devices are supported. These restrictions place limits on a customer's expansion capabilities. In order to preserve system upgrade opportunities, VAXBI options should be used to address application needs wherever possible. Customers who currently employ the UNIBUS should have an account plan to migrate to the VAXBI. Customers who require UNIBUS support should order the VAX 6200 UNIBUS cabinet: DWMUA-AA/AB (208/416 V). This 60-inch cabinet contains a dedicated VAXBI channel with interface to the VAX 6200 and a BA11-AW/AX UNIBUS enclosure. DD11-DK backplanes should be ordered as needed. This option will be available in Q1/FY89 and can be ordered now. Reliability and Maintainability The VAX 6200 systems are designed for high reliability. CMOS technology is inherently cool-running and reliable, and the high circuit density of the CVAX chip and CMOS gate arrays means fewer parts can fail. The device's low power consumption allows for smaller power supplies which dissipate less heat. Less power consumption also creates lower cooling costs, and simpler cooling components. The VAX 6200 series incorporates a low-velocity cooling system that is exceptionally quiet. Each module and/or subsystem in the processor runs a power-up self-test to ensure data integrity. Errors are displayed on the console terminal and via LED indicators. At power-up or reboot, if a failing component is one of the multiple processors or memory modules (VAX 6220, VAX 6230 and VAX 6240 systems) the failing module will be dropped from the configuration and the system will run with available processors and memory. Work can continue and maintenance can be scheduled for a convenient time. Extensive parity is employed in the Cache and data paths. In many cases, transient errors will be recovered by automatic retry mechanisms in the hardware data paths, including the system interconnect, the processor data path and the VAXBI. System Packaging All VAX 6200 systems use the same 60-inch cabinet which contains the system interconnect, two 5-slot VAXBI channels and a TK50 tape drive. This package, with a full complement of processors and memory (256 MB), Ethernet, and other VAXBI options, occupies only 6.4 square feet. Since all the VAX 6200 systems use the same enclosure, uniprocessor, dual processor and triprocessor systems can grow with the user's requirements by adding the appro priate processor and memory modules. SOFTWARE SUPPORT VAX 6200 systems are supported by VMS Version 5.0 and layered software products. ULTRIX-32 support is planned for a future release. SERVICES All VAX 6200 systems include a one-year warranty on all system hardware and the operating system. The hardware warranty features DECservice with a committed two-hour response. The software warranty features telephone support, right-to-copy updates, and access to the Digital Software Information Network (DSIN). Services are described elsewhere. GIA/EUROPEAN SPECIFIC INFORMATION VAX 6200 systems are available worldwide, subject to Department of Commerce export regulations as follows: o VAX 6210 and VAX 6220 systems can be shipped to Class "A" and Class "B" countries under the terms of Digital's Distribution License (DL), but will require an Individually Validated License (IVL) to other countries. o VAX 6230 and VAX 6240 systems can be shipped to Class "A" countries using Digital's DL, but will require an IVL for all other countries. AVAILABILITY VAX 6200 systems are available now and will begin shipping in volume immediately. You should book orders immediately, specifying Q4 delivery. Certain VAX 6200 options can be ordered now, but will not be available until Q1/FY89. The options listed below should be specified as partial shipments during Q4, to ensure Q4 delivery of the majority of the system order. o H7231-N Battery Back-Up Option o DWMBA-BA/BB/CA VAXBI Expansion Options o DWMUA-AA/AB UNIBUS Expansion Options o VAX 6200 Series Performance Enhancement Packages ================================================================================ Note 43.15 VAX SYSTEM ANNOUNCEMENTS 15 of 25 EISNER::OSUDAR "John Osudar" 14 lines 22-APR-1988 19:57 -< Remember the UNIBUS >- -------------------------------------------------------------------------------- Limited support for UNIBUS devices?! You ain't kiddin'... We were seriously considering changing our order from an 8350 cluster to a single 62xx series system, until we saw that one. I've got an 11/785 with two Unibuses full of hardware. I can get that down to one Unibus full of stuff on an 8350 -- but since we have no DR11-W's, DUP11's, or LP11's, I couldn't do diddly with a Unibus on a 6200. And I'd really love to pay about $30K or so for the extra BI (when I don't need both of the first two) plus the expander box, which of course won't arrive until N months after the rest of the system... Was this the brilliant idea of some idiot in marketing, or is there a serious technical reason for this ridiculous situation? but still smouldering... ================================================================================ Note 43.16 VAX SYSTEM ANNOUNCEMENTS 16 of 25 EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 11 lines 23-APR-1988 10:29 -< I am sure it is Market driven... >- -------------------------------------------------------------------------------- Yes I remember the UNIBUS, and I am sure it is a marketing decision, for places like us that have three DMZ32s, used for nothing but driving LP26s and an LP27. When (if) we ever get rid of the 785s we would just get the BI controller for the printers. Do I really want to pay the extra to support something I will never use? Yes, maybe the cost of the new controller would be the same as the cost of "proper" UNIBUS support... But, yes, it does leave you out in a lurch... Seton ================================================================================ Note 43.17 VAX SYSTEM ANNOUNCEMENTS 17 of 25 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 4 lines 25-APR-1988 09:58 -< Takin' the last (uni)bus out of town. >- -------------------------------------------------------------------------------- We are looking to trade our 782 in on a better machine, since V5.0 doesn't support asymmetric multiprocessing. DEC gave us an offer, but said we could keep all the Unibus stuff. They just weren't interested. ================================================================================ Note 43.18 VAX SYSTEM ANNOUNCEMENTS 18 of 25 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 1 line 28-APR-1988 15:59 -< And how about all our 3rd party Ubus devices? >- -------------------------------------------------------------------------------- ================================================================================ Note 43.19 VAX SYSTEM ANNOUNCEMENTS 19 of 25 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 13 lines 29-APR-1988 08:41 -< No Unibus *stinks* to me >- -------------------------------------------------------------------------------- > >< Note 43.14 by EISNER::DEC_NEWS_1 "News about DEC from DEC" > > 6200 Series > Only the DR11-W, DUP11 and LP11 devices are supported. Amongst other things, this means there is no programmable communications board allowed (e.g. KCT-32). Is it safe to assume that it would not be unwise for DEC to be working on a BI programmable comm board? Or from now on do we have to buy a computer to do what we used to do with a board??? :-( I'd say more about how the closed BI system is starting to *really* rear it's ugly head, but that would belong in SOAPBOX. ================================================================================ Note 43.20 VAX SYSTEM ANNOUNCEMENTS 20 of 25 EISNER::TINKELMAN "Bob Tinkelman - CCA" 7 lines 29-APR-1988 12:33 -< Simpact BI programmable comm board >- -------------------------------------------------------------------------------- > Amongst other things, this means there is no programmable > communications board allowed (e.g. KCT-32). Is it safe to > assume that it would not be unwise for DEC to be working > on a BI programmable comm board? Doesn't Simpact have one? With a DEC approved BI-license agreement? Joint marketing, too? ================================================================================ Note 43.21 VAX SYSTEM ANNOUNCEMENTS 21 of 25 EISNER::HANRAHAN "Jamie Hanrahan" 10 lines 29-APR-1988 14:03 -------------------------------------------------------------------------------- > > Amongst other things, this means there is no programmable > > communications board allowed (e.g. KCT-32). Is it safe to > > assume that it would not be unwise for DEC to be working > > on a BI programmable comm board? > > Doesn't Simpact have one? With a DEC approved BI-license agreement? > Joint marketing, too? Yes, we do. ================================================================================ Note 43.22 VAX SYSTEM ANNOUNCEMENTS 22 of 25 EISNER::TINKELMAN "Bob Tinkelman - CCA" 139 lines 2-AUG-1988 22:44 -< 05/17/88 VAX 62xx Startup Service Packages >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! VAX 62xx STARTUP SERVICE PACKAGES - 17-May-1988 ******************************************************************************* o The VAX 6210/6220/6230/6240 SSPs provide comprehensive, cross-functional services that complement the one-year integrated HPS/SPS software warranty that comes with the system o DECnet-VAX DECstart in the VAXcluster SSPs contributes to network effectiveness Startup Service Packages (SSPs) for the mid-range VAX 6210/6220/6230/6240 systems provide Digital's customers with cross-functional, comprehensive services which enhance the one-year integrated hardware/software warranty and ensure successful system installation, startup, and usage. These SSPs, which are eligible for DBA discount, are available only in Levels II and III. Software Product Service One-year SPS Media and Documentation Update Service (SMS) complements the telephone support, DSIN, and Rights-to-Use software warranty attributes. Update Installation/Impact Planning Analysis is the DECsupport Uplift service component in the VAX 6210/6220/6230/6240 Level III SSPs. Software Services DECnet-VAX DECstart Services are a predefined set of tasks delivered at the customer's site and structured to ensure an efficient startup of the Digital DECnet-VAX network. DECstart assists the customer to configure the new DECnet-VAX network. This key service component is included in both the new and upgrade 6210/6220/6230/6240 VAXcluster SSPs. The Software Services DECstart, Educational Services training, and SPG Initial Media and Documentation traditional SSP components remain in the VAX 6210/6220/6230/6240 Startup Service Packages as key service elements. The service components included in each VAX 6210/6220/6230/6240 SSP are: Level III (QT-001AC-B5) Level II (QT-001AC-75) o Initial software installation o Initial software installation (VMS and all qualified products) (VMS and all qualified products) o VMS DECstart Plus (QSC49-SZ) o VMS DECstart (QSC48-SZ) o Initial Media and Documentation o Initial Media and Documentation (VMS and all qualified products) (VMS and all qualified products) o Educational training ($7,080) o Educational training ($3,540) Level III (QT-001AC-B5) Level II (QT-001AC-75) o 1-year SPS Media and Documentation o 1-year SPS Media and Documentation Update Service (SMS) Update Service (SMS) (VMS and all qualified products) (VMS and all qualified products) o 1-year SPS Update Installation/ Impact Analysis Planning VAX 6210/6220/6230/6240 ALL-IN-1 SSPs Level III (QT-AAAAC-B5) LEVEL II (QT-AAAAC-75) o Initial software installation o Initial software installation (VMS and all qualified products) (VMS and all qualified products) o ALL-IN-1 software installation o All-IN-1 software installation o ALL-IN-1 DECstart Plus (QSC32-SZ) o ALL-IN-1 DECstart (QSC31-SZ) o VMS DECstart (QSC48-SZ) o VMS DECstart (QSC48-SZ) o Initial Media and Documentation o Initial Media and Documentation (VMS, ALL-IN-1 and all qualified (VMS, ALL-IN-1 and all qualified products) products) o Educational training ($17,870) o Educational training ($11,420) o 1-year SPS Media and Documentation o 1-year SPS Media and Documentation Update Service (SMS)(VMS, ALL-IN-1, Update Service (SMS)(VMS, and all qualified products) ALL-IN-1, and all qualified o 1-year SPS Update Installation/ products) Impact Analysis Planning VAX 6210/6220/6230/6240 NEW VAXcluster SSPs Level III (QT-025AC-B5) Level II (QT-025AC-75) o Initial software installation o Initial software installation (VMS and all qualified products) (VMS and all qualified products) o VAXcluster software installation o VAXcluster software installation o VMS VAXcluster DECstart Plus o VMS VAXcluster DECstart (QSC76-SZ) (QSC75-SZ) o DECnet-VAX DECstart (QSN2A-SZ) o DECnet-VAX DECstart (QSN2A-SZ) o Initial Media and Documentation o Initial Media and Documentation (VMS and all qualified products) (VMS and all qualified products) o Educational training ($9,950) o Educational training ($5,790) o 1-year SPS Media and Documentation o 1-year SPS Media and Documentation Update Service (SMS) Update Service (SMS) (VMS and all qualified products) (VMS and all qualified products) o 1-year SPS Update Installation/ Impact Analysis Planning VAX 6210/6220/6230/6240 Upgrade VAXcluster SSPs Level III (QT-026AC-B5) Level II (QT-026AC-75) o VAXcluster software installation o VAXcluster software installation o VAX-to-VAXcluster DECstart o VAX-to-VAXcluster DECstart (QSC55-SZ) (QSC55-SZ) o DECnet-VAX DECstart (QSN2A-SZ) o DECnet-VAX DECstart (QSN2A-SZ) o Initial Media and Documentation o Initial Media and Documentation (for ONE NEW LICENSED PRODUCT) (for ONE NEW LICENSED PRODUCT) Level III (QT-026AC-B5) Level II (QT-026AC-75) o Educational training ($3,390) o Educational training ($2,510) o 1-year SPS Media and Documentation o 1-year SPS Media and Documentation Update Service (SMS) Update Service (SMS) o 1-year SPS Update Installation/ Impact Analysis Planning **NOTE: NODE SSPs (-7Z/-BZ) are not required, due to the integrated warranty attributes of telephone support and DSIN. ***NOTE: The ULTRIX-32 operating system is not yet available for the VAX 62xx series. PRICING/ORDERING INFORMATION U.S. Order No. Description List Price QT-001AC-75 VAX 62xx VMS SSP LEVEL II $29,030 QT-001AC-B5 VAX 62xx VMS SSP LEVEL III 46,400 QT-AAAAC-75 VAX 62xx ALL-IN-1 SSP LEVEL II 50,960 QT-AAAAC-B5 VAX 62xx ALL-IN-1 SSP LEVEL III 70,910 QT-025AC-75 VAX 62xx NEW VAXcluster SSP LEVEL II 45,310 QT-025AC-B5 VAX 62xx NEW VAXcluster SSP LEVEL III 60,435 QT-026AC-75 VAX 62xx UPGRADE VAXcluster SSP LEVEL II 29,750 QT-026AC-B5 VAX 62xx UPGRADE VAXcluster SSP LEVEL III 38,050 NOTE: 16MT9 media is not available for the VAX 62xx series at this time. ================================================================================ Note 43.23 VAX SYSTEM ANNOUNCEMENTS 23 of 25 EISNER::TINKELMAN "Bob Tinkelman - CCA" 81 lines 2-AUG-1988 22:45 -< 05/17/88 Phase-down of VAX-11/785 upgrade >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! PHASE-DOWN OF VAX-11/785 UPGRADE - 17-May-1988 ******************************************************************************* o Last order date for VAX-11/785 Module Upgrade is July 1, 1988 o All orders must be placed with local sales offices This is the LAST CHANCE customers will have to order the VAX-11/785 Module Upgrade. Final orders must be placed by July 1, 1988. Orders will be filled on a first-come, first-served basis and are subject to availability. To ensure that customers are satified with their level of service, sales representatives are highly encouraged to follow these steps with their current VAX-11/780 accounts. 1. Identify all existing VAX-11/780 installed base accounts (end-users and volume OEMs who service customers with VAX-11/780 systems). 2. Analyze customers' growth requirements for that installation. 3. Position the VAX-11/785 Module Upgrade as one alternative for short-term future growth and as an "interim" step for customers who need more processing power but are not able to migrate to VAXBI yet. 5. Refer to the Product Shipping Guide for product delivery information. Contact Manufacturing Operations if you have any questions regarding availability. POSITIONING THE UPGRADE VAX-11/780 users have multiple growth options available to them. Position the VAX-11/785 Module Upgrade as just one of those options. For example: Problem Solution Needs short-term performance Upgrade in cabinet to VAX-11/785 enhancement. with the intention of later upgrading to VAXBI or VAXclusters. Wants more system availability. Create a VAXcluster or LAVc and distribute shared applications to multiple systems. Problem Solution Needs to consolidate multiple Replace with VAX 8800 Series, or CPUs to conserve floor space VAX 6200 Series, or VAX 8550. or needs a long-term higher performance system. Naturally, these growth options are dependent on the specific customer requirements, applications, growth rates, service level objectives and user environment. We recommend that customers implement performance management techniques in order to achieve maximum productivity from existing systems, while planning future computing resources to meet their business growth expansion. ORDERING INFORMATION Since orders cannot be handled by DECdirect, all orders must be placed with your local sales offices. U.S. Model No. Description List Price 780UP-XA VAX-11/785 Module Upgrade Kit $ 96,600 Includes VAX-11/785 CPU and FP785 board set. (11/780 and FP780 board sets must be returned to Digital.) 780UP-XB Same as 780UP-XA but without 90,300 floating point. *Installation included. ================================================================================ Note 43.24 VAX SYSTEM ANNOUNCEMENTS 24 of 25 EISNER::TINKELMAN "Bob Tinkelman - CCA" 155 lines 2-AUG-1988 22:47 -< 06/29/88 VAX 8842/8840 start-up packages >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! THE VAX 8842/8840 - 29-June-1988 ******************************************************************************* WITH SOLUTION START-UP AND SUPPORT PACKAGE o Three areas pertaining to the VAX 8842/8840 Solution Start-up and Support Package and the High Data Availability - Ordering Rules - Package Contents - Process/Resources to be used when selling a VAX 8842 or a VAX 8840 WITH a Solution Start-up and Support Package ORDERING RULES There are three components to the new VAX 8842/8840 Systems. 1. The System Building Blocks - SBB (VAX 8842, Part No. 889CD-AP/DP and VAX 8840, Part No. 888CC-AW/DW). 2. The Solution Start-up and Support Package, Part No. 889CC-AP/DP 3 The High Data Availability Package, Part No. 889CH-AP/DP The Solution Start-up and Support Package, described below, may only be ordered WITH (AND ON THE **SAME ORDER** AS) a VAX 8842 SBB (889CD-AP/DP) or a VAX 8840 SBB (888CC-AW/DW). The High Data Availability Package may ONLY be ordered for use with the VAX 8842 and VAX 8840 SBBs. Although it is strongly encouraged that this package be ordered at the same time an SBB is ordered, it can be ordered at a later date. Customers looking for the total system solution (hardware, software, installation and services) similar to that offered with the VAX 897x Systems must order the 8842/8840 SBB, the Solution Start-up and Support Package and the High Data Availability Package. PACKAGE CONTENTS VAX 8842/8840 Solution Start-up and Support Package - Part No. 889CC-AP/DP The SOLUTION START-UP & SUPPORT PACKAGE for the VAX 8842/8840 offers much more than just a Start-up Service Package. The appropriate VAXcluster SSP Level III is a key component of the Solution Start-up and Support Package. The Solution Start-up and Support Package is comprised of the following: o 2.5 gigabyte SA482 Storage Array o One-year Integrated HPS/SPS Warranty o 8-line, SC008 Star Coupler (DECservice/DECsupport Level) o 1 HSC70 Tape Channel Card o Pre-site Solution-level Site Plan o DECserver 200 Terminal Server o Ed Services DECplan Account ($21,750 o TA79 Dual-Density Tape System included) o 4 HSC70 Disk Channel Cards o SWS Sys. Eng. Res. (6 mo. QS846-SZ) o VAXcluster Console System w/color o SWS UPGRADE to VAX Performance graphics terminal and 4 FOCHA-AE o Optomization (VPO) (QSP78-SZ) o VAXcluster Systems Wallchart o SWS UPGRADE to VPO o VAX Performance Advisor license Inc.NODE (QSP79-SZ) o Solution-level, Customized o SWS VPO QTR Extension (QSP76-SZ) Configuration Control Document o SWS VPO QTR EXT. Inc. NODE (QSP77-SZ) o Local Network Interconnect (DELNI)o NEW VAXcluster SSP LEVEL III * o 4 BN25J-20 and 4 BNE3L-20 cables o NEW VAXcluster SSP NODE LEVEL III ** * QT-025AD-BM SSP LEVEL III Components: o SW Update Installation/Impact Planning Analysis o VMS VAXcluster DECstart Plus (QSC76-SZ) o DECnet-VAX DECstart (QSN2A-SZ) o VPCS/Performance Analysis (QSP72-SZ) o Initial Media and Documentation (H-Kits) (for VMS and all qualified products) o Educational training ($9,950 included) o One year SPS Media and Documentation Update Service (for VMS and all qualified products) **QT-025AD-BZ SSP NODE LEVEL III Components: o Incremental Update Installation/Impact Planning Analysis per Node o VPCS/Performance Analysis Incremental Node (QSP73-SZ) o Documentation (G-Kits) for VMS and all qualified products o One-year SPS Additional Update Service (-KZ) (for VMS and all qualified products) NOTE: On SBBs ordered without Solution Start-up and Support Package, software installation is provided but not media (except for VMS and systems integrated products) and documentation. VAX 8842/8840 High Data Availability Package (Part No. 889CH-AP/DP) o 1 HSC70 Storage Controller o 4 HSC70 Disk Channel Cards o 1 HSC70 Tape Channel Card o 1 BNCIA-20 CI Cable o 8 BC26V-25 Storage Cables o VAX Volume Shadowing License - VAX Volume Shadowing installed by Digital if H.D.A. package is purchased at the same time as VAX 8842/8840 SBB. It is customer-installed otherwise. - Documentation included ONLY if the H.D.A. package is purchased on the same order as Solution Start-up and Support Package. Otherwise order QA-AB2AA-GZ. (Documentation is built into SSP included in the Solution Start-up and Support Package. Media comes with VMS in the System Building Blocks). o Present an integrated system solution that includes all the right parts including planning, documentation and installation of the entire solution. o Encourage customers to purchase the complete package which includes services and support that have proven to be effective for the successful implementation of a system solution. The VAXcluster Series Program Team ensures that all manufacturing, software services, field service, etc. issues are resolved transparent to your customer. - Customer Preliminary Site Survey--Analyze the customer's computer room environment, access aisles and receiving facilities. (A detailed environmental analysis will occur after the order is certed.) - Customer Configuration Document (CCD)--Details the customer's equipment list, cabinet and box information, cable layouts, systems environmental information, pregenerated software information, packing and unpacking information, along with detailed plots that configure the entire system within the customer's computer room and is performed by the Customer Project Support Group (CPSG). - Customer Support Plan (CSP)--Documented Digital and customer commitments is required for a successful installation and ongoing service delivery by the Sales Account Team. - An analysis of the proposed solution (hardware, software, services and any third party involvement) by the Sales Account Team. - A SWS analysis to understand how the customer will best use the Resident Systems Engineer and the identification of the engineer. These activities must be started before an order can be certed. As with the the VAX 897x systems, partial ships or splitting of orders is not allowed. The VAXcluster Series Program Team is committed to 90-day availability after an order is certed, provided that the preliminary Customer Configuration Document is completed and the customer site is ready for installation by the time the system is delivered. RESOURCES FOR ADDITIONAL INFORMATION If you are considering a VAX 8842, a VAX 8842/8840 Solution Start-up and Support Package and the High Data Availability Package to meet your customers' needs, please have your Digital Representative contact any of the following people in the VAXcluster Series Program Team: ================================================================================ Note 43.25 VAX SYSTEM ANNOUNCEMENTS 25 of 25 EISNER::TINKELMAN "Bob Tinkelman - CCA" 50 lines 2-AUG-1988 22:51 -< 06/13/88 Price increases due to higher DRAM costs >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! NEW PRICE INCREASES DUE TO HIGHER DRAM COSTS - 13-July-1988 ******************************************************************************* o New prices for all VAX systems and for most VAX and PDP-11 add-on memory options o All new prices were effective June 13, 1988 The price increases for systems with embedded memory average approximately 3.5 percent; add-on memory product increases average approximately 35 percent. These changes reflect the higher than expected costs for DRAM chips. Although most of the industry is experiencing a shortage of DRAM chips, Digital continues to meet demand for computer systems from our customers worldwide. The tight supply and increased price of DRAM memory chips is an industry- wide phenomenon, and all memory and system vendors buying DRAMS from the merchant market are equally effected. PRICING/ORDERING INFORMATION S E O U.S. E / E Instal. Model No. Description List Price G U M BMC Charge MS650-AA 8 MB ECC MEM MicroVAX 3XXX BA23 $ 6,000 XX Y 2 N/C $350 MS650-AF 8 MB ECC MEM MicroVAX 3XXX BA21X 6,000 XX Y 2 N/C 350 MS630-BB 4 MB EXPAN. MEMORY 1,968 XX T T N/C 325 MS630-BF 4 MB EX MEM FOR BA21X 1,975 XX T T N/C 325 MS630-CA 8 MB PARITY MEM 3,583 XX T T N/C 325 MS630-CF 8 MB MicroVAX II FOR BA21X 3,583 XX T T N/C 325 MS730-CA 1 MB VAX-11/730 MEM 1,335 XX T T N/C 455 MS750-CA 1 MB VAX-11/750 MEM 1,335 XX T T N/C 455 MS780-FA 2 MB VAX-11/780 MEM 2,483 XX T T N/C 500 MS780-JA 8 MB VAX-11/780 4,561 XX T T N/C 512 MSV11-QA 1 MB Q-BUS PAR MEM 1,240 XX T T $29 335 MSV11-QB 2 MB Q-BUS PAR MEM 1,430 XX T T 36 335 MSV11-QC 4 MB Q-BUS PAR MEM 2,000 XX T T 72 350 MSV11-QF 1 MB Q-BUS 11/53 BA21X 1,240 XX T T 29 335 MSV11-MB 1 MB DUAL MOS PARITY 658 XX T T 16 318 MS11-PB 1 MB ECC MOS MEMORY 1,865 XX T T 48 320 MS88-DA 64 MB 88XX/87XX/86XX MEM 37,000 XX J Q N/C 578 MS86-DA 64 MB 8600/8600 MEM 37,000 XX V V N/C 575 MS86-CA 16 MB 8600/8650 MEM 12,000 XX V V N/C 575 MS820-CA 16 MB 83XX/83XX BI MEM 12,000 XX S Q N/C 465  ================================================================================ Note 44.0 RDxx FCT Format 7 replies EISNER::ETHINGTON "Superior Klingon Technology" 33 lines 9-SEP-1987 01:23 -------------------------------------------------------------------------------- I need some information on the format of the FCT (whatever that stands for) which is recorded on the last cylinder of all DEC RDxx winchester drives, and records the manufacturer's detected bad block information. I know the following: 1. The FCT is recorded on each surface of the disk, with the FCT block repeated 16 times on the surface to ensure that at least one copy will be readable. This information is on the last cylinder of the disk. 2. The first 256 bytes of an FCT block consist of 64 4 byte entries, and is terminated by an all zero entry. 3. Each entry encodes the following bad block information, describing a media defect for the surface you are reading it from: 15 5 4 0 +--------------------+----------+ ! Cylinder Number ! Head ! +--------------------+----------+ | Offset from index | +-------------------------------+ OK, all well and good. What I want to know at this point is: What information is recorded in the final 256 bytes of each FCT block? On the drives I have examined, the first two bytes of the remaining 256 are non-null; they do not appear to be the drive serial number, at least in any radix I can think of; they are not the same on different drives. Help?!?!?!?! Jerry Ethington ================================================================================ Note 44.1 RDxx FCT Format 1 of 7 EISNER::KILLEEN "Jeff Killeen" 5 lines 9-SEP-1987 08:24 -< A THOUGHT >- -------------------------------------------------------------------------------- > -< RDxx FCT Format >- Have you tried re-running the formater with a new serial number to see what happens? In PDP-11 land it is R ZRQC?? ================================================================================ Note 44.2 RDxx FCT Format 2 of 7 EISNER::ETHINGTON "Superior Klingon Technology" 15 lines 12-SEP-1987 15:27 -< The Point of All This... >- -------------------------------------------------------------------------------- I guess I should have supplied a spot more information. The intent of all this is to write an application, running on Pro series hardware, to permit one to type in the serial number and bad block information from the label on top of a drive and rewrite the FCT from scratch. I'm not familiar with the current diagnostics. Older ones didn't even use to ask for a serial number, if I remember correctly. I also don't think I've ever seen one that would permit you to reeenter the bad block info. I have access to a bunch of official DEC drives that have had their FCTs destroyed by a rogue version of RT-11; also, I would obviously like to use cheap third party drives as opposed to paying DEC triple the price for the same drive. I already have a program running on the Pro that formats the last cylinder and records the bad block info, and it works fine and dandy; I'm just trying to figure out what the last part of the FCT is, and whether it is important. ================================================================================ Note 44.3 RDxx FCT Format 3 of 7 EISNER::KENNEDY "Terry Kennedy" 29 lines 12-SEP-1987 21:14 -< Here's the inside scoop... >- -------------------------------------------------------------------------------- 1) FCT = Format Control Table 2) The two bytes you see *should* be the controller model and software revision of the controller that initted the drive. 3) You can (with RQDX1, 2, or 3) initialize a drive even if the FCT is shot by answering 'Y' to 'continue if FCT is inaccessible' and 'N' to 'use existing bad block info', etc. 4) The following drives are the ones that DEC uses - you can use any of these as long as it is supported on your controller type. All DEC does is format them. Use of other drives is possible by burning new EPROMS for the controller, but the parameter determination is a bit sticky. DEC DRIVE MODEL REAL MANUFACTURER/MODEL MINIMUM CONTROLLER REQ'D --------------- ----------------------- ------------------------ RD51 SEAGATE ST-412 RQDX1 RD52 (primary) QUANTUM Q-540 RQDX1 W/ ROM UPGRADE 1 RD52 (alternate) ATASI 3046 RQDX1 W/ ROM UPGRADE 1 RD53 MICROPOLIS 1325 RQDX2 W/ ROM UPGRADE 2 RD54 MAXTOR (model?) RQDX3 RD3? (unknown) SEAGATE ST-225 RQDX3 Rom upgrade 1 is DEC kit # EQ-01361-01, containing parts 23-172E5-00 and 23-173E5-00. Rom upgrade 2 (no kit #), contains parts 23-188E5-00 and 23-189E5-00. For completeness, the current RQDX3 ROMS are 23-243E5-00 and 23-244E5-00. 5) All this and more can be found by ordering DEC model QP-905-GZ, if you can bear the price and can get DEC to sell you one - it is the MSCP Proto- col Description and Programmer's Kit. [no, I don't have one, either] ================================================================================ Note 44.4 RDxx FCT Format 4 of 7 EISNER::KILLEEN "Jeff Killeen" 3 lines 12-SEP-1987 21:22 -< XT2190 >- -------------------------------------------------------------------------------- > RD54 MAXTOR (model?) RQDX3 MAXTOR XT2190 I believe ================================================================================ Note 44.5 RDxx FCT Format 5 of 7 EISNER::HAHN 10 lines 21-APR-1988 21:47 -< Where to buy? >- -------------------------------------------------------------------------------- 44.3 notes the following RD52 equivalent can you give me a lead as to a vendor: DEC DRIVE MODEL REAL MANUFACTURER/MODEL MINIMUM CONTROLLER REQ'D --------------- ----------------------- ------------------------ RD52 (primary) QUANTUM Q-540 RQDX1 W/ ROM UPGRADE 1 RD52 (alternate) ATASI 3046 RQDX1 W/ ROM UPGRADE 1 Welcome any help. Pierre ================================================================================ Note 44.6 RDxx FCT Format 6 of 7 EISNER::KILLEEN "Jeff Killeen" 13 lines 21-APR-1988 21:58 -< THE SAFE CHOICE >- -------------------------------------------------------------------------------- > -< Where to buy? >- > > 44.3 notes the following RD52 equivalent can you give me a lead > as to a vendor: If you want a safe buy I would go with Trimarchi. They will supply the skid plate, supply cables, and format (tell them which controller). They charge 1 or 2 hunderd more but it is worth the trouble it saves. 818 234-5659 P.S. Why not a RD53? - The price diff is very small. ================================================================================ Note 44.7 RDxx FCT Format 7 of 7 EISNER::KENNEDY "Terry Kennedy" 20 lines 21-APR-1988 23:47 -------------------------------------------------------------------------------- > If you want a safe buy I would go with Trimarchi. They will supply the > skid plate, supply cables, and format (tell them which controller). They > charge 1 or 2 hundred more but it is worth the trouble it saves. See discussion in 93.* for other vendors. Yes, Trimarchi will sell you the drive. However, 'street' is still about *half* of Trimarchi's price. Also, my experiences with Trimarchi have been interesting to say the least - They will do anything to help you - 'for a price'. I once was attemting to integrate a RD53 'clone' into a box (w/controller) that they sold me. They kindly offered to make my RD53 'DEC compatible' for only $700 - which is the street value of the drive. The actual change they would have performed was soldering in 1 jumper! Not bad, huh? Also, you might have a problem if you work for a large organization - they do not tell you, but their terms are net 10. > P.S. Why not a RD53? - The price diff is very small. Agreed - as above, $700 from sources listed in 93.*. Or $1700 for a RD54. Of course, you'll need RQDX2 and RQX3, respectively...  ================================================================================ Note 45.0 No SIXEL through VT125 printer port? 12 replies EISNER::LEDERMAN "Bart Z. Lederman" 23 lines 9-SEP-1987 17:35 -------------------------------------------------------------------------------- I have what appears to be a hardware problem (or possibly a hardware design problem). I have an LA12 connected to the printer port of my VT125. I want to print documents which contain "pictures" in SIXEL format. If I send ReGIS commands to the VT125 to draw a picture which includes a "dump screen to printer port" command the LA12 prints out the SIXEL picture as it should (or at least as well as can be expected from an LA12): however, embedding ReGIS commands into text and including a screen dump command doesn't seem to work, and some of my pictures aren't in ReGIS. If I connect the LA12 directly to the VAX and TYPE the file containing the SIXEL data it prints correctly, so I'm fairly certain the problem doesn't lie in the printer. If I TYPE a file to the VT125 which contains the "printer port on" command ([5i) at the beginning of the file and the "printer port off" command ([4i) at the end, the text goes through the VT125 and out the printer port to the LA12: except that when it comes to the portion containing the SIXEL data, the picture draws on the screen of the VT125 and does not pass through to the printer port. Does anyone know why the VT125 is sucking out the SIXEL data and how I get it to pass that on along with the ASCII text? ================================================================================ Note 45.1 No SIXEL through VT125 printer port? 1 of 12 EISNER::HAHN 6 lines 10-SEP-1987 00:22 -< Will check further >- -------------------------------------------------------------------------------- I have to go back and check the manual, the CTRL PRINT SCREEN will only print the characters while the SHIFT PRINT SCREEN prints the graphics on the VT240. I do not remember the combinations on the VT125, will look it up, and reply further. I know that I had no problem graph printing to a LA50. Pierre ================================================================================ Note 45.2 No SIXEL through VT125 printer port? 2 of 12 EISNER::LEDERMAN "Bart Z. Lederman" 9 lines 10-SEP-1987 10:11 -< Not quite what I need >- -------------------------------------------------------------------------------- > I have to go back and check the manual, the CTRL PRINT SCREEN will > only print the characters while the SHIFT PRINT SCREEN prints the I'm afraid neither of these is what I need. I need to have the proper control characters in the file: I can't enter a TYPE command and then hit the print screen key on my terminal and get a multi-page document to print out. Besides, on my VT125 neither of the above commands work to print out the screen anyway. ================================================================================ Note 45.3 No SIXEL through VT125 printer port? 3 of 12 EISNER::HAHN 19 lines 10-SEP-1987 13:50 -< VT125 SCREEN HARD COPY >- -------------------------------------------------------------------------------- In the VT125 user guide (EK-VT125-UG-001) p 5-56/57 the following: Screen Hard Copy Output: S(H (P[]) [][- -------------------------------------------------------------------------------- > S(H (P[]) [][- -------------------------------------------------------------------------------- Let see if I understand the problem: 1) you have mixed ASCII and REGIS on the screen 2) you want to dump the screen to a LA12 The S(H... command is a REGIS command and will only dump the REGIS (by converting to SIXELS) portion of the screen. Since the REGIS command only dumps REGIS I would make sure the ASCII was put out in a graphic mode. Pierre ================================================================================ Note 45.6 No SIXEL through VT125 printer port? 6 of 12 EISNER::LEDERMAN "Bart Z. Lederman" 19 lines 10-SEP-1987 21:36 -< NOT A SCREEN DUMP >- -------------------------------------------------------------------------------- > 1) you have mixed ASCII and REGIS on the screen > 2) you want to dump the screen to a LA12 NO, I do NOT want to dump a screen to the LA12. I have a file which contains ASCII text and images (preferably in SIXEL). It will print correctly if the LA12 is connected directly to the VAX, but normally it isn't; it's connected to the printer port on my VT125. I want the entire file to go through the VT125 and end up on the LA12 without being modified or affected by the VT125. I can do it with text, but the VT125 won't pass on the SIXEL commands. I tried ReGIS as an alternative, but it didn't work either, and I really don't want to do ReGIS, I would rather do SIXEL. Does this explain what I want to do? ================================================================================ Note 45.7 No SIXEL through VT125 printer port? 7 of 12 EISNER::KENNEDY "Terry Kennedy" 13 lines 10-SEP-1987 21:47 -< Outboard solution? >- -------------------------------------------------------------------------------- > I want the entire file to go through the VT125 and end up on the LA12 > without being modified or affected by the VT125. If you don't mind an outboard solution, you can get something like the Code Operated Switch (COS-4P) from Black Box. It will hook up like an A-B switch, and will respond to an escape sequence to switch the CPU from A to B, where one is the 125 and the other is the printer. This means you can't use any of the 125 screen copy stuff, because the printer isn't connected to the 125 anymore, but from what I understand, that doesn't matter to you. You can call Black Box technical assistance 8:30AM-6:30PM EST @ (412) 746-5564. The COS-4P costs $450.00 E/U Q1. ================================================================================ Note 45.8 No SIXEL through VT125 printer port? 8 of 12 EISNER::TABOR "Bill Tabor" 11 lines 11-SEP-1987 18:19 -< Try [5i >- -------------------------------------------------------------------------------- > I want the entire file to go through the VT125 and end up on the LA12 > without being modified or affected by the VT125. Have you tried "[5i". From what I have read everything after this escape sequence is suppose to be passed to the printer port until you sent an "[4i". I know this will allow passing the font changes and other escape sequences through to the la210 that is slaved to my VT220. ================================================================================ Note 45.9 No SIXEL through VT125 printer port? 9 of 12 EISNER::LEDERMAN "Bart Z. Lederman" 9 lines 12-SEP-1987 19:01 -< That is exactly what DOESN'T work. >- -------------------------------------------------------------------------------- > Have you tried "[5i". From what I have read everything after > this escape sequence is suppose to be passed to the printer port > until you sent an "[4i". That's what I thought too: I'm sure I said in my first note that I already have these escape sequences at the beginning and end of my file, and everything goes through EXCEPT the SIXEL data. That is what is so frustrating: all other data in the file goes through but NOT the pictures in SIXEL. ================================================================================ Note 45.10 No SIXEL through VT125 printer port? 10 of 12 EISNER::KILLEEN "Jeff Killeen" 9 lines 12-SEP-1987 19:33 -< QUESTIONS >- -------------------------------------------------------------------------------- DUMB QUESTIONs... Bart is the SIXEL stuff 8 bit chars? If so have you tried any other 8 bit data? I had a VT100 clone that changed all the incomming 8 bit to 7 bit on it's way to the printer port. ================================================================================ Note 45.11 No SIXEL through VT125 printer port? 11 of 12 EISNER::LEDERMAN "Bart Z. Lederman" 12 lines 13-SEP-1987 16:35 -< No 8 bits. >- -------------------------------------------------------------------------------- > Bart is the SIXEL stuff 8 bit chars? The nature (definition) of SIXEL precludes 8 bit data. In any event, the stuff I'm working with isn't 8 bit. > If so have you tried any other 8 bit data? Yes, that's not the problem. The problem is that the VT125 refuses to pass any SIXEL data on: nothing at all comes out of the printer port, but the SIXEL picture appears on the CRT screen. All other data does go through the printer port to the LA12. ================================================================================ Note 45.12 No SIXEL through VT125 printer port? 12 of 12 EISNER::HANRAHAN "Jamie Hanrahan" 20 lines 15-SEP-1987 16:08 -< I don't think it'll work. >- -------------------------------------------------------------------------------- I take it that displaying the sixel data on the screen, then doing a graphics screen print, won't work, perhaps because your sixel pictures go beyond the bounds of the VT125's coordinates, or some such? I believe you are probably out of luck (barring an outboard hardware solution like the code-operated switch mentioned in an earlier note), because of the the way in which the VT125 is "added on" to the basic VT100. The VT125 graphics board connects "in between" the external RS232 connector and the place on the VT100 board where the RS232 connector would normally go, and it looks at everything coming in to see if it's interested in it. If it is, it grabs it and processes it, and doesn't pass it on to the VT100 board. And the printer is connected to the VT100... oh, well. (The video is handled the same way: Video comes out of the VT100, goes to the graphics board where the graphics stuff is "or"d in, and thence to the monitor... this is why text and graphics planes are SO independent on that beast!)  ================================================================================ Note 46.0 New MicroVAXen? 13 replies EISNER::PRIGOT "Jonathan M. Prigot" 4 lines 9-SEP-1987 22:03 -------------------------------------------------------------------------------- Now that the µVAX III (a.k.a. 35xx and 36xx) have been announced at DECworld, how about some details from those in the know. I would never confuse a marketing tool like DECworld for a real conference like the DECUS symposia. ================================================================================ Note 46.1 New MicroVAXen? 1 of 13 EISNER::KILLEEN "Jeff Killeen" 1 line 9-SEP-1987 22:47 -< E-STORE >- -------------------------------------------------------------------------------- The E-STORE has a write-up on it. ================================================================================ Note 46.2 New MicroVAXen? 2 of 13 EISNER::BYRNE_C "Charlie Byrne" 2 lines 10-SEP-1987 14:30 -< What is E - STORE >- -------------------------------------------------------------------------------- If I said I had no idea what the E-STORE is, would you hold it against me? ================================================================================ Note 46.3 New MicroVAXen? 3 of 13 EISNER::KILLEEN "Jeff Killeen" 9 lines 10-SEP-1987 19:42 -< MORE INFO WAS NEEDED >- -------------------------------------------------------------------------------- > If I said I had no idea what the E-STORE is, would you hold it > against me? My fault not yours. It is DEC's Electronic Store. Call 1-800-332-3366 (1-80-DEC-DEMO) ================================================================================ Note 46.4 New MicroVAXen? 4 of 13 EISNER::HYDE "Mark Hyde" 8 lines 11-SEP-1987 13:01 -< nit pickin >- -------------------------------------------------------------------------------- re: .0 >Now that the µVAX III (a.k.a. 35xx and 36xx) have been announced Please note that the MicroVAX 3500/3600 is NOT a "MicroVAX III"! mark ================================================================================ Note 46.5 New MicroVAXen? 5 of 13 EISNER::BYRNE_C "Charlie Byrne" 8 lines 11-SEP-1987 14:06 -< 8400? Are rumors allowed? >- -------------------------------------------------------------------------------- >Now that the µVAX III (a.k.a. 35xx and 36xx) have been announced I see there was no announcement of the rumoured 8400 (are we allowed to talk about rumors here?). Any one have any word on a BI system that is obviously needed to fill the large uniprocessor gap between the 8250 (Approx 1.2 Mips, $80K) and the 8530 (Approx 4 Mips, $300K)? ================================================================================ Note 46.6 New MicroVAXen? 6 of 13 EISNER::KILLEEN "Jeff Killeen" 22 lines 11-SEP-1987 16:36 -< RUMOR HAS IT >- -------------------------------------------------------------------------------- > I see there was no announcement of the rumoured 8400 (are we > allowed to talk about rumors here?). Any one have any word on There is nothing in DECUServe policy that states no can not talk about rumors - however - since the system is suppose to be a store house of useful knowledge I would label any discussion based on rumor as a "rumor has it" discussion. My other DECUS job is to work with DEC on product planning. Thanks to C. Matco and Digital Review DEC has become tense about non-disclosure information. The user community has been hurt in general by Matco and company because any non-disclosure information give outside the company has to be cleared by the Policy & Pricing committee. This means Digital is telling it's customers less than it use to and we have Digital Review to thank for it. Please be careful not to violate any non-disclosures you may have with DEC. We don't need DECUS being identified as a source of leaks! ================================================================================ Note 46.7 New MicroVAXen? 7 of 13 EISNER::PRIGOT "Jonathan M. Prigot" 12 lines 13-SEP-1987 22:17 -< So noted >- -------------------------------------------------------------------------------- > re: .0 > >Now that the µVAX III (a.k.a. 35xx and 36xx) have been announced > > Please note that the MicroVAX 3500/3600 is NOT a > "MicroVAX III"! > > mark So noted. (I heard the story of why, too. See note 46.6) ================================================================================ Note 46.8 New MicroVAXen? 8 of 13 EISNER::BYRNE_C "Charlie Byrne" 10 lines 14-SEP-1987 11:20 -< A Touch of Grey >- -------------------------------------------------------------------------------- RE: < Note 46.6 by EISNER::KILLEEN "Jeff Killeen" > -< RUMOR HAS IT >- > Please be careful not to violate any non-disclosures you may have with > DEC. We don't need DECUS being identified as a source of leaks! I certainly would not violate non-disclosure agreements I have with DEC. Question though: Once a "rumoured product" is discussed in mass-media, do further open dicussions about it by the "non-disclosees" constitute violations of trust to DEC? I suppose this may come down to ethical rather than legal opinions. ================================================================================ Note 46.9 New MicroVAXen? 9 of 13 EISNER::HASSINGER "Bob Hassinger" 2 lines 14-SEP-1987 11:47 -< Legal issues may be more restrictive than ethical ones. >- -------------------------------------------------------------------------------- ================================================================================ Note 46.10 New MicroVAXen? 10 of 13 EISNER::HAHN 8 lines 14-SEP-1987 12:18 -< non-disclosure=~ secret >- -------------------------------------------------------------------------------- Non-disclosure is like classified documents, you cannot even acknowledge that you have knowledge about the subject. Even when the rumor mills comment on it, you cannot affirm or negate the rumor. You have insider knowledge which must remain yours and yours alone. You are released from your bond when DEC releases the information. Pierre ================================================================================ Note 46.11 New MicroVAXen? 11 of 13 EISNER::KILLEEN "Jeff Killeen" 8 lines 14-SEP-1987 12:42 -< YOU ARE GAGED >- -------------------------------------------------------------------------------- > I certainly would not violate non-disclosure agreements I have with > DEC. Question though: Once a "rumoured product" is discussed in mass-media, > do further open dicussions about it by the "non-disclosees" constitute > violations of trust to DEC? I suppose this may come down to ethical > rather than legal opinions. Until it is an announced product you cannot comment - period. ================================================================================ Note 46.12 New MicroVAXen? 12 of 13 EISNER::BYRNE_C "Charlie Byrne" 8 lines 15-SEP-1987 15:01 -< Lucky I was not a big mouth :-O >- -------------------------------------------------------------------------------- < Note 46.11 by EISNER::KILLEEN "Jeff Killeen" > -< YOU ARE GAGED >- > Until it is an announced product you cannot comment - period. In that case, I am glad that any non-disclosure info I have is not related to any products previously mentioned. Thanks for the clarifications. ================================================================================ Note 46.13 New MicroVAXen? 13 of 13 EISNER::POOLE "Clyde T. Poole" 5 lines 18-SEP-1987 20:19 -< This Belongs in Soapbox Conference >- -------------------------------------------------------------------------------- Ok folks enough. Please take this discussion to the Soapbox conference. I am setting this note NOWRITE to incourage you to move the discussion elsewhere. ctp  ================================================================================ Note 47.0 Can You Speed up an 11/70? 65 replies EISNER::BYRNE_C "Charlie Byrne" 16 lines 10-SEP-1987 15:08 -------------------------------------------------------------------------------- I have an 11/70 running RSX-11M+ V2.1 that is totally saturated, both CPU and busses. We are moving the application to VAXen, but meanwhile I need some short term help. The software is optimized to @#$%. This is a real-time application, and we cannot split it across another CPU. Benchmarks run on 11/84 showed great CPU improvement but the system died when the Unibus traffic started piling up. The system gets input from 1Mb DMR, manipulates the data, stores it on a Bulk Core device (Dataram), then sends it out over both another 1 Mb DMR and a 19.2 COMM-IOP-DUP-11 (yes someone still has one!). One area I am interested in looking at is faster memory. We have the original DEC memory in the system. Are there any third parties out there still making 11/70 memory that is faster than DEC's? Any other ideas out there for speeding up an 11/70 ? The only help I can see in the near future is going to M+ V3 which is supposedly faster, but any quick hardware ideas would be appreciated. ================================================================================ Note 47.1 Can You Speed up an 11/70? 1 of 65 EISNER::KASPER "Beverly T. Kasper" 15 lines 10-SEP-1987 16:10 -< What size are the memory boards? >- -------------------------------------------------------------------------------- If your "original DEC memory" is those 32K boards we had in our 11/70 (since replaced by an 11/84), I'd bet that you could, in fact, get faster memory than that. I can't really help with details; We had a couple of DataRam 256K boards in our system, but they were added long before I got there. If the system is on contract, and you have a good relationship with your Field Service Reps, you might be able to get them to let you swap the boards out for denser ones long enough to run benchmarks. Midwest has used 11/70 memory; I doubt they're the only ones. As for going to V3.0, it may be faster, but your application might need work to make it play. Do you really want to go through that? It would probably be worth it if you were planning to stay with M+, but I'd be wary of it as a short-term fix. Also, keep in mind that part of the advertised speed increase is due to disk caching, which is sometimes Not A Win for applications, depending on many factors. ================================================================================ Note 47.2 Can You Speed up an 11/70? 2 of 65 EISNER::BYRNE_C "Charlie Byrne" 8 lines 10-SEP-1987 17:09 -< The memory now is 32KB MK-11 >- -------------------------------------------------------------------------------- > < Note 47.1 by EISNER::KASPER "Beverly T. Kasper" > > -< What size are the memory boards? >- > If your "original DEC memory" is those 32K boards we had in our 11/70 (since > replaced by an 11/84), I'd bet that you could, in fact, get faster memory > than that. The memory in the machine is 1.5 MB of MK-11's, which I believe is "the original" 16KW MOS chip memory supplied with the 70's. ================================================================================ Note 47.3 Can You Speed up an 11/70? 3 of 65 EISNER::KILLEEN "Jeff Killeen" 2 lines 10-SEP-1987 19:45 -< HMMMMMMM.... >- -------------------------------------------------------------------------------- Did you know the VAX 750 and 11/70 use the same array cards. Check the M numbers. ================================================================================ Note 47.4 Can You Speed up an 11/70? 4 of 65 EISNER::ETHINGTON "Superior Klingon Technology" 37 lines 12-SEP-1987 15:11 -< Memory Interleaving >- -------------------------------------------------------------------------------- There are still a few thrid party outfits making 11/70 memory, but I don't honestly know whether or not they are any faster that the MK11-C arrays from DEC. A recent Digital Review had a big table of all third party memory vendors; check it for the names of the outfits offering 3rd party memory, and give them a call. Another option, which used to be obscenely expensive but, with current used equipment prices is actually quite reasonable, is to install a second MK11-C memory box on the 11/70. You can then turn the thumbwheels on the two memory boxes and select 4 way interleaving. Interleaving is a method of speeding up memory access times by performing accesses to more than one memory chip in parallel, as opposed to accessing fewer chips multiple times. I'm not a hardware expert, but the way I think it works is along these lines; a chip can roll out data over say 8 pins in one memory cyle, so you can get 1 byte per memory cycle per chip. If you want 16 bits, you can either read the same chip twice at successive addresses, but taking twice the access time; or, with the appropriate hardware, you can in parallel talk to two chips simultaneously, getting the 16 bits in one memory cyle instead of two. The 11/70 CacheBus actually performs 32 bit fetches to load the cache; going to 4 way interleaving pays off. We installed it on a 70 and gained about 15% raw CPU performance. The MK11-C memory controller performs 2 way interleaving within the single box, but 11/70s can support up to 4 MK11-C controllers and use up to 8 way interleaving. You'll probably have to read the controller manual and set the switches yourself; field circus doesn't seem to know what interleaving is and tends to look at you with a blank stare. M+ V3 has some minor performance improvements in the way of interrupt latency that may be significant on an application as heavily I/O bound as yours. Between the memory change and going to V3, you might reasonably expect to gain about 20% performance. Hope this is of some help..... Jerry ================================================================================ Note 47.5 Can You Speed up an 11/70? 5 of 65 EISNER::KENNEDY "Terry Kennedy" 14 lines 12-SEP-1987 20:59 -< Watch FS w/ interleaved memory >- -------------------------------------------------------------------------------- > controllers and use up to 8 way interleaving. You'll probably have > to read the controller manual and set the switches yourself; field > circus doesn't seem to know what interleaving is and tends to look > at you with a blank stare. Worse than that - here we had a big fight with FS about interleaving - they refused to support the machine until we turned interleaving off. We fought that up to regional level, and got to keep our interleaving. After that, however, we noticed that after PM's, the machine was slower. FS was sneaking into the box during PM and disabling interleaving! (They get up- set when the diagnostic bombs because of interleaved memory). So, if you do this, watch your FS guy every PM. Terry ================================================================================ Note 47.6 Can You Speed up an 11/70? 6 of 65 EISNER::BYRNE_C "Charlie Byrne" 11 lines 14-SEP-1987 11:30 -< WIll check out with in house techs >- -------------------------------------------------------------------------------- RE: < Note 47.4 by EISNER::ETHINGTON "Superior Klingon Technology" > -< Memory Interleaving >- > ...read the controller manual and set the switches yourself; field > circus doesn't seem to know what interleaving is and tends to look > at you with a blank stare. Luckily we have in-house Field Circus for our PDP's. I'll check this out and see if the stares are any blankless. :-) Thanks. ================================================================================ Note 47.7 Can You Speed up an 11/70? 7 of 65 EISNER::BYRNE_C "Charlie Byrne - Reuters Info Svcs" 7 lines 21-SEP-1987 15:41 -< Any Software changes required? >- -------------------------------------------------------------------------------- I have scrounged around and gotten another MK-11, and we are just about ready to try out quad interleaving. Is there any reason to expect any software problems? I had imagined all along this would be totally transparent; however I was just asked to state that fact without reservation and I could not because obviously I am not sure. Can anyone verify that splitting memory into 2 MK-11 controllers will have require no S/W changes? Thanx. ================================================================================ Note 47.8 Can You Speed up an 11/70? 8 of 65 EISNER::KENNEDY "Terry Kennedy" 7 lines 21-SEP-1987 20:02 -< User code, no; O/S, maybe? >- -------------------------------------------------------------------------------- > Can anyone verify that splitting memory into 2 MK-11 controllers > will require no S/W changes? Thanx. About the only thing that can bite you is a hard-coded timing loop, which will execute faster. User programs should be unaffected. You may have to tell the O/S about it. You'd need to find an appropriate RSX guru here (We are talking RSX, right?) ================================================================================ Note 47.9 Can You Speed up an 11/70? 9 of 65 EISNER::LEDERMAN "Bart Z. Lederman" 6 lines 21-SEP-1987 21:06 -< No Changes Required >- -------------------------------------------------------------------------------- I ran with a number of 11/70s, and when we flipped the switches to go between interleaved and non-interleaved memory between two memory chassis, we did absolutely nothing whatsoever to your RSX-11M or RSX-11M-Plus systems and they booted and ran. No changes to O/S or user programs. ================================================================================ Note 47.10 Can You Speed up an 11/70? 10 of 65 EISNER::MCCARTHY 15 lines 22-SEP-1987 21:07 -< MK11 interleaving requires no software change >- -------------------------------------------------------------------------------- As Bart mentions, you do not need to change the software to run MK11s interleaved. We have four boxes of MK11 memory that we've been running for 8 years (the RSX-11M-PLUS development system) and that's the one thing about the memory we never had to change the software for. A word of caution, however: On the 11/74, we did change the boot rom to check for having half of an interleaved memory. The only thing to worry about here is that the symptoms of having every third and fourt word of memory missing can be very bizarre, so check for that first in any bizarre situation. -Brian ================================================================================ Note 47.11 Can You Speed up an 11/70? 11 of 65 EISNER::BYRNE_C "Charlie Byrne - Reuters Info Svcs" 15 lines 24-SEP-1987 16:35 -< Dataram sells 11/70 memory >- -------------------------------------------------------------------------------- > < Note 47.4 by EISNER::ETHINGTON "Superior Klingon Technology" > > -< Memory Interleaving >- > There are still a few thrid party outfits making 11/70 memory, but > I don't honestly know whether or not they are any faster that the > MK11-C arrays from DEC. A recent Digital Review had a big table > of all third party memory vendors; check it for the names of the > outfits offering 3rd party memory, and give them a call. Thanks. Digital Review July 13 1987 listed no less than 321 DEC and 3rd party memory boards produced by 25 vendors. There was exactly 1 (one) board for the 11/70 Cachebus. That was the DR-175 manufactured by Dataram. Unfortunatly Dataram says this board does not have a performance edge over DEC's MK11, it is there because no one else (including DEC) sells new 11/70 memory :-( ================================================================================ Note 47.12 Can You Speed up an 11/70? 12 of 65 EISNER::LEDERMAN "Bart Z. Lederman" 12 lines 24-SEP-1987 21:10 -< Caveat Emptor!!!! >- -------------------------------------------------------------------------------- We tried to put Dataram memory on our 11/70 once: following the installation instructions didn't work. They sent out new instructions: they didn't work either. Datatram sent out a serviceman to install it: he couldn't get it to work. Unfortunately, we had a DEC Field service man there to see that the Dataram person didn't damage anything, but we forgot to tell him not to make suggestions, and of all things he was the one to figure out what Dataram was doing wrong! The memory appeared to work for a while, but we started getting random crashes and system failures, which were eventually all traced back to the Dataram memory. We finally sent the memory back, but I don't think they ever refunded any of our money, and it cost us a LOT of money for the memory, the lost time, our effort, buying good DEC memory to replace it, etc. ================================================================================ Note 47.13 Can You Speed up an 11/70? 13 of 65 EISNER::BYRNE_C "Charlie Byrne - Reuters Info Svcs" 149 lines 29-SEP-1987 15:56 -< Results of memory interleaving on PDP >- -------------------------------------------------------------------------------- OK folks, here are the bizzare results. We set up two 11/70's basically identically configured with the following difference: Machine NIL has 768KW memory non- interleaved. Machine IL has 1024KW memory hanging off two MK-11 controllers. One MK-11 has start address 000, interleave switch set to 2, force panel switch enabled. Second MK-11 is same but interleave switch set to 3. This is as per instructions in MK-11 manual for interleaving with 2 MK-11's. There is nothing running on either system. The program is task built with priority 150. just in case. We create test program (below) which is just a loop that moves 8KB from one location to another using MOV instructions (in real life we use a modified version of $BLXIO for block data transfer). The program sends a message to the console every 10 seconds how many times it has executed the loop. We let this run for about a minute and average the results. *Note* that though runs for system NIL and system IL are listed adjacent below, they were done at different times (we actually moved the whole system pack and re-booted for each test). We then run a second set of tests where we install the test program with a different name and run two copies simultaneously. I figured this would give a truer flavor to the system activity. Here are the results of both test sets: *Single Task Running* System IL System NIL --------------------------- ------------------------------------ Called TIMER 881. times. Called Timer 880. times. Called TIMER 880. times. Called Timer 880. times. Called TIMER 881. times. Called Timer 880. times. Called TIMER 880. times. Called Timer 880. times. Called TIMER 881. times. Called Timer 880. times. Average = 880.6 Average = 880.0 Ok, so the difference is noise, and it looks like interleaving is not doing a damn thing, or worse. (BTW, we first tested the interleaving on our real application (see .0 for details of application) and got no difference. I couldn't believe that, hence these tests). 880/10 seconds * 8KBytes = 704K Bytes/second Is that reasonable even considering that it is being done about as inefficiently as possible? Now I say, well maybe interleaving won't help with a single task because everything is basically happening syncronously. Lets run two copies of this at the same time and see if we can cause some overlapping that might make a difference. So, INSTALL LOOP/TASK=LOOP2 RUN LOOP RUN LOOP2 ...after a minute or so ABO LOOP ABO LOOP2 Throw out the first few iterations while things are starting up and: *Two Tasks Running* System IL System NIL --------------------------- ------------------------------------ Called TIMER 446. times. Called Timer 448. times. Called TIMER 444. times. Called Timer 449. times. Called TIMER 446. times. Called Timer 447. times. Called TIMER 440. times. Called Timer 444. times. Called TIMER 444. times. Called Timer 450. times. Average = 444.0 Average = 447.6 The Non-Interleaved system is transfering more data per 10 seconds than the interleaved! :-(((( I am at a loss here, the only thing I can think of is that somehow that hardware is not properly configured, however looking at the MK-11 controller handbook, everything looks ok. Any help would be appreciated because the walls are getting a beating around here from people banging their heads against them. PS: Yes, I know the program below is not perfect, that is not the issue here. START: MRKT$S ,#10.,#2,#MAST 10$: MOV #IN.BUFFER,R1 MOV #OUT.BUFFER,R0 MOV #XFR.SIZE/2,R2 CALL TIMER ADD #1, COUNT+2 ADC COUNT BR 10$ TIMER: ; This is the test loop MOV (R1)+,(R0)+ SOB R2, TIMER RETURN MAST: MOV R0,-(SP) ;Save MOV R1,-(SP) MOV R2,-(SP) MOV #OUTBUF,R0 ; Set up message MOV #PARAM,R1 MOV #ARGBLK,R2 MOV #COUNT,ARGBLK CALL $EDMSG SUB #OUTBUF,R0 QIOW$S #IO.WLB,#5,#1,,,,<#OUTBUF,R0,#40> ; reset MRKT$S ,#10.,#2,#MAST CLR COUNT+2 CLR COUNT MOV (SP)+,R2 MOV (SP)+,R1 MOV (SP)+,R0 TST (SP)+ ASTX$S COUNT: .WORD 0,0 IOSB: .WORD 0,0 ARGBLK: .WORD COUNT PARAM: .ASCIZ /Called TIMER %T. times./ .EVEN OUTBUF: .BLKB 512. .EVEN IN.BUFFER: .BLKB 8.*1024. XFR.SIZE = . - IN.BUFFER .EVEN OUT.BUFFER: .BLKB 8.*1024. .END START If you have gotten this far, thanks for such help! ================================================================================ Note 47.14 Can You Speed up an 11/70? 14 of 65 EISNER::KENNEDY "Terry Kennedy" 5 lines 29-SEP-1987 19:30 -< Cache? >- -------------------------------------------------------------------------------- Well, I'm not a '70 expert, but... Do 70's have cache? If so, your sample program is so small, you're running entirely in cache and never going out to 'real' memory (once the code is loaded). Try a large Fortran program, with lots of array references. If you still don't get anything, then I'd say to check that your memory is really set up correctly. ================================================================================ Note 47.15 Can You Speed up an 11/70? 15 of 65 EISNER::LEDERMAN "Bart Z. Lederman" 15 lines 29-SEP-1987 20:36 -< Need READ, no cache >- -------------------------------------------------------------------------------- > Well, I'm not a '70 expert, but... Do 70's have cache? Yes, 70's have cache, and the program itself is probably all in cache. I think the reason you are not seeing much improvement is that none of the data that's being moved can really be cached. Neither would I expect interleaving to help. The purpose of interleaving is to get more data transferred from memory to the CPU/Cache on reads. Writes go directly through cache to memory one word (or whatever) at a time to insure memory always gets updated (11s don't have write-back cache, at least on on 70s). To see if interleaving is helping, I'd try a program that did mostly reads on a large array in memory. The program and data should be large enough so that it won't fit into cache, so the change in speed in main memory can be seen. ================================================================================ Note 47.16 Can You Speed up an 11/70? 16 of 65 EISNER::BYRNE_C "Charlie Byrne - Reuters Info Svcs" 14 lines 30-SEP-1987 11:22 -< What's the difference once beyond 2KB? >- -------------------------------------------------------------------------------- RE: < Note 47.15 by EISNER::LEDERMAN "Bart Z. Lederman" > > -< Need READ, no cache >- > To see if interleaving is helping, I'd try a program that did >mostly reads on a large array in memory. The program and data should >be large enough so that it won't fit into cache, so the change in speed >in main memory can be seen. 1) RE: Program running in cache. As I mentioned, we ran these tests because our application (which consists of perhaps 10 large tasks) showed *no* improvement. So I don't think the code running in cache is a factor here. 2) I don't see why a LARGE array will show anymore than the 8KB area I am using. The 11/70 cache is 2KB; don't I guarantee invalidation as soon as I reference beyond that? ================================================================================ Note 47.17 Can You Speed up an 11/70? 17 of 65 EISNER::BYRNE_C "Charlie Byrne - Reuters Info Svcs" 12 lines 6-OCT-1987 15:02 -< Oh my God! He's got a gun! >- -------------------------------------------------------------------------------- Well, the failure of interleaving to help me seems to have stumped the experts! (that oughta get any self-respecting guru's attention). :-) So the kid is getting desperate. I see in the October 1987 DEC PRO (page 68) where a claim is made that once upon a time (? why not now ?) there was a way to change the 11/70 clock speed, and with a properly tuned board (whatever that means) you could get a 20% inprovement (in what?). I've never heard of this one. Sounds extremely dangerous. Comments? ================================================================================ Note 47.18 Can You Speed up an 11/70? 18 of 65 EISNER::DELARISCH 18 lines 6-OCT-1987 20:10 -< Performance Enhancement PKG for the PDP 11/70 >- -------------------------------------------------------------------------------- There has been someds in the DEC Pro and Digital Review for a company who has performance improvment products for various PDP products. Its one of the original UNIBUS designer's from DEC who started his own company. They offer very HIGH SPEED memory boards and UNIBUS isolators. They postcard infered that they had an enhancement package for the PDP 11/70. You may want to check into it. Please Note that I DO NOT WORK FOR NOR HOLD ANY INTEREST IN THE FOLLOWING COMPANY (so there)! Digital Data Systems, Inc. 1551 NW 65th Avenue Plantation, FL 33313-9951 (305) 792-3290 ================================================================================ Note 47.19 Can You Speed up an 11/70? 19 of 65 EISNER::KENNEDY "Terry Kennedy" 27 lines 6-OCT-1987 20:15 -< Not as crazy as it sounds >- -------------------------------------------------------------------------------- > So the kid is getting desperate. I see in the October 1987 DEC PRO > (page 68) where a claim is made that once upon a time (? why not > now ?) there was a way to change the 11/70 clock speed, and with > a properly tuned board (whatever that means) you could get a 20% > improvement (in what?). Well, I haven't seen the article and I don't have a '70 (yet), but here are some thoughts. DEC generally designs hardware with a very wide safety margin, since the maximum 'safe' speed of the parts involved varies. If you have a board with parts which just happen to exceed spec, it is possible to increase the CPU speed by changing various parts in the system timing circuits. Notable cases are the oscillator in the 11/23 (KDF11) CPU. It is possible to run it at 150% of rated speed on most boards. With some re-design, VAX 750's can also run about 30% faster (remember, the more complicated it is, the harder it is to get a useful improvement). So, it is probably possible to do the same to a '70. You speed up the system until it crashes, and then back off about 10-15%. Of course, later on you WILL have problems and have to drop back some more. Your best bet is to buy a complete speed-up kit from someone (if someone makes one for the '70) and get a guarantee. The whole thing is not as crazy as it sounds - this is a major after-market area for IBM AT computers. ================================================================================ Note 47.20 Can You Speed up an 11/70? 20 of 65 EISNER::KILLEEN "Jeff Killeen" 26 lines 6-OCT-1987 21:35 -< WISHES >- -------------------------------------------------------------------------------- > They offer very HIGH SPEED memory boards and UNIBUS isolators. > Digital Data Systems, Inc. > 1551 NW 65th Avenue > Plantation, FL 33313-9951 > (305) 792-3290 I was in contact with a RSTS site that tried this on a PDP-11/44. They gave up and sent the boards back. The kit worked fine when the system was configured with all DEC boards. When they had their EMULEX TC03 controller attached they would get bus errors. Their feeling was great care should be used if your system contains non-DEC boards. Finding out that EMULEX did not design a safety margin into there boards would not come as a surprise. I believe this kit is not intended for an 11/70 - I may be wrong. Speeding up the UNIBUS would not gain much if all the disks are on the MASSbus. I have oft wondered why someone does not come out with a fast 4MB memory board to install in the CPU box to replace to MK11 memory cabinet. With that board and SC71 disk controller that would be a fast and cheep system to maintain - most likely faster than 11/84 in a commercial production shop. ================================================================================ Note 47.21 Can You Speed up an 11/70? 21 of 65 EISNER::BYRNE_C "Charlie Byrne - Reuters Info Svcs" 46 lines 7-OCT-1987 08:53 -< Thanks >- -------------------------------------------------------------------------------- >Note 47.18 Can You Speed up an 11/70? 18 of 20 >EISNER::DELARISCH 18 lines 6-OCT-1987 20:10 >They offer very HIGH SPEED memory boards and UNIBUS isolators. >Digital Data Systems, Inc. >(305) 792-3290 Thanks, I have spoken to this company, right now their Unibus offering are not for the 70. They said in late August they were working on something for the 70, so it is about time for me to get back to them. >Note 47.19 Can You Speed up an 11/70? 19 of 20 >EISNER::KENNEDY "Terry Kennedy" 27 lines 6-OCT-1987 20:15 >DEC generally designs hardware with a very wide safety margin, since the >maximum 'safe' speed of the parts involved varies. If you have a board >with parts which just happen to exceed spec, it is possible to increase >the CPU speed by changing various parts in the system timing circuits. > > Notable cases are the oscillator in the 11/23 (KDF11) CPU. It is possible >to run it at 150% of rated speed on most boards. With some re-design, VAX >750's can also run about 30% faster (remember, the more complicated it is, >the harder it is to get a useful improvement). Yes in fact the DEC PRO article I referenced was about "turbo-charging" a 750, and Dave Mallery mentions people also tuning the KDJ11-AA. >Note 47.20 Can You Speed up an 11/70? 20 of 20 >EISNER::KILLEEN "Jeff Killeen" 26 lines 6-OCT-1987 21:35 > They offer very HIGH SPEED memory boards and UNIBUS isolators. >I believe this kit is not intended for an 11/70 - I may be wrong. As noted above you are correct. >I have oft wondered why someone does not come out with a fast 4MB memory >board to install in the CPU box to replace to MK11 memory cabinet. >With that board and SC71 disk controller that would be a fast and >cheep system to maintain - most likely faster than 11/84 in a >commercial production shop. Yes, as I have delved into this problem it seems unbelievable that there is nothing like this out there. Maybe the 11/70 product that this company is working on will fill a void. There has got to be still a *lot* of 70's out there that could use some help. Note: It is really great to see what everyone on this system (DECUServe) is willing to do for others. A refreshing view of humanity. ================================================================================ Note 47.22 Can You Speed up an 11/70? 22 of 65 EISNER::ROCHELEAU "Rod Rocheleau" 16 lines 21-OCT-1987 09:23 -< PEP70 product by Digital Data Systems >- -------------------------------------------------------------------------------- >> < Note 47.0 by EISNER::BYRNE_C "Charlie Byrne" > >> -< Can You Speed up an 11/70? >- >> >>One area I am interested in looking at is faster >>memory. We have the original DEC memory in the system. Are there >>any third parties out there still making 11/70 memory that is faster >>than DEC's? Any other ideas out there for speeding up an 11/70 ? >> If you received the DEXPO WEST 87 product preview, Look on page 4. If not call Digital Data Systems, the product is called a PEP70 and is slated as the "WORLDS FASTEST UNIBUS MEMORY". They claim that performance will double. need I state that I have no connections to DDS. Just information. r ================================================================================ Note 47.23 Can You Speed up an 11/70? 23 of 65 EISNER::BYRNE_C "Charlie Byrne - Reuters Info Svcs" 9 lines 23-OCT-1987 16:09 -< I have ordered a PEP70 >- -------------------------------------------------------------------------------- RE:< Note 47.22 by EISNER::ROCHELEAU "Rod Rocheleau" > > -< PEP70 product by Digital Data Systems >- > If you received the DEXPO WEST 87 product preview, Look on page 4. > If not call Digital Data Systems, the product is called a PEP70 > and is slated as the "WORLDS FASTEST UNIBUS MEMORY". They claim > that performance will double. Thanks. I called them and I now have an evaluation unit on order. Will provide results when available (probably early December). ================================================================================ Note 47.24 Can You Speed up an 11/70? 24 of 65 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 15 lines 22-JAN-1988 18:32 -< PEP-70 looks promising >- -------------------------------------------------------------------------------- > RE:< Note 47.22 by EISNER::ROCHELEAU "Rod Rocheleau" > >> -< PEP70 product by Digital Data Systems >- >> If you received the DEXPO WEST 87 product preview, Look on page 4. >> If not call Digital Data Systems, the product is called a PEP70 >> and is slated as the "WORLDS FASTEST UNIBUS MEMORY". They claim >> that performance will double. > > Thanks. I called them and I now have an evaluation unit on order. > Will provide results when available (probably early December). We have received the PEP-70. Initial results look very interesting. I don't expect double the performance from anything, but if we get say, 15%, that might be just what we need. More details coming shortly... ================================================================================ Note 47.25 Can You Speed up an 11/70? 25 of 65 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 28 lines 18-FEB-1988 21:22 -< Yes, you can speed up an 11/70! >- -------------------------------------------------------------------------------- We have completed our evaluation of the PEP-70 and the results are in. On our application, which is a combination of heavy CPU, heavy Massbus, and heavy Cachebus traffic, system throughput increased by 15 to 20%. We also ran standalone M+ directive tests, and directives increased by 14 to 30%. I/O directives (to NL:) recorded that greatest gains. Some examples (in seconds): Directive No PEP-70 With PEP-70 Percentage Gain --------- --------- ----------- --------------- CLEF/Local .00025 .00020 20 RDEF/Global .00025 .00021 16 SETF/Group .00030 .00025 17 WTLO/Local .00039 .00029 26 WTSE/Group .00048 .00037 23 QIOW/IOS .00137 .00103 25 MAP/Same .00172 .00126 27 MAP/Swap .00168 .00129 23 From all I have heard, the performance gains are extremely application dependent, so 'your mileage may vary'. For those with no 11/70 migration though (see previous notes), and willing to spend the considerable bucks that this board costs, it seems to fill a niché that was giving us a lot of headaches. Standard Disclaimer 102: I have no vested interest in this product. ================================================================================ Note 47.26 Can You Speed up an 11/70? 26 of 65 EISNER::FRISBIE "Alan E. Frisbie" 12 lines 19-FEB-1988 11:36 -< This came along just at the right time >- -------------------------------------------------------------------------------- >> We have completed our evaluation of the PEP-70 and the results >> are in. On our application, which is a combination of heavy >> CPU, heavy Massbus, and heavy Cachebus traffic, system throughput >> increased by 15 to 20%. Thank you! One of my clients is looking for a way to speed up an 11/70 application. It currently takes about a week(!) to run on an 11/70, and an improvement like this will be very welcome. What is the price of this gem (in singles and in quantity 10-20)? Alan ================================================================================ Note 47.27 Can You Speed up an 11/70? 27 of 65 EISNER::KILLEEN "Jeff Killeen" 0 lines 19-FEB-1988 12:53 -< MLP / QUANT 1 / 18K / PEP-70 >- -------------------------------------------------------------------------------- ================================================================================ Note 47.28 Can You Speed up an 11/70? 28 of 65 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 21 lines 19-FEB-1988 21:49 -< My evaluation unit was just beyond aprototype >- -------------------------------------------------------------------------------- > < Note 47.26 by EISNER::FRISBIE "Alan E. Frisbie" > >What is the price of this gem (in singles >and in quantity 10-20)? I don't know what the quantity discounts are; I assume the high cost right now is to recover R&D expenses. It is "plug-and-play". It may seem expensive considering that I imagine you can get an entire 11/70 for less than the board cost now, but as I've said, if you have the bucks and need a quick boost, it is quite painless. Note that this is a new product and there may be a few bugs that DDS are still working out, but the support appears to be very good. Call the company and you will probably get the chief engineer who will talk to you for literally hours if you are so inclined. As a new product, my biggest concern would be reliability; I'm not sure how long they have been running this and what the failure rates might be. I'd appreciate any feedback if you do any in-depth reserch. ================================================================================ Note 47.29 Can You Speed up an 11/70? 29 of 65 EISNER::TINKELMAN "Bob Tinkelman - CCA" 38 lines 20-FEB-1988 12:55 -< More about DDS/SETASI >- -------------------------------------------------------------------------------- I received the following from Barton Bruce, a fellow employee at Cambridge Computer Associates, Inc., after I showed him one of the PEP-70 notes.  I should preface my 'two-cents' with the statement that I have NO direct personal experience with DDS/SETASI products, but have gleaned some info that I hope is accurate, and will share. I have not seen all that has been on DECUServ in this area, but only saw a bit mailed to me by a fellow employee who offered to post this. The PEP-70 is by the same folks that do the hop-up job on 11/44s. It seems to be not one but actually two companies. DDS seems to run the ads but when an inside DEC person with, say, the stature of a product line manager speaks glowingly of 'them' the SETASI name seems to be used. They also have a Unibus repeater that keeps noise on one segment from being propagated to an adjacent segment, and that will not tolerate a device that is not driving the bus properly. It will force a marginal device to fail solidly and thus be identifiable for replacement. This goodie is showing up in F/S Unibus troublshooting kits, currently enjoys F/S Class-B support, and should soon be Class-A. The 11/44 hop-up consists of replacing memory with 4 one meg screaming fast card, and adding the repeater/isolator after the CPU backplane. The 11/44's own cache is removed as it slows everything down! Our site rep. has a guy in his unit that put in an 11/44 hop-up and loves it. All these pieces for the 4-meg 11/44 hop-up cost so much that for a trifle more you can have the fastest 11 of all - just buy the PEP-70 ($18K) and a used 11/70 (probably less that $2k complete with the MOS box for which you can easily find a 'taker' as you won't need it or its cabinet). Before now, the largest 11/70s I ever saw were 2 that had 3 meg, 1 that had 2 meg, 1 or 2 at 1 meg. Vast numbers of them only have around 1/2 meg. Memory was too expensive, and then was even worse to maintain. You may sell this to your management on speed alone, but the smallest PEP-70 is 2 meg and that's only is a tad less expensive than the full 4 meg unit. There may soon be many more 4 meg 11/70s that ever! RSX folks using M+ 4.0 can use the deferred caching for .tmp files, and may want to dig back and find such things as the FX driver or VS driver on an old SIG to help use all the memory creatively. ================================================================================ Note 47.30 Can You Speed up an 11/70? 30 of 65 EISNER::FRISBIE "Alan E. Frisbie" 21 lines 9-MAR-1988 12:59 -< More PEP-70 raves >- -------------------------------------------------------------------------------- Yesterday, I received a call from a friend, Dan Tannenbaum, who is evaluating the PEP-70 for the city of Los Angeles. He was nothing short of ecstatic about its performance. The MINIMUM improvement he saw was 15%. He stated that he would recommend the PEP-70 to anyone with an 11/70 that needs more performance. Just as an aside, his 11/70 is the Los Angeles Police Department dispatching system. Besides the dispatcher and 911 operator consoles, it supports the terminals in the police cars. In all, there are ~500 terminals on the system, plus comm links to NCIC, DMV and the phone company data base. Try THAT on your VAX! In fact, DEC proposed an 8700 for an upgrade proposal, but couldn't meet the same response times they are already getting on the 11/70. There is more to this, but I won't clutter up this note with it. Alan ================================================================================ Note 47.31 Can You Speed up an 11/70? 31 of 65 EISNER::JOE 27 lines 27-APR-1988 00:48 -< PEP-70, THE MIRACLE IS HERE! >- -------------------------------------------------------------------------------- We have finished our evaluation of the PEP-70, and it is nothing short of phenominal. Dan Tannenbaum has been doing the benchmarks, and we seem to have about a 30%-40% actual thru-put performance increase. How many of you have users who get upset at 30-second responses from the keyboard? The users on our system are the primary E911 answering positions that interface with the LAPD dispatching system. At peak load, they were experiencing up to 60 seconds and better delay at the keyboard. All this while some irate citizen is attempting to tell them the sky is falling. With the PEP-70, we experienced a raw command-processing increase of at least 20%-25%. Add to this the factor of NO NOTICABLE DELAY IN KEYBOARD RESPONSE at that higher level of thru-put, and an as yet unquantifiable percentage that consists of major system functions that were previously disabled to ensure the system would not crash. This one device has given the gift of renewed life to the LAPD dispatch system. We would like to extend an invitation to one and all to call and or visit us. Contact myself or Dan Tannenbaum at (213) 485-2146, or write to us at Information Services Division 200 N. Main St. MS-232 Los Angeles, Ca. 90012 Regards, KMJ ================================================================================ Note 47.32 Can You Speed up an 11/70? 32 of 65 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 19 lines 27-APR-1988 12:24 -< A tour worth taking >- -------------------------------------------------------------------------------- >> This one device has given the gift of renewed life to the LAPD dispatch >> system. We would like to extend an invitation to one and all to call and or >> visit us. I took advantage of the invitation to visit the LAPD Emergency Operations Center and was very impressed at what they are doing with a pair of PDP-11/70's. Besides handling all of the 911 operators, it also handles all of the dispatchers and the terminals in all of the police cars. ~500 terminals on an 11/70!!! The system is also connected to the NCIC data base in Washington, the California Department of Motor Vehicles, the telephone company and ghod knows what else. That is one busy CPU! I hope that they will give a session at the Fall '88 DECUS symposium. It should be a real chuckle to see how they made a PDP-11 run rings around DEC's proposed VAX solution. Alan ================================================================================ Note 47.33 Can You Speed up an 11/70? 33 of 65 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 7 lines 4-MAY-1988 15:30 -< Make you 11/70 as fast as a µV3000??? >- -------------------------------------------------------------------------------- Well now that we've finally agreed the PEP-70 is the greatest thing since sliced processing, I see someone called Quickware has a direct replacement for the 11-70 processor that supposedly runs 300% faster (article in May 2 1988 Digital Review page 20). Availability June or July. Price less than PEP-70. Any early reports out there on this? ================================================================================ Note 47.34 Can You Speed up an 11/70? 34 of 65 EISNER::MERRIMAN "George Merriman - CCA/NY" 1 line 4-MAY-1988 17:02 -< What does "300% faster" mean? >- -------------------------------------------------------------------------------- ================================================================================ Note 47.35 Can You Speed up an 11/70? 35 of 65 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 5 lines 5-MAY-1988 11:08 -< We shall see >- -------------------------------------------------------------------------------- > < Note 47.34 by EISNER::MERRIMAN "George Merriman - CCA/NY" > > -< What does "300% faster" mean? >- I hope to find out when I receive the literature I sent for yesterday. ================================================================================ Note 47.36 Can You Speed up an 11/70? 36 of 65 EISNER::DELARISCH "Arnold S. De Larisch" 7 lines 5-MAY-1988 13:50 -< New Cache Memory Boards for PDP 11/70 >- -------------------------------------------------------------------------------- I just read in the DEC Professional about some new Cache boards that replace those in the PDP-11/70. They are made by the SAME people who brought you PEP-70. Is this what you are talking about??? -Arnold ================================================================================ Note 47.37 Can You Speed up an 11/70? 37 of 65 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 12 lines 5-MAY-1988 14:55 -< Quickware .NE. PEP-70 >- -------------------------------------------------------------------------------- > < Note 47.36 by EISNER::DELARISCH "Arnold S. De Larisch" > > -< New Cache Memory Boards for PDP 11/70 >- >I just read in the DEC Professional about some new Cache boards that >replace those in the PDP-11/70. They are made by the SAME people who >brought you PEP-70. >Is this what you are talking about??? No. This Quickware is an entirley different outfit as far as I know. And I just happened to notice (in LOCAL_HAPPENINGS) that they are giving a presentation this month to MITLUG. Maybe we can get some comments from some attendees after the show. THE PEP-70 is a memory replacement, this Quickware thing sounds like a CPU replacement. ================================================================================ Note 47.38 Can You Speed up an 11/70? 38 of 65 EISNER::JOE 10 lines 11-MAY-1988 00:44 -< DOUBTING THOMAS >- -------------------------------------------------------------------------------- I spoke to the Quickware people today, and they have a three (3) board implementation of an 11/70 replacement. This package contains their own processor bundled with a faster access memory. They claim improvment is partialy due to not having to access a memory bus. I am skeptical about their performance ( 300% ). They seem to make you want to believe that their product can do all that an 11/70 can do. My concern is the compatibility of the instruction set. They claim it is identical, but only time will tell. ================================================================================ Note 47.39 Can You Speed up an 11/70? 39 of 65 EISNER::MAXWELL "Gary Maxwell" 15 lines 15-MAY-1988 02:18 -< Performance is not always performance >- -------------------------------------------------------------------------------- < I spoke to the Quickware people today, and they have a three (3) board < implementation of an 11/70 replacement. This package contains their own < processor bundled with a faster access memory. They claim improvment is < partialy due to not having to access a memory bus. This may be a case of "using selective performance metrics." Part of the performance of an 11/70 is the MASSBUS architecture, of having "high-speed" (at least back then) I/O channels tied directly into the memory/cache bus. Raw CPU power, which I suspect is what this group is referring to, is one piece of the total performance issue. If they don't have to access a "memory bus," then I am curious about how their I/O architecture is set up. ================================================================================ Note 47.40 Can You Speed up an 11/70? 40 of 65 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 10 lines 16-MAY-1988 14:02 -< Tom Provost, can you help? >- -------------------------------------------------------------------------------- >< Note 47.39 by EISNER::MAXWELL "Gary Maxwell" > > -< Performance is not always performance >- >< I spoke to the Quickware people today, and they have a three (3) board > If they don't have to access a "memory bus," then I am curious >> about how their I/O architecture is set up. I saw in LOCAL_HAPPENINGS that these guys are going to be giving a talk at a LUG in Boston. (Mitlug?). Maybe someone who attends cam shed some light. ================================================================================ Note 47.41 Can You Speed up an 11/70? 41 of 65 EISNER::ETHINGTON "Superior Klingon Technology" 45 lines 29-MAY-1988 00:47 -< Unbridled Speculation >- -------------------------------------------------------------------------------- > If they don't have to access a "memory bus," then I am curious > about how their I/O architecture is set up. With absolutely nothing whatsoever but unbridled speculation to back this up, I suspect they are talking about eliminating the actual memory bus, not the 11/70 CacheBus. The memory bus came off of the CacheBus, as did the MassBus. Since the memory bus had to be designed for, by current standards, absurdly slow nasty core memory (remember that? ick...), and had to have timings that could tolerate the nearly 20 foot cable runs necessary to daisy chain memory boxes together, I could easily beleive MAJOR improvements could be made here. In an attempt to speed up the pathetically slow memory access times of the day, the 11/70 memory controllers could be hooked up in series with interleaving between each controller; you could also get two way interleaving within each memory box. So in theory, if you daisy chained 4 memory boxes, and set each one for internal interleaving, and set each memory controller correctly for 4 way interleaving, you could actually get 8 way interleaving, pulling a full byte of memory from 8 different chips simultaneously in one memory cycle. The physical constraints of 4 separate memory boxes, the cabinets to hold them in, and the 11/74 multiprocessor project required ribbon cable runs of as much as 20 feet, and accordingly the bus cycles had to be tuned down to tolerate this. Brian McCarthy, a member of the 11/74 development team, has a funny story about how the 11/74 cabinets had to be positioned weird with a raised floor in between them to cover the memory bus ribbon cables - since they could not exceed 20 feet of cable, the cables weren't long enough to reach the floor level! These days, with 1 megabit high speed DRAMS, and needing only 32 chips to get the full 4 MB of memory that an 11 can eat, it is clearly silly to have a bus at all - just put the memory on the board right next to the CPU. I suspect this is a large part of what QuickWare might be doing, and if so I can certainly beleive there might be a huge performance boost. The MassBus and CacheBus would be intact, there just wouldn't necessarily be a nasty slow memory bus off one end of the CacheBus. I hope to hear more about this product - they could be on to a real winner here, in view of the rock bottom prices of used 11/70 CPUs these days. Jerry Ethington ================================================================================ Note 47.42 Can You Speed up an 11/70? 42 of 65 EISNER::KILLEEN "Jeff Killeen" 20 lines 29-MAY-1988 11:38 -< SOUNDS GOOD >- -------------------------------------------------------------------------------- Jerry's comments are 100 percent correct. The PEP-70 people gave a BOF at DECUS. What they did was create a 4MB cache and never have to go out to memory - therefore no need for memory or a memory bus. An 11/70 CPU only talks to the cache via the cache bus. The cache talks to the Memory, the massbus, and the unibus. They have gained performance by the following..... 1. It has all 4MB of memory right in cache. They claimed that since most programs today use large threaded code that the 11/70 cache has become less effective. The programs are jumping around so much that the cache hits have dropped because the 11/70 cache is so small. 2. The Mass buss passed data through cache to memory without really using it. Now the data that the mass buss loads is going right to high speed memory. Alan any other comments - you where there. ================================================================================ Note 47.43 Can You Speed up an 11/70? 43 of 65 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 23 lines 29-MAY-1988 13:26 -< One of the best sessions at Spring '88 DECUS >- -------------------------------------------------------------------------------- >> The PEP-70 people gave a BOF at DECUS. >> Alan any other comments - you where there. When I heard about the BOF, it immediately raised the red commercialism flag in my mind. To my surprise and delight, it was in the "old time DECUS" tradition: a pure technical session of the "this is how we did it" type. I hope that the engineer will repeat it as a regular session for the fall symposium. He spent most of the time explaining the engineering and economic trade-offs of the original 11/70 design (he was part of the team). By explaining the 1973(?) constraints versus today's technology/economics, he hardly had to tell us what he had done with the PEP-70; it was obvious. He is an expert on cache design (his PhD subject) and I wish that he would present a session at DECUS on it. After all that, I can't even remember his name. Does anyone know? Alan ================================================================================ Note 47.44 Can You Speed up an 11/70? 44 of 65 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 16 lines 30-MAY-1988 21:01 -< Quickware .NE. PEP-70? >- -------------------------------------------------------------------------------- >< Note 47.43 by EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" > >-< One of the best sessions at Spring '88 DECUS >- >To my surprise and delight, it was in the >"old time DECUS" tradition: a pure technical session of the >"this is how we did it" type. >I hope that the engineer will repeat it as a regular session for the >fall symposium. >After all that, I can't even remember his name. Does anyone know? When I was talking to them when this first came out I spoke on the phone often with a Bud Gallagher; I think the speaker was probably either him or another fellow named something like Jim Jones. On another note, I am now getting confused - I thought this "Quickware" device was something entirely different from the Setasi/Digital Data Systems PEP-70. Am I wrong? ================================================================================ Note 47.45 Can You Speed up an 11/70? 45 of 65 EISNER::KILLEEN "Jeff Killeen" 11 lines 30-MAY-1988 23:07 -< PEP70 <> QUICKWARE >- -------------------------------------------------------------------------------- > On another note, I am now getting confused - I thought this > "Quickware" device was something entirely different from the > Setasi/Digital Data Systems PEP-70. Am I wrong? Quickware *CLAIMS* they are developing there *OWN* 24MHZ plus CPU. Their plan is to replace the entire CPU/CACHE/MEMORY card set. PEP-70 only replaces the cache and memory. If Quickware pulls this off it is going to be a whole new ball game in PDP-11 land. P.S. The PDP-11 instruction set patents expire in October. ================================================================================ Note 47.46 Can You Speed up an 11/70? 46 of 65 EISNER::DELARISCH "Arnold S. De Larisch" 41 lines 31-MAY-1988 00:27 -< Dr. John Jones >- -------------------------------------------------------------------------------- >> < Note 47.43 by EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" > >> -< One of the best sessions at Spring '88 DECUS >- >> >> >> The PEP-70 people gave a BOF at DECUS. >> >> Alan any other comments - you where there. >> >> When I heard about the BOF, it immediately raised the red >> commercialism flag in my mind. To my surprise and delight, >> it was in the "old time DECUS" tradition: a pure technical >> session of the "this is how we did it" type. I hope that >> the engineer will repeat it as a regular session for the >> fall symposium. >> >> He spent most of the time explaining the engineering and >> economic trade-offs of the original 11/70 design (he was >> part of the team). By explaining the 1973(?) constraints >> versus today's technology/economics, he hardly had to tell >> us what he had done with the PEP-70; it was obvious. >> >> He is an expert on cache design (his PhD subject) and I >> wish that he would present a session at DECUS on it. >> >> After all that, I can't even remember his name. Does anyone >> know? Alan, his name is Dr. John Jones and I was fortunate enough to chair the BOF. I agree with Alan, the session was in the old DECUS tradition! The speaker was VERY knowlegable and very Tech! I had lunch with them after the BOF and convinced him to give two sessions at the Fall Symposium. One session will be on the Unibus and the other will be on the PDP 11/70. I would also like to note that Dr. Jones was also on the Design of the PDP 11/74 (he was one of the designers of the IIST device!) I was trying to get him to give a talk on the hardware aspects of multiprocessing al. la. PDP 11/74. Unfortunately, I don't think it will fly (I couldn't submit a CFP in time. I'll try to schedule it as a BOF if there is interest. -Arnold P.S. Be on the lookout in the Hardware Micro SIG for his sessions in the Fall! ================================================================================ Note 47.47 Can You Speed up an 11/70? 47 of 65 EISNER::KILGALLEN "Larry Kilgallen" 8 lines 1-JUN-1988 17:36 -< In what part of the PDP-11 Universe? >- -------------------------------------------------------------------------------- > Note 47.42 > EISNER::KILLEEN "Jeff Killeen" > > 1. It has all 4MB of memory right in cache. They claimed > that since most programs today use large threaded code > that the 11/70 cache has become less effective. The programs Most programs or most *Basic* programs (e.g., most *RSTS* programs)? ================================================================================ Note 47.48 Can You Speed up an 11/70? 48 of 65 EISNER::KILGALLEN "Larry Kilgallen" 5 lines 1-JUN-1988 17:40 -< What were the patent claims? >- -------------------------------------------------------------------------------- > Note 47.45 > EISNER::KILLEEN "Jeff Killeen" > P.S. The PDP-11 instruction set patents expire in October. Also, can't one get one extension after the first 17 years period? ================================================================================ Note 47.49 Can You Speed up an 11/70? 49 of 65 EISNER::MERRIMAN "George Merriman - CCA/NY" 8 lines 1-JUN-1988 18:38 -< I don't think extensions apply here >- -------------------------------------------------------------------------------- > Also, can't one get one extension after the first 17 years period? I think extensions are only given if the holder can show economic hardship -- for instance, drug companies can get extensions because the regulatory process delayed entry into market for their products. I think extensions can be granted for inventions that result in extraordinary benefit to mankind as well, but I doubt the PDP-11 instruction set qualifies for that. ================================================================================ Note 47.50 Can You Speed up an 11/70? 50 of 65 EISNER::CONROY 9 lines 7-JUL-1988 15:56 -< Am I missing something? >- -------------------------------------------------------------------------------- Am I missing something here? We looked into the PEP-70 products and were dismayed at the cost. To get both boards cost about $25,000. This week, I have seen 11/70s selling for about $1700. Now, even if the PEP-70 stuff could make your 11/70 run 12 times faster (!) it would still be more cost effective to buy more used '70s. Of course, I realize that if you want a dedicated application to speed up, then the multi-11/70 point is moot. I could also see (if I try really hard) being able to spread the cost over several years, but even having to buy FPUs ($300) memory, etc it would still take a LONG time. ================================================================================ Note 47.51 Can You Speed up an 11/70? 51 of 65 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 8 lines 7-JUL-1988 16:35 -< COST OF OWNERSHIP >- -------------------------------------------------------------------------------- > -< Am I missing something? >- Yup - If you buy a PEP-70 you get to trash the MK11 memory box. The cost of maintaining a 4MB MK11 under contract equals the cost of the PEP-70 in about 14-18 months. Now you could go around buying MK11's and maintain them yourself - but then you have to put up with the reliability issue. ================================================================================ Note 47.52 Can You Speed up an 11/70? 52 of 65 EISNER::KENNEDY "Terry Kennedy" 17 lines 7-JUL-1988 16:37 -< It depends on your goals >- -------------------------------------------------------------------------------- > Am I missing something here? Well, as a prospective new '70 owner myself, I see a few reasons. Whether they are *good* reasons or not probably depends on your intended use of the system. o Despite DEC's claims that a faster single-instance 11 is not being demanded by customers, *we* need something faster than a stock '70, and multiprocessor solutions won't help. o The cost of the upgrades is well spent as it removes the single worst aspect of the system - the memory controller and memory. o For those of us loons to whom 'Massbus is a Massbus is a Massbus' who are going to try RP07's and TU78's on said '70, the cost of those items outweighs the upgrade. Besides, you can get the upgrade on evaluation. ================================================================================ Note 47.53 Can You Speed up an 11/70? 53 of 65 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 3 lines 11-JUL-1988 11:18 -< A *SINGLE* faster 70 was the goal >- -------------------------------------------------------------------------------- Yes, as Terry mentioned in *-1 and I mentioned in .0, this is real nice where you can not spread an application over multiple cpu's. ================================================================================ Note 47.54 Can You Speed up an 11/70? 54 of 65 EISNER::CONROY "Alan Conroy" 8 lines 18-JUL-1988 14:19 -< Hmmm >- -------------------------------------------------------------------------------- I realize the advantage in a situation where multiple CPUs don't work. We don't have that situation here, and several of the other replies seemed to indicate the same. I suppose I can also see a long term benefit. However, we hope to get rid of all of our remaining PDPs sometime in the next year or two and the cost of the PEP was too high for me to justify it to management. I guess I was just disappointed by the high cost. I don't believe 4Mb costs that much on a VAX, for instance (granted, it's faster than VAX memory, but still...) ================================================================================ Note 47.55 Can You Speed up an 11/70? 55 of 65 EISNER::KENNEDY "Terry Kennedy" 10 lines 18-JUL-1988 15:38 -< Apples and oranges... >- -------------------------------------------------------------------------------- > I don't believe 4Mb costs > that much on a VAX, for instance (granted, it's faster than VAX memory, but > still...) Actually, VAX memory costs far more, if you consider the cost of the mem- ory controller as well. On many VAXen, 'memory' boards are just that - boards full of memory chips. *just* memory chips. In contrast, 11-type memories have the controller on the card (for ECC or parity, bus interface, etc.) It is an apples and oranges situation. Also, the rising cost of RAM chips isn't help- ing. ================================================================================ Note 47.56 Can You Speed up an 11/70? 56 of 65 EISNER::CONROY "Alan Conroy" 3 lines 19-JUL-1988 12:48 -< I bow to your greater wisdom >- -------------------------------------------------------------------------------- OK, OK. Maybe it is a good deal, but I know I'll never convince management to spend $25K on our existing 11/70s when we could buy five more equivalent machines for that price. * sigh * ================================================================================ Note 47.57 Can You Speed up an 11/70? 57 of 65 EISNER::CONROY "Alan Conroy" 3 lines 31-AUG-1988 10:26 -< Quickware? >- -------------------------------------------------------------------------------- Note 198.17 makes reference to an 11/70 CPU replacement by a company called Quickware. Anyone know anything about this? Is it for real, or have I been duped? ================================================================================ Note 47.58 Can You Speed up an 11/70? 58 of 65 EISNER::SIMONS "Paul, Not that CONVEX!" 5 lines 1-SEP-1988 11:40 -< Flying in a holding pattern >- -------------------------------------------------------------------------------- Quickware Engineering's "PDP 11/94" got stomped on by DEC. Something to do with "patent infringement". However, both Setasi and Quickware are waiting until Halloween (When DEC's patent expires (see earlier note)) to wheel out some *new* processors. Rumor has it the 11/74 is alive and well and living in Florida. ================================================================================ Note 47.59 Can You Speed up an 11/70? 59 of 65 EISNER::CONROY "Alan Conroy" 4 lines 1-SEP-1988 12:30 -< I can hardly wait! >- -------------------------------------------------------------------------------- > However, both Setasi and Quickware > are waiting until Halloween (When DEC's patent expires Looks like Christmas is coming early this year! ================================================================================ Note 47.60 Can You Speed up an 11/70? 60 of 65 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 7 lines 1-SEP-1988 16:20 -< hot 70's in mere weeks >- -------------------------------------------------------------------------------- QuickWare's QED-94 may have been what got stompped (if anything), but they say their QED-95 will be in test this month, and in production in Nov. This thing may RAISE the price of empty used 11/70 CPU boxes if the demand outstrips what hasn't been fed to the scrap crushers. ================================================================================ Note 47.61 Can You Speed up an 11/70? 61 of 65 EISNER::SIMONS "Paul, Not that CONVEX!" 3 lines 2-NOV-1988 15:05 -< Any word? >- -------------------------------------------------------------------------------- I was unable to attend the Anahiem symposia. Can someone talk about Jim Jones presentation? Any word about new processors? Enquiring minds need to know! ================================================================================ Note 47.62 Can You Speed up an 11/70? 62 of 65 EISNER::TINKELMAN "Bob Tinkelman - CCA" 14 lines 2-NOV-1988 17:00 -< In the sessions, no, but in the halls... >- -------------------------------------------------------------------------------- > I was unable to attend the Anahiem symposia. Can someone talk about > Jim Jones presentation? Any word about new processors? Enquiring > minds need to know! I went to the morning presentation - something like "all about the Unibus". He bent over backwards to make it 100% noncommercial. He didn't even mention that he was assocaited with a company that might have a product. I suspect his afternoon session was also "strictly by the book, noncommercial". However, since he was there I'm sure that there were many attendees who talked to him about "new processors". I happen to know that at least one of them logs into DECUServe quite regularly. Maybe he'll pass on his knowledge. [OK Barton - your turn.] ================================================================================ Note 47.63 Can You Speed up an 11/70? 63 of 65 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 84 lines 2-NOV-1988 18:46 -< Hints of 11 futures >- -------------------------------------------------------------------------------- >>>< Note 47.61 by EISNER::SIMONS "Paul, Not that CONVEX!" > >>> -< Any word? >- >>> Can someone talk about Jim Jones presentation? PresentationS - he did 2 - 2 hour ones the same day! All right out of his head and on a blackboard. The first was basically about unibuses. The second was about 11/70s. >>> Any word about new processors? No, and yes. He was being totally careful about any commercialism, and simply talked theory and what could be done. He did get to a part where the 'what could be done' was his PEP-70, and almost said it, and caught himself. It was simply used as an example, and the chair said he could use the name, so he did. It simply tied his discussion to the products people have been watching. Off line, a bit more seemed to be available. What he has running in the lab sounds as though we are in for super things in the future, but it did not sound as though it would be next week. Can you say 'times 25'! Other folks, whose HOT product keeps slipping another month, seem to be claiming vast speed advantages over his product. He says this will NOT be so, as it is simply an issue of specs-manship, and he does not believe they will be faster. As to their price, his answer was that he has product NOW, they have paper. It will be interesting to see if and when Quickware delivers what they promise, what John Jones's Setasi does to lowers their price. Both companies seemed to be keeping mum on one front! But a DEC s/w type readily KNEW the staff compliment at Quickware, and exactly how many were coop students. Indications from DEC and the other companies seem to say they have been talking but won't talk to us about it. In another session, Dec fielded an older white-haired gentleman who introduced all those s/w + h/w that worked for him. ALL the 11 related upper staff we normally hear of at Decus work for him! He was saying that DEC has reevaluated its position on 11 products, reshuffled staff, and wants to know what users want! The sad part is that maybe 1/2 that hall was filled with HIS staff that had just introduced themselves, and the rest were the few die-hard users you would expect to be there. There were NOT hordes of people to greet this man and respond to his message. If DEC truely now wants to support users who NEED faster 11s, they could be planning to make one themselves, possibly with the 40mhz (if well cooled, 30mhz otherwise) CMOS Harris chip, or they may license these other companies for any relevant DEC patents that have not expired, and in the process CONTROL how they distribute their cards to US. It was even suggested that DEC might put these brand X cards in a DEC box to sell as a system. This might be good, but if we can't then also get just the cards at a fair price (or at all - think quad 11/73-83 cards...) this is terrible. I think DEC is seriously talking to the other vendors, and there may be some good to come out of it, if DEC has listened to the cries of the users. I doubt anything has been decided yet, and now is the time to make your feelings known anywhere you can. Dec DID come to listen, only few spoke. But back onto John Jones's sessions. He is a very good speaker, and clearly KNOWS the subject VERY well. I was totally impressed. When he is saying that others specs don't mean they are faster than he is, I believe him. He is convinced the unibus is a VERY potent bus, and has years of good life (that he can exploit) left. From their Business cards (I have NO financial connections): Setasi - John Jones (pres) (305) 963-1267 Quickware - Alan Sieving (vp) (617) 647-3800 or (800) 237-1185 ================================================================================ Note 47.64 Can You Speed up an 11/70? 64 of 65 EISNER::BYRNE_C "Charlie Byrne" 3 lines 5-APR-1989 10:12 -< Status of Quickware? >- -------------------------------------------------------------------------------- Has anyone heard anything about Quickware lately? Anyone have their product(s)? Are they alive? ================================================================================ Note 47.65 Can You Speed up an 11/70? 65 of 65 EISNER::SIMONS "Paul, Not that CONVEX!" 6 lines 6-APR-1989 17:28 -< Well, sort of ... >- -------------------------------------------------------------------------------- Quickware is alive and well, but not on the 11/70 front. We are actively involved with both Setasi and Quickware. We have been tentatively approved as a beta test site for anything Quickware comes out with. We are using Setasi's PEP-70 and Hyper-Cache products. When we told Setasi we wanted to wait for their new 11/70, they said it would be a while (Months), they were negotiating with DEC.  ================================================================================ Note 48.0 FCO kit for TA/TU78 5 replies EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sa" 14 lines 14-SEP-1987 06:57 -------------------------------------------------------------------------------- > CAN A TA/TU78 SUBSYSTEM BE UPGRADED INTO A TA/TU79? > > No. The TA/TU79 incorporates significant mechanical and cabinetry > changes from the TA/TU78. However, the electronic changes incorporated > in the TA/TU79 are field upgradeable into the TA/TU78 via an FCO kit. > Contact your Field Service representative to check availability of the > FCO kit. I guess this means Field Service might make some changes to my TA78 if I tell them about this. Any idea what the FCO kit number is and what specifically the "electronic changes" would be ? Also is this FCO "free" under my several thousand dollar monthly maintenance plan with Digital ? ================================================================================ Note 48.1 FCO kit for TA/TU78 1 of 5 EISNER::NORTON 4 lines 14-SEP-1987 14:28 -------------------------------------------------------------------------------- My FS guy says the changes are "free", but the kits are being distributed on an allocation basis to the branches. Changes supposed to include new head and read/write boards. ================================================================================ Note 48.2 FCO kit for TA/TU78 2 of 5 EISNER::KENNEDY "Terry Kennedy" 6 lines 14-SEP-1987 19:48 -< Two FCO's - not the same >- -------------------------------------------------------------------------------- Yes, the kit is free (or will be, when it becomes available). The head- change and card changes are NOT the 79 upgrade ECO. The cards changed are the formatter CRC card (gets a shield on the back) and the formatter microprocessor card (gets new ROMS). The 79 upgrade ECO involves changing boards in the top part of the drive, not the formatter. Some ROM changes in the formatter are also included, but that is not the major part. ================================================================================ Note 48.3 FCO kit for TA/TU78 3 of 5 EISNER::TINKELMAN "Bob Tinkelman - CCA" 21 lines 16-SEP-1987 00:02 -< TU78 FCOs 1399 and 1400; and TU77s >- -------------------------------------------------------------------------------- > I guess this means Field Service might make some changes to my TA78 if I > tell them about this. Any idea what the FCO kit number is and what > specifically the "electronic changes" would be ? Some time ago, Barton Bruce (not yet invited to DECUServe) sent me this mail related to some TU77s at one of our client sites. From: CCAVAX::BRUCE "Barton" 8-MAY-1987 16:30 To: TINKELMAN,WOLFSON,BRUCE Subj: TU77 FCOs The TU78 drives have various FCOs available to improve operation. Some are to the mechanics of the drive, and some are to the read/write electronics. The mechanical ones improve tape loading (much needed) and apply equally well to the TU77 drives which are built around the same transport. The TU78 read/ write ECOs do not apply to TU77s. Field service will probably NOT have bothered to apply these to TU77s in the field, but certainly can and should be asked to. The FCOs are listed being TU78 ones, and are numbers 1399, and 1400. I think they had been mentioned at DECUS by Bruce Gordon, the product line manager. ================================================================================ Note 48.4 FCO kit for TA/TU78 4 of 5 EISNER::BUTLER 8 lines 18-NOV-1987 15:27 -< TA/TU78 FCO experience wanted >- -------------------------------------------------------------------------------- My local FS guy has procured the TA/TU FCOs for my 4 drives. However, as he looked into it, he was told that the upgrade FCOs are really just for "problem" drives. As all of our drives seem to be operating "normally" (they were installed in Jan 87), we are wondering if we should leave well enough alone, and "don't fix what ain't broke". Anybody out there have any experience with the FCOs in question? Negative or positive feedback of any kind would be appreciated. ================================================================================ Note 48.5 FCO kit for TA/TU78 5 of 5 EISNER::KENNEDY "Terry Kennedy" 12 lines 18-NOV-1987 21:00 -< FCO's well worth it >- -------------------------------------------------------------------------------- > Anybody out there have any experience with the FCOs in question? I had one of those 'problem' drives - it had 165 hours on the meter, of which 145 were from Field Service! If your drives are that new, you probably have the FCO's. To reiterate, look at the air opening on the lower lip of the housing where you plop the tape in. If there is a rubber gasket (usually brown) there, you have most, if not all, of the FCO's you need. My drive was a 1985 vintage unit (but we bought it with only 20 hours on it last July - I guess the previous owner wasn't too happy with it either). After the FCO's, the unit is just wonderful!  ================================================================================ Note 49.0 microVAX-II to microVAX-3xxx CPU upgrade 2 replies EISNER::FRISBIE "Alan E. Frisbie" 20 lines 14-SEP-1987 13:28 -------------------------------------------------------------------------------- After the microVAX 3xxx-series announcement at European DECUS last week, I had a discussion with some of the DEC people about the CPU board. They know that some OEMs and end-users would like to upgrade existing microVAX-II systems to microVAX-III CPUs. The problems that DEC has include production capacity and the desire to not eat into the market for new systems. The proposed solution is to make an upgrade available, but at a relatively high price. Then, as production ramps up and demand stabilizes, reduce the price. They have already announced this policy with the new TK52-K cartridge for the TK70, and it was well received. Nothing will be officially decided for some time, so don't hold your breath. The real-time people within DEC are bothered, however, by the way that users are leaving DEC products, and want to stop it. This is one way to do it. Alan ================================================================================ Note 49.1 microVAX-II to microVAX-3xxx CPU upgrade 1 of 2 EISNER::KILLEEN "Jeff Killeen" 7 lines 14-SEP-1987 17:26 -< MAYBE FOR DOD >- -------------------------------------------------------------------------------- >hold your breath. The real-time people within DEC are >bothered, however, by the way that users are leaving DEC >products, and want to stop it. This is one way to do it. The MV3xxx hardware may be capable of doing the job but I will bet you the price DEC will charge will drive them away. ================================================================================ Note 49.2 microVAX-II to microVAX-3xxx CPU upgrade 2 of 2 EISNER::FRISBIE "Alan E. Frisbie" 9 lines 14-SEP-1987 18:34 -< CPU upgrade = not cheap >- -------------------------------------------------------------------------------- >> The MV3xxx hardware may be capable of doing the job but I will >> bet you the price DEC will charge will drive them away. I agree, but they know that some people will be willing to pay the price for three times the CPU speed. "Whatever the market will bear." Alan  ================================================================================ Note 50.0 DMZ-32 Bit Droppings 2 replies EISNER::MERUSI 17 lines 15-SEP-1987 08:41 -------------------------------------------------------------------------------- We are having a problem using the DMZ-32. We here at Pratt & Whitney Aircraft are using the DMZ-32 for a pseudo-real-time data acquisition system. We are inputting binary hex data at 9600 baud, therefore, we do not use XON/XOFF. We have had some DEC representatives in house to determine the nature of the situation, and they discovered one thing darn perculiar, and that is, if you set all of the terminal characteristics we require "too quickly" they evidentally do not all go into effect properly! After breaking up a certain DCL command procedure it was realized that the data was being acquired properly. The symptom that we experienced had was a random number of bytes being dropped from the input stream. As it turns out, the terminal characteristics setting situation has not completely addressed our circumstance, we are still seeing (although less frequently) bytes being dropped from the input stream. Anybody out there have any similar experience with the DMZ-32? Supposedly, the DMZ-32 has been updated to be the latest rev ... Thanx. ================================================================================ Note 50.1 DMZ-32 Bit Droppings 1 of 2 EISNER::ROCHELEAU "Rod Rocheleau" 9 lines 8-OCT-1987 17:02 -< Could your cables be a problem >- -------------------------------------------------------------------------------- < Note 50.0 by EISNER::MERUSI > -< DMZ-32 Bit Droppings >- One question ... how long are the cables from the DMZ's and are they the DEC or T1 spec cables. I have about 30 DMZ's and have had a few interesting problems but they all had to do with cable or grounding problems. ================================================================================ Note 50.2 DMZ-32 Bit Droppings 2 of 2 EISNER::RENES 9 lines 25-OCT-1987 17:41 -< Rev. F1 of M8398 is the fix >- -------------------------------------------------------------------------------- Your problem is one of seven that are experienced with out of rev. DMZ-32 boards: Ask your DEC FE to install EQ-01457-01 FCO which changes the M8398 modules, Rev. E1 or earlier to Rev. F1. We had experienced loss of characters for 2 years now, until this fix.  ================================================================================ Note 51.0 RX33 for BA23 or BA123 boxes? 12 replies EISNER::FRISBIE "Alan E. Frisbie" 8 lines 15-SEP-1987 23:06 -------------------------------------------------------------------------------- Does anybody know if there is an RX33 drive available for existing BA23 or BA123 boxes? Since the RX50 is "obsolete" now, it would seem that the RX33 should be available, but it doesn't seem to be. I would even accept a 3rd-party solution, as long as it is a "clean" installation (that is, skidplate and faceplate included). Alan ================================================================================ Note 51.1 RX33 for BA23 or BA123 boxes? 1 of 12 EISNER::KILLEEN "Jeff Killeen" 16 lines 16-SEP-1987 09:01 -< RX33 ARE AROUND >- -------------------------------------------------------------------------------- > Does anybody know if there is an RX33 drive available for > existing BA23 or BA123 boxes? Since the RX50 is "obsolete" We have installed RX33 eqivs from Trimarchi (814) 234-5659 in both BA23 and BA123 boxes. As long as you get the dual drive config it is clean. The single drive config does not have the nice face plate. The cost is well under $200 per drive. Make sure you get the stacking hardware, the "Y" power connector, and the right firmware for the RQDX3 (the early RQDX3's do not have the right firmware). DEC and CDC have put them under Field Service with no questions. DEC lists the RX33A-AA (first drive) and RX33A-AB (second drive) in the price book. Even through they only list it for the BA23 they should work in the BA123 also. If you want to pay 4.5 times as much for the extract same drive. ================================================================================ Note 51.2 RX33 for BA23 or BA123 boxes? 2 of 12 EISNER::FRISBIE "Alan E. Frisbie" 13 lines 27-OCT-1987 02:53 -< Trimarchi "RX33"s almost running >- -------------------------------------------------------------------------------- Now that I have installed my RX33 (equivalent) drives in my VAXstation, how do I format blank floppies? This must be documented some place, but my sleep-deprived brain can't seem to find it. Also, I noticed a curious jumper on the M9058 signal Distribution Board. Next to the "RX" connector are three posts with a a single jumper between the center one and the one labeled "RX50". Should the jumper be moved to the unlabeled position for RX33 drives, or is there some other meaning for this jumper? Alan ================================================================================ Note 51.3 RX33 for BA23 or BA123 boxes? 3 of 12 EISNER::RICE "Gary Rice" 10 lines 27-OCT-1987 10:50 -------------------------------------------------------------------------------- > Now that I have installed my RX33 (equivalent) drives in > my VAXstation, how do I format blank floppies? This must... According to the people at DILOG (who talked us into THEIR RQDX3 controller), the RQDX3 controller CANNOT format floppies. Does your RX33 equivalent use the RQDX3 or something else? Gary ================================================================================ Note 51.4 RX33 for BA23 or BA123 boxes? 4 of 12 EISNER::COLE 13 lines 27-OCT-1987 12:34 -< /DENSITY=DOUBLE qualifier for RX33 format >- -------------------------------------------------------------------------------- >> Now that I have installed my RX33 (equivalent) drives in >> my VAXstation, how do I format blank floppies? This must... > According to the people at DILOG (who talked us into THEIR > RQDX3 controller), the RQDX3 controller CANNOT format floppies. There is a qualifier on the INIT command used to format RX33 floppies, /DENSITY=DOUBLE. I don't know about the controller capability, but it works fine on uVAX 2000. Steve Davis. ================================================================================ Note 51.5 RX33 for BA23 or BA123 boxes? 5 of 12 EISNER::FRISBIE "Alan E. Frisbie" 28 lines 27-OCT-1987 14:56 -< I must still be missing something >- -------------------------------------------------------------------------------- >> According to the people at DILOG (who talked us into THEIR >> RQDX3 controller), the RQDX3 controller CANNOT format floppies. >> Does your RX33 equivalent use the RQDX3 or something else? I am using the RQDX3. The ROMs are labled: DEC86 DEC86 LM8709 LM8705 244E5 243E5 Since DEC does not supply RX33-formatted floppies (only RX50), I would expect that there must be some way to format them. This belief is reinforced by the fact that RSX-11M-Plus (Update C) has this capability. >> There is a qualifier on the INIT command used to format >> RX33 floppies, /DENSITY=DOUBLE. I don't know about the >> controller capability, but it works fine on uVAX 2000. When I try that, I get the message, %INIT-F-FORMAT, invalid media format A SHOW DEVICE command identifies the drives as RX33s and I am able to read RX50 diskettes. I haven't tried writing to an RX50 yet. Alan ================================================================================ Note 51.6 RX33 for BA23 or BA123 boxes? 6 of 12 EISNER::KENNEDY "Terry Kennedy" 11 lines 27-OCT-1987 19:59 -< ZRQF??.BI? for PDP-11's >- -------------------------------------------------------------------------------- > According to the people at DILOG (who talked us into THEIR > RQDX3 controller), the RQDX3 controller CANNOT format floppies. They are partially in error. The RQDX3 *can* format anything. It won't (with factory firmware) format RX50's, therefore preserving DEC's investment in this obsolete medium. It can, however, format RX33's if you have the latest firmware (which you do). For PDP-11's, the formatter is ZRQF??.BI?. I don't know what operating systems support direct formatting at the command line level - I would be surprised if any do. RX02 'reformatting' is simply a matter of re- writing the data, not actually formatting anything. ================================================================================ Note 51.7 RX33 for BA23 or BA123 boxes? 7 of 12 EISNER::KILLEEN "Jeff Killeen" 12 lines 28-OCT-1987 00:17 -< WHAT YOU WANTED TO BE ABLE TO USE THE DRIVE! >- -------------------------------------------------------------------------------- > Should the jumper be moved to the > unlabeled position for RX33 drives, or is there some other > meaning for this jumper? The jumpers are the same for the RX33 and RX50. They are also the same jumper settings whether install one or two RX33's. Dec does *NOT* sell formated RX33's. In PDP-11 land you must format them using the MICRO-11 Formatter disk (BL-FN7AD-MC). That disk has the ZRQF?? formatter on it. You are going to have to get the VAX equivalent disk. ================================================================================ Note 51.8 RX33 for BA23 or BA123 boxes? 8 of 12 EISNER::FRISBIE "Alan E. Frisbie" 16 lines 28-OCT-1987 02:08 -< RSX has it now... >- -------------------------------------------------------------------------------- >> Dec does *NOT* sell formated RX33's. In PDP-11 land you must format them >> using the MICRO-11 Formatter disk (BL-FN7AD-MC). That disk has the >> ZRQF?? formatter on it. You are going to have to get the VAX equivalent >> disk. Not ALL PDP-11 operating systems are brain-damaged. :-) RSX-11M-Plus v3.0 (Update C and above) allows you to format an RX33 with the following command: FMT DU2:/DENSITY=HIGH Buying the $2000 diagnostic software simply to format RX33 diskettes on the VAX seems like a bit much. There must be a better way. Alan ================================================================================ Note 51.9 RX33 for BA23 or BA123 boxes? 9 of 12 EISNER::KILLEEN "Jeff Killeen" 7 lines 28-OCT-1987 05:40 -< PDP-11 ARE ALWAYS BETTER AT SOME JOB'S ;-} >- -------------------------------------------------------------------------------- > Buying the $2000 diagnostic software simply to format RX33 > diskettes on the VAX seems like a bit much. There must > be a better way. Sound like the better way for the short term is to plug a PDP-11/%3 board set into BA123 and format a bunch of disks. ================================================================================ Note 51.10 RX33 for BA23 or BA123 boxes? 10 of 12 EISNER::COLE 19 lines 29-OCT-1987 20:32 -< When at first... >- -------------------------------------------------------------------------------- >> There is a qualifier on the INIT command used to format >> RX33 floppies, /DENSITY=DOUBLE. I don't know about the >> controller capability, but it works fine on uVAX 2000. > When I try that, I get the message, > %INIT-F-FORMAT, invalid media format We have used /DENS=DOUBLE on our uVAX 2000 with IBM AT floppies. There is also a off-line diagnostic that will do it on the 2000. Are you using a brand new floppy? Your welcome to come use our machine to format some if you like, we're in Costa Mesa. If I recall correctly, you and I met sometime around 1982 in Pasadena. I dropped by your place to pick up old disk packs to be used in juggling. Steve Davis ================================================================================ Note 51.11 RX33 for BA23 or BA123 boxes? 11 of 12 EISNER::FRISBIE "Alan E. Frisbie" 19 lines 29-OCT-1987 21:38 -< Disks, flying and otherwise >- -------------------------------------------------------------------------------- >> Are you using a brand new floppy? Yes. It is totally unformatted. Can the uVAX 2000 format totally blank floppies with the INIT command? If so, I wonder what the problem is with the uVAX-II. >> You're welcome to come use our machine to format some if >> you like, we're in Costa Mesa. Thanks for the offer, but there is a closer site I can use. >> I dropped by your place to pick up old disk packs to be >> used in juggling. Did you ever manage to juggle them? I would love to see it at a symposium. Alan ================================================================================ Note 51.12 RX33 for BA23 or BA123 boxes? 12 of 12 EISNER::FRISBIE "Alan E. Frisbie" 9 lines 7-NOV-1987 01:08 -< DSIN has a solution(?) >- -------------------------------------------------------------------------------- >> I just checked to see if DSIN had anything on formatting RX33 floppies. The only note on the subject suggested using the RX33 on the HSC70! I'm sure that everyone with RX33's on a MicroVAX-II is going to rush right out and buy an HSC70. Alan  ================================================================================ Note 52.0 ?? Hot Sites ?? 3 replies EISNER::WALLIS "Barry Wallis - Fleetwood Enterprises" 7 lines 17-SEP-1987 17:55 -------------------------------------------------------------------------------- We are looking for hot-sites for our computing services in case of a catastrophe. We are open to having the hot-site anywhere. It needs to at least have an 8650 (and a cluster of them would be preferred). Any ideas or suggestions? Be gentle with me I'm new at this. ================================================================================ Note 52.1 ?? Hot Sites ?? 1 of 3 EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sap" 7 lines 17-SEP-1987 23:34 -< Digital Has It Now >- -------------------------------------------------------------------------------- > We are looking for hot-sites for our computing services in case > of a catastrophe. Good old Digital Equipment Corporation has a fairly well thought out disaster recovery program. They do have "hot sites" like you want. It seemed good but was a little expensive. ================================================================================ Note 52.2 ?? Hot Sites ?? 2 of 3 EISNER::PINSLEY "Howard W. Pinsley" 10 lines 18-SEP-1987 00:56 -< Funny you should bring this up today >- -------------------------------------------------------------------------------- > Good old Digital Equipment Corporation has a fairly well thought > out disaster recovery program. They do have "hot sites" like you > want. It seemed good but was a little expensive. This might seem hard to believe but as I read your hot site suggestion (DEC) I had just returned from their NJ hot site after spending the day dumping my entite system in preparation for a "disaster" weekend. I won't go into the details of why we are declaring the disaster but the bottom line is that we are using DEC's service for real for the first time today. It does indeed exist. ================================================================================ Note 52.3 ?? Hot Sites ?? 3 of 3 EISNER::WALLIS "Barry Wallis - Fleetwood Enterprises" 6 lines 25-SEP-1987 14:40 -< Any other hot sites? >- -------------------------------------------------------------------------------- Does anyone know of any other hot sites utilizing DEC equipment? I believe ther is one in Colorado called Business Recovery Systems (BRS). I'm not specifically looking for recommendations, just a place to begin searching.  ================================================================================ Note 53.0 MUXserver 100 & DECMux II 2 replies EISNER::MACNEIL "Laird of the Highlands" 3 lines 18-SEP-1987 23:22 -------------------------------------------------------------------------------- Has anybody used the MUXserver 100/DECMux II hardware? And if so, what problems have you encountered? Real hot for any information about using these products and LAT and DECNET with them. ================================================================================ Note 53.1 MUXserver 100 & DECMux II 1 of 2 EISNER::NORTON 16 lines 21-SEP-1987 09:56 -------------------------------------------------------------------------------- >Has anybody used the MUXserver 100/DECMux II hardware? We are using one of these. It has 16 ports, one of which is a spooled LN03. The users seem to be happy with server access to the network, as well as transmission speed over the 9600 baud line. It has some neat features, like being able to connect thru the muxserver to the decmux II's to see counters and set port parameters. >And if so, what problems have you encountered? About the only weirdness I've noticed is that when any of the three boxes involved (muxserver & 2 mux II's) needs to be restarted, they all need it, or some parts/ports don't come up again. ================================================================================ Note 53.2 MUXserver 100 & DECMux II 2 of 2 EISNER::MCCORMICK_B "Bob McCormick" 29 lines 28-SEP-1988 22:11 -< Not happy with them! >- -------------------------------------------------------------------------------- We've got two sites that are using these things; each site has three remote sites. Five of them are MS100 - 9.6kbps -Z- 9.6kbps - DMII One is MS100 - special cable local - DMII One of the modem ones above recently added a second DMII box. With all devices at 4800 baud; typically one printer and 5 or 6 VT220's or VT320's we find that the users at the remote site are NOT happy with the performance. We find better performance with CODEX 6003 4 or 8 port muxs with 9.6kbps modems. We have had a rash of problems - field service is glad to come out and swap the boxes, but then you must reprogramme the DECmux II or MUXserver 100 boxes. Our sites can't do this; they are sorta turnkey users. We had significant problems when we added the second DECmux II box at one remote site that needed to add more than 8 peripherals. These were resolved by replacing ROMS (and FCO or ECO) in equipment in both ends of the link. We are not happy with them at all. We don't configure them any longer on our systems. The alternative is the Codex gear (works right out of the box; no programming) which is reliable and has good support. We plug it directly into DEC423 ports (using BC16E-XX cables and H8571-D adaptors on the mux) on DS200 or DS500's. Glad to share more info off line.  ================================================================================ Note 54.0 SA482? 4 replies EISNER::PRIGOT "Jonathan M. Prigot" 7 lines 21-SEP-1987 09:05 -------------------------------------------------------------------------------- I have a project coming up where I need to add new disks. I can either go with multiple RA81s or an SA482 array. Since they have come out, I have never seen any comments on the SA482/RA82 drives. Anyone have any comments/war stories that they can share? (Moderator: You have my permission to move this to SOAPBOX should the need arise.) ================================================================================ Note 54.1 SA482? 1 of 4 EISNER::DROPPERS "Seton R. Droppers" 8 lines 21-SEP-1987 12:20 -< SA482? No Problems Yet... >- -------------------------------------------------------------------------------- They came up two months ago, ran from the start, we haven't heard a thing from them since, except lots and LOTS of reads and writes, (the intended kind). Bottom Line: We put them on line and have had no problems of any sort. Just one note: Because of the large number of blocks (1,216,665) DEC does not support them with a cluster size of 1, only 2 or greater (see VAX/VMS Dispatch from sometime last spring, as I remember) ================================================================================ Note 54.2 SA482? 2 of 4 EISNER::BORCHARDT "RODNEY P. BORCHARDT" 3 lines 21-SEP-1987 15:28 -< SA482 >- -------------------------------------------------------------------------------- ~r I HAVE HAD ALL FOUR VOLUMES BOUND TOGETHE{ AS ONE LOGICAL VOLUME FOR ABOUT FOUR MONTHS. SO FAR NO PROBLEMS. ================================================================================ Note 54.3 SA482? 3 of 4 EISNER::HAMPARIAN "Don Hamparian - Battelle" 4 lines 23-SEP-1987 10:34 -< 82s are working fine >- -------------------------------------------------------------------------------- We have had 2 SA482s since Nov 1986 and have not had any problems. We also have about 37 RA81s also and have had *many* problems with them. The 82s seem to be much more reliable then the 81s. Not much faster though. ================================================================================ Note 54.4 SA482? 4 of 4 EISNER::KALLICK 2 lines 24-OCT-1988 16:50 -< a few problems >- -------------------------------------------------------------------------------- We added 4 RA82s in June 88. Within 2 months, we had replaced two of the HDAs. No problems since.  ================================================================================ Note 55.0 Problem with borrowed RA81 3 replies EISNER::NORTON 17 lines 21-SEP-1987 11:30 -------------------------------------------------------------------------------- I borrowed an RA81 from our VAXcluster to use on an 11/70 under RSX11M-Plus. Problem is that when I try to run the bad block scan on it, the RCT task reports "Write back caching data lost. Unit write locked", and BAD dies. The manual says this is due to the "writeback cache in use" bit being set in the disk's RCT, and that this is very rare. Question is, how do I clear this bit? Is there a way to clear it online? Would reformatting the RA81 on the HSC do it, or would I have to format from the 11/70? ================================================================================ Note 55.1 Problem with borrowed RA81 1 of 3 EISNER::DOHERTY "Bob Doherty" 23 lines 22-SEP-1987 00:08 -< RCT disk locking >- -------------------------------------------------------------------------------- > Problem is that when I try to run the bad block scan on it, the > RCT task reports "Write back caching data lost. Unit write locked", > and BAD dies. We observed this problem several years ago when the RA disks were first announced. One of our RA-80's had exactly this behavior under RSX (I believe v4.0). We were told by Field Circus that this was due to the disk being used on a system which did disk caching (the hypothesis was RSTS or VMS) and being dismounted dirty. If you can move it back to your VMS system and dismount it cleanly, then that dirty bit should be turned off. BTW, if the disk is part of a volume set, the disks must be dismounted in reverse numerical order, or the whole dismount is marked dirty. In the absence of moving it back to the VMS system, the only way out, and the one we used at the time, was to hardware format the volume, writing a new RCT table. For the RA-80 this took several hours, so count on a lot longer for an RA81. Bob. ================================================================================ Note 55.2 Problem with borrowed RA81 2 of 3 EISNER::NORTON 7 lines 23-SEP-1987 11:52 -------------------------------------------------------------------------------- > If you can move it back to your VMS system and dismount it > cleanly, then that dirty bit should be turned off. That didn't help. Neither did reformatting and verifying on the HSC70, although it doesn't look like the HSC formatter completely rebuilds the RCT. ================================================================================ Note 55.3 Problem with borrowed RA81 3 of 3 EISNER::KENNEDY "Terry Kennedy" 7 lines 23-SEP-1987 22:27 -< A silly solution >- -------------------------------------------------------------------------------- > That didn't help. Neither did reformatting and verifying on the HSC70, > although it doesn't look like the HSC formatter completely rebuilds > the RCT. Well, the hardware is acting like it's broken, even if it isn't! There- fore, if you have a contract, you should call Field Circus to have the thing fixed!  ================================================================================ Note 56.0 When is the HSC saturated? 3 replies EISNER::WALLIS "Barry Wallis - Fleetwood Enterprises" 8 lines 25-SEP-1987 14:44 -------------------------------------------------------------------------------- How can I know when I am going to run out of HSC power? I am currently using SPM to monitor the "Work Available %" on the HSC for any given prime time period. My problem is determining what an acceptable threshold of performance is (e.g., should I start pricing another controller when I am 80% saturated)? ================================================================================ Note 56.1 When is the HSC saturated? 1 of 3 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 6 lines 25-SEP-1987 16:33 -< Don't sweat. >- -------------------------------------------------------------------------------- DEC, to the best of my knowledge, has no drives that are high enough in performance such that an HSC would be overdriven if filled to its capacity. An HSC can handle 32 drives at most. An HSC70 is only 50 - 60% utilized by such a load. An HSC50 would only be about 80-85% (maybe 90%). Why worry, especially if you have tape drives on the HSC as well, which reduces the number of disks that can be physically attached? ================================================================================ Note 56.2 When is the HSC saturated? 2 of 3 EISNER::WALLIS "Barry Wallis - Fleetwood Enterprise" 10 lines 25-SEP-1987 19:52 -< To sweat or not to sweat... >- -------------------------------------------------------------------------------- Are you saying that there is no need to worry until the HSC (in this case an HSC50 driving 3 tape drives and 8 RA81s) hits 100%? Our HSC consistently runs a higher workload than any of our disks (and usually it is higher than all of our disks put together). I am not worried about driving the HSC from the disk side but from the CPU side. How much demand is too much? Please forgive me for not phrasing my question(s) too well and let me know if I am not making myself clear. ================================================================================ Note 56.3 When is the HSC saturated? 3 of 3 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 18 lines 28-SEP-1987 08:48 -< What I mean is... >- -------------------------------------------------------------------------------- We had HSC50s with three tapes and eight RA81s as well (actually, we had 16 RA81s, but since they were dual ported to dual HSCs, we had one HSC control eight and the other control eight as well), and we saw peak usage of 70% or less. I don't think your CPU can overdrive the HSC, We had 8600s (two) and 780s (two). Of course, your I/O loading can be (probably is) different. It's my understanding from some document I read given me by my DEC rep that five completely I/O bound 780s can overdrive an HSC50, so maybe two (or three) I/O bound 8600s can as well, but with our workloads, it was unlikely all CPUs would be I/O bound simultaneously. We have since purchased another 8600 (in 1986), replaced the HSC50s with HSC70s, and added four more RA81s. We will be adding an SA482 in the course of the next couple months. The HSC70s have no trouble whatsoever with handling 20 RA81s, even when one is down and the other has to take up the entire load. What burns us is the speed of the RAs. The HSCs are just sitting there waiting on the disks, due to their relatively slow latency. The document I mentioned before stated that the main bottleneck in the CPU to HSC to disk pathway would be, as we have verified, the HSC to disk path.  ================================================================================ Note 57.0 Laser Printer Questions 32 replies EISNER::KASPER "Beverly T. Kasper" 13 lines 25-SEP-1987 16:10 -------------------------------------------------------------------------------- Our documentation group has decided it wants a Laser Printer; they went out and bought an LN03, and now they wants us to tell them how to make it do the things they need. Specifically multiple fonts on the same page. The documentation for the LN03 rivals the old RSX I/O Operations Manual for readability; is it in there, or is there really no way to do this? The documentation people are about ready to give up on the LN03 and get something else. Hardware seems like the likeliest place to ask this, which isn't saying much . . . ================================================================================ Note 57.1 Laser Printer Questions 1 of 32 EISNER::HAHN 5 lines 25-SEP-1987 16:20 -< LNO3 with which WP? >- -------------------------------------------------------------------------------- Which kind of WP are you using, with WPS-PLUS it is extremely simple with other WP it means sometimes setting up escape sequences, I have the manual and can give you some of the sequences? Pierre ================================================================================ Note 57.2 Laser Printer Questions 2 of 32 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 2 lines 25-SEP-1987 16:35 -< LN03 and TeX font loading >- -------------------------------------------------------------------------------- There are also TeX packages (if you use TeX) that can load fonts to the LN03. One such is available from Kellerman and Smith, I believe. ================================================================================ Note 57.3 Laser Printer Questions 3 of 32 EISNER::HASSINGER "Bob Hassinger" 19 lines 25-SEP-1987 17:10 -< It's quite do-able, depending... More info please. >- -------------------------------------------------------------------------------- Yes, the prime question is what you are doing on the software side. Selecting fonts from the built in ones or ones in plug in cartridges is simple to do (once you get past the documentation) if you are in a position to insert escape sequences in your output. In the WPS-PLUS case you can do it by using the User Modifiable Printer Tables that are now (more or less) documented - this allows inserting control blocks in document that will cause the desired font selection sequences to be sent. In the DECmate-WPS case it is even easier, or at least more direct. Both TeX and the DECpage add-on for WPS-PLUS do down line font loads and then format their output with the required stuff to use what they have loaded. More details on your situation, what you are trying to do and what you have tried so far will help us give more specific assistance. ================================================================================ Note 57.4 Laser Printer Questions 4 of 32 EISNER::NAGY "Frank J. Nagy, VAX Guru" 6 lines 28-SEP-1987 07:54 -< TeX for LN03 on L&T SIG tape >- -------------------------------------------------------------------------------- To add to 57.2 (Brian Tillman): A complete TeX system with drivers for the LN03 and fonts for downloading the LN03 is available on the most recent Languages & Tools SIG tape (LT86A). This tape was distributed last year, I do not know if it is in the DECUS Library. ================================================================================ Note 57.5 Laser Printer Questions 5 of 32 EISNER::KASPER "Beverly T. Kasper" 10 lines 28-SEP-1987 09:03 -< More info >- -------------------------------------------------------------------------------- The text they're trying to print is straight edit files. We are, it turns out, a Beta Test site for VAX Script Printer Software. This package loads the fonts and lets you use them, but the documentation on it is apparently very unclear. I didn't realize when I posted the original topic that this was a Field Trial product -- I'd just been told it was a new LN03 package from DEC. This greatly decreases my hopes of getting info here, but if any of you knows anything about this, I'd appreciate it. ================================================================================ Note 57.6 Laser Printer Questions 6 of 32 EISNER::NORTON 10 lines 28-SEP-1987 10:12 -< Better LN03 Programmer Manual >- -------------------------------------------------------------------------------- BTW - There is a newer Programmer Reference Manual for the LN03, that I'm told is MUCH better in the area of programming than the one that comes with the unit. The part number is EK-OLN03-RM-002. Good luck, -Bill ================================================================================ Note 57.7 Laser Printer Questions 7 of 32 EISNER::HASSINGER "Bob Hassinger" 84 lines 28-SEP-1987 10:28 -< Well, that is a different story. ABOUT the LN03R... >- -------------------------------------------------------------------------------- Well..., that changes things Bev. It just so happens I have had an LN03R sitting here for the last couple of months, doing very little. I believe this is the Script Printer you are talking about - looks exactly like an LN03 except the plug in cartridge slots are closed up and the test page you get when you turn it on or push the test button looks quite different. It turns out to be quite a different beast however when you go to use it. It only understands the Adobe Postscript language. It does not have the native capability to accept even straight ASCII such as you would send to a Teletype, let along the DEC escape sequences an LN03 or LN03S (LN03 Plus) understand. On the other hand, Postscript is very powerful. This printer can do anything the others can do and MUCH more if you give it the right help. Mine has not been getting much work done because it has been waiting for the long overdue software package for the VAX that has been called the "translator". From the users point of view it is supposed to be similar to the software that drives DEC's larger 40 page-per-minute laser printer. The major difference that has been discussed is of course that the one uses Ethernet and the other uses RS232. I believe there is also an issue of downloading software for the built in VAX engine in the larger one. I assume the software you are talking about is this same "translator" stuff that we are now supposed to get the middle of October. It is supposed to be, in effect, a print symbiont that knows how to translate ASCII, Sixel, Regis, and Tek 4010/4014 (I think) into something that will print on the LN03R. It is my understanding that when you drive the LN03R from a VAX via this translator package you can send just about anything you could send to a normal LN03 or LN03S and it will be output. At the Spring Symposium I looked into it carefully and was told explicitly that it would handle the outputs that the TeX package mentioned here sends. They include bit map font down loads. This function had not been explicitly documented as working anywhere else. It is my understanding that this means you can use the TeX package available on the L&T tape and elsewhere, just as you would on a real LN03. This leaves two additional areas. The first is the issue of getting the translator to notice and act on requests for downloads of fonts. In the existing TeX case the font loads are handled by the TeX back end. It just stuffs the whole load sequence with all the bit maps for the characters from the fonts it needs in a preamble to each output file. The better thing in this case would be for the translator to notice something in the input that would specify fonts that IT would go out and find and include in the down load. Even better would be for the translator to keep track of what is down loaded across print jobs and avoid extra work. (It is not clear how easy this is in a situation where the printer is shared between systems.) In any case, I got a hint that this capability was to be included in the translator but no one could say anything about how it was to work. I am very interested in finding out because I have a custom logo font I would like it to support for me. Based on what you say Bev, I suspect Colorado is going to start getting calls from me for clarification a day or so after the software finally gets here. The other variation is that your LN03R already contains a lot of built in fonts of high quality. They are stored in generic outline form. When a request is sent to the printer the appropriate set of bit maps for the selected font, at the selected scaling, aspect ratio and so on are computed. The printer is bright enough to save these across jobs and avoid recalculation when possible - it even spends it's idle time after startup pre-computing some standard choices of font and size. At the moment, getting to these fonts from TeX is a little hard but there are products that are very nearly able to do it. There are some EXPENSIVE typesetting packages for the VAXstation that can use them too. Of course there are a lot of typesetting and page composition packages for micros that can produce Postscript output too (Pagemaker for example). While waiting for the translator software I have been getting some output from our LN03R by two means. First, you can write raw Postscript and send it. There is a lot of potential in this - the printer is a programmable computer of considerable power. You can do a lot going this way. Second, I have been using VAX GKS to send text and graphics to the printer with great results. Some of this discussion drifts into areas that may belong in the DESKTOP_PUBLISHING conference. I suggest looking there too. Bob H ================================================================================ Note 57.8 Laser Printer Questions 8 of 32 EISNER::KASPER "Beverly T. Kasper" 9 lines 28-SEP-1987 16:35 -< Sounds like expected bad news for the LN03 >- -------------------------------------------------------------------------------- Thanks, Bob. I actually started putting this in DESKTOP_PUBLISHING originally; I would have done so if I'd known it was stricly a software problem. I'll take it there now. All of what you suggest, by the way, sounds good to me, but won't to the techwriters. They want to be able to use this thing without extensive training, and there isn't really anyone available to do the s/w work who has any experience with this sort of thing. ================================================================================ Note 57.9 Laser Printer Questions 9 of 32 EISNER::CONROY "Alan Conroy" 33 lines 26-JUL-1988 18:29 -< HELP!! Only 65 lines per page >- -------------------------------------------------------------------------------- I have a problem getting our LN03 to print 66 lines on a page. It is a remote printer connected through a DECServer 200 to various and sundry VAX systems. I set the LTA port to NOSPOOL just to make sure that the LATSYM symbiont didn't do anything "helpful" whilst I tried to debug... All output was done via the COPY command. I turned the printer on and off to make sure it was properly reset. I then sent the RIS codes (ESC c) just to make really sure. Then I copied a file to the printer. Its format was: 1 2 3 4 5 6 7 8 9 10 1 2 . . . and so on, up to 70. Now, I would expect that the first 66 lines would end up on the first page and the remainder on the next one. No such luck! The first 65 lines ended up on the first page, but line 66 and up were on the next page. You might think that one line would be a minor detail, but it is a major pain. So I went to the programmer reference manual and looked up the RIS code. The manual says that the lines per page are set to 66 by this, and the top margin is set to 1 and the bottom margin to 66. So why only 65 lines? So I sent the code to specifically set the bottom margin to 66. No effect. Then I tried to set the printer to origin mode, but the printer got a blinking 2. After a day of going around and around on this I'm getting rather upset. If anyone can offer me any insights into this, I'd sure like to here them! ================================================================================ Note 57.10 Laser Printer Questions 10 of 32 EISNER::MAYHEW "Bill Mayhew" 11 lines 26-JUL-1988 22:05 -< Print another page and see what THAT does >- -------------------------------------------------------------------------------- < Note 57.9 by EISNER::CONROY "Alan Conroy" > -< HELP!! Only 65 lines per page >- Have you tried printing 3 pages? How much do you get on page 2? If you get the full 66 on page 2, then I would think something funny is happening at the beginning of, or before, your COPY (like maybe the RIS code is followed by a newline?). If you get only 65 on page 2, then it's definitely some kind of LN03 setup problem. I saw this same problem at one point and I can't remember what the fix turned out to be, but it was, I think, something I was doing in "setting up" the printer. ================================================================================ Note 57.11 Laser Printer Questions 11 of 32 EISNER::HAHN "DECUServe Vice Chair" 8 lines 27-JUL-1988 09:18 -< VMI may be the culprit >- -------------------------------------------------------------------------------- From using the LN03 with a decmate I can remember that I could only type 65 lines/page. to print 66 lines per page I had to decrease the vertical motion index (VMI) to print about 6+ lines to the inch Check the Printing commands on page 93 (PFS,SGR,DECSTBM commands) and/or change the VMI (# of 1/300 ths of an inch per line). Pierre ================================================================================ Note 57.12 Laser Printer Questions 12 of 32 EISNER::HASSINGER "Bob Hassinger" 5 lines 27-JUL-1988 11:01 -< Sounds familiar... BTW who DO you call for support on this? >- -------------------------------------------------------------------------------- I think Pierre is on the right track. Remember, the LN03 can't print the whole page. There is a margin all around the page that it can not image in. This can lead to a number of problems along these lines. Bob H ================================================================================ Note 57.13 Laser Printer Questions 13 of 32 EISNER::PROVOST 16 lines 27-JUL-1988 11:17 -< 66 lines here >- -------------------------------------------------------------------------------- We have 4 LN03+'s on DECserver200's with queues on 5 uVAXen and 2 VAXen. I printed 159 lines with successive numbers on each. I get 66 lines/page. I do get a blank line at the bottom, so there's paper enough for 67 lines. We use "$ set terminal/perm/device=LN03/nobroad/nowrap/width=132 res$device" We map the applications ports as follows: "SET PORT RES$DEVICE /APPLICATION /NOTDE=TS01 /SERVICE=RES_LN03 /NOLOG" Any surprises here? Maybe it's due to our's being "+" version of LN03? Tom ================================================================================ Note 57.14 Laser Printer Questions 14 of 32 EISNER::HAHN "DECUServe Vice Chair" 8 lines 27-JUL-1988 12:00 -< 6.25 lines per inch to print 66 lines >- -------------------------------------------------------------------------------- Looking into the manual a little further I see that in order to get 66 lines on 8.5x11 paper in portrait orientation you should set the SVS vertical spacing to 6.25 lines per inch (see the paragraph 4.7 Spacing and 4.8 Page Print Area and Margions). It would be interesting if Provost would give an actual measure of the lines per inch on the printout mentioned in the previous note. Pierre ================================================================================ Note 57.15 Laser Printer Questions 15 of 32 EISNER::HASSINGER "Bob Hassinger" 9 lines 27-JUL-1988 13:25 -< Are they on queues Tom? >- -------------------------------------------------------------------------------- > It would be interesting if Provost would give an actual measure > of the lines per inch on the printout mentioned in the previous > note. Yes, and also Tom, are these printers used directly through the ports or are they on print queues. If on queues then are there any setup modules, etc. that might be controlling the SVS setting, etc. Bob H ================================================================================ Note 57.16 Laser Printer Questions 16 of 32 EISNER::LEDERMAN "Bart Z. Lederman" 20 lines 27-JUL-1988 14:33 -< Where do you print after setting the printer? >- -------------------------------------------------------------------------------- I have an LN03+ and have used LN03s. I recall there are at least two portrait page layouts: one with a margin, one without. I think you can get 66 lines on either page. I think Bill's suggestion is probably right. If your test file looks like this: [xxxxx 1 2 3 4 5 .... Then the escape sequence which sets up the printer characteristics is printing on line 1, and the line you have numbered as 1 is on line 2, etc. so you end up miscounting the printing lines. Make sure the line you are counting as 1 has it's first printing character at the end of the escape sequence which sets up the printer or is the first character after a form-feed. ================================================================================ Note 57.17 Laser Printer Questions 17 of 32 EISNER::MAYHEW "Bill Mayhew" 6 lines 27-JUL-1988 17:34 -< LN03 (basic) should print 66 lpp @ power-up >- -------------------------------------------------------------------------------- Re: .12 and some successive notes... The power-up default of the LN03 is 6.25 lines per inch, which fits 66 lines on the page. So if 66 lines really don't fit (to be confirmed by counting the lines on page 2), then some software somewhere is tweaking the printer (or possibly the firmware is way out of date?) ================================================================================ Note 57.18 Laser Printer Questions 18 of 32 EISNER::CONROY "Alan Conroy" 23 lines 27-JUL-1988 17:49 -< It works! >- -------------------------------------------------------------------------------- Well, I finally solved the problem with the help of all of you who responded. The problem was that the RIS code set the lpi to 6, whereas you must have the lpi set to 6.25. At 6 lpi only 65 lines show up, at 6.25 all 66 will show up. Part of my confusion was due to the manual. Section 5.4.2 talks about vertical line spacing and gives examples of output at various lines per inch. Nothing there is listed as 6.25 lpi. However, on the next page in section 5.4.3. it discusses vertical pitch, including how to get 6.25 lpi. Comparing the two sections, I now see what they were getting at in 5.4.2, but it certainly could have been stated more clearly. That was the first problem. The second problem was that I assumed (quite incorrectly) that RIS would set the vertical pitch to 6.25 (or whatever was needed to put 66 lines on a page). Surprise! RIS does nothing to pitch, either vertical or horizontal. If I had been more careful in reading the section on RIS, I would have noticed. * sigh * So, after putting RIS and DECVERP codes in the file, suddenly I got 66 lines on the page. For those of you who could benefit from my ordeal, the code you need is: ESC 1 1 z (that is, an escape, two one's, and a lower-case Z). I placed this code into a module in a .TLB and set up the que: /SEPARATE=(RESET=(RESET,10PITCH,66LPP)) where RESET is a module containing RIS, 10PITCH is a module containing the code to change to 10 cpi, and 66LPP is a module which contains the above code. Once again, thank you for all your wonderful help! :-) ================================================================================ Note 57.19 Laser Printer Questions 19 of 32 EISNER::CONROY "Alan Conroy" 9 lines 27-JUL-1988 17:53 -< one other thing... >- -------------------------------------------------------------------------------- > The power-up default of the LN03 is 6.25 lines per inch, which > fits 66 lines on the page. So if 66 lines really don't fit (to > be confirmed by counting the lines on page 2), then some software > somewhere is tweaking the printer (or possibly the firmware is way > out of date?) I forgot to say this, but the power-up condition leaves the printer at 6 lpi, which has the problems mentioned in my previous message (i.e. you need to specifically set the printer to 6.25 lpi to get 66 lines on the page). ================================================================================ Note 57.20 Laser Printer Questions 20 of 32 EISNER::MAYHEW "Bill Mayhew" 20 lines 28-JUL-1988 00:12 -< Not that I want to argue with success, but... >- -------------------------------------------------------------------------------- < Note 57.19 by EISNER::CONROY "Alan Conroy" > >> The power-up default of the LN03 is 6.25 lines per inch, which > I forgot to say this, but the power-up condition leaves the printer at > 6 lpi, Something still smells, here. Page 125 of the LN03 Programmer's Reference Manual, EK-OLN03-RM-002, "Initial Operating Values", specifically states the vertical spacing is set to 6.25 lpi at power-up. Page 123 specifically states that the initial margins produce "a page size of 66 lines per page for 8.5 x 11 and A4 paper." I copied a file like yours (numbered lines) to my LN03 immediately after powering it up and it printed 66 lines on the first page. This was done with no RIS or any other stuff sent before my text. The firmware rev on my LN03, according to the status-test sheet, is DEC004.4. My printer is attached to a PDP-11, and there's no "symbiont" involved. ================================================================================ Note 57.21 Laser Printer Questions 21 of 32 EISNER::HAHN "DECUServe Vice Chair" 4 lines 28-JUL-1988 09:31 -< c = 1 line >- -------------------------------------------------------------------------------- The problem may have been with the original set up an c may have been followed with a which would be a null line; that may be line 0 and therefore only 65 line numbers were printed! Pierre ================================================================================ Note 57.22 Laser Printer Questions 22 of 32 EISNER::BURNET "John Burnet" 3 lines 31-JUL-1988 01:12 -< define/form/margin=top=? >- -------------------------------------------------------------------------------- Check the definition for the form being used by the laser-printer queue as the default (set up by init/queue or set queue as /default=form=foo). Does the form specify a top margin? ================================================================================ Note 57.23 Laser Printer Questions 23 of 32 EISNER::CONROY "Alan Conroy" 41 lines 1-AUG-1988 14:35 -< DECVERP <> SVS >- -------------------------------------------------------------------------------- > Something still smells, here. Page 125 of the LN03 Programmer's > Reference Manual, EK-OLN03-RM-002, "Initial Operating Values", > specifically states the vertical spacing is set to 6.25 lpi at > power-up. Page 123 specifically states that the initial margins > produce "a page size of 66 lines per page for 8.5 x 11 and A4 paper." Vertical spacing is NOT the same as vertical pitch. This is what confused me in the first place. RIS does nothing with either vertical or horizontal pitch. Pitch is what you need to change (see page 80 - DECVERP) to get 66 lpp. And yes, I know that page 123 states that this sets the printer up for 66 lpp, however, you can tell that something is wrong with the statement right off. The LN03 only prints 66 lpp if the font you're using prints 66 lpp. Now, I thought that the default font would print that way, however it didn't. > I copied a file like yours (numbered lines) to my LN03 immediately > after powering it up and it printed 66 lines on the first page. > This was done with no RIS or any other stuff sent before my text. > The firmware rev on my LN03, according to the status-test sheet, > is DEC004.4. Hmmm. Well, my firmware rev is also DEC004.4. After reading your message I thought that maybe the RIS was causing the problem, but I did the same thing you did and only got 65 lpp. The only thing I can figure is that if yours works and mine doesn't, then mine must have a problem. What's the model number on yours? Mine is LN03A-A2. Also, was the paper switch set for NA or 8.5x11? Mine was set 8.5 x 11. For those of you suggesting that maybe a spurious CRLF sequence found its way into my test file, you can rest assured that this was not the case. If you were to use EDT, the file would look like: c1 2 3 . . . Also, though the margin settings on the que could cause this problem, in this case they are not. The only difference between the test file that worked and the one that didn't is that the one that did had the DECVERP codes after the RIS, to change the pitch to 6.25. Its working now, but I think I'm going to talk to the maintenance people... ================================================================================ Note 57.24 Laser Printer Questions 24 of 32 EISNER::MAYHEW "Bill Mayhew" 13 lines 1-AUG-1988 15:55 -< Nothing like a good menage-a-trois! >- -------------------------------------------------------------------------------- < Note 57.23 by EISNER::CONROY "Alan Conroy" > > What's the model number on > yours? Mine is LN03A-A2. Also, was the paper switch set for NA or 8.5x11? > Mine was set 8.5 x 11. Mine is an LN03-A2 (not LN03A, i.e. mine is the earlier model) and the paper switch is set for 8.5x11. (I believe the other setting is "A2", a European paper-size standard, not "NA", but anyway...) Now that I've shown you mine and you've shown me yours, I think we need somebody else, esp. someone with an LN03A-A2, to try this and report back. ================================================================================ Note 57.25 Laser Printer Questions 25 of 32 EISNER::COY "Dale E. Coy" 1 line 1-AUG-1988 18:07 -< Confirmation >- -------------------------------------------------------------------------------- LN03-A2 set to 8 1/2, Rev DEC004.4 prints 66 lines. ================================================================================ Note 57.26 Laser Printer Questions 26 of 32 EISNER::CONROY "Alan Conroy" 0 lines 2-AUG-1988 15:10 -< Anyone with an LN03A-A2 who can provide info? >- -------------------------------------------------------------------------------- ================================================================================ Note 57.27 Laser Printer Questions 27 of 32 EISNER::WILSON_S 11 lines 30-MAR-1989 16:58 -< LN03 truncates screen dumps from GPX >- -------------------------------------------------------------------------------- We just recieved a LN03 plus and we're using it to output a portion of a GPX screen using the Workstation Options menu "print (portion of) screen". It works well until we attempt to print an area that exceeds a certain size. Attempting these larger areas generates a picture that is truncated. There are RAM packs that may be plugged into the ln03. Does anyone know if these RAM packs will allow larger screen dumps to be printed on the ln03? Any comments are greatly appreciated! ================================================================================ Note 57.28 Laser Printer Questions 28 of 32 EISNER::BURNET "John Burnet" 31 lines 30-MAR-1989 20:26 -< RAM cartridges only give you more font storage space >- -------------------------------------------------------------------------------- > There are RAM packs that may be plugged into the ln03. Does anyone > know if these RAM packs will allow larger screen dumps to be printed > on the ln03? The RAM cartridges won't help; they provide font memory, not bitmap memory. The LN03+ is configured (straight out of the box) with one meg of bitmap memory, which is all it needs. Quoting from the *LN03 Plus Programmer Reference Manual* (EK-LN03S-RM-001): The LN03+ prints Tektronix images on a fixed page area within the physical page. This fixed area is known as the Tekpage and is equal to 19.3 x 25.64 cm (7.68 x 10.24 inches). ... The LN03+ prints within the same size Tekpage whether it is using 8-1/2 x 11 inch or A4 paper. ... The unit of screen addressing is the Tekpoint. By default, there are 1024 Tekpoints horizontally and 768 Tekpoints vertically within the Tekpage. ... You can achieve much higher resolution if you activate the enhanced graphics module (EGM) feature by sending an extra address byte. The number of Tekpoints increases to 4096 horizontally and 3072 vertically. The size of each Tekpoint goes down to 0.75 pixel, with increased positioning resolution. You need 12 bits to address the printer in EGM mode. The EGM referred to is a Tek option for Tek terminals; the LN03+ implements some of the option's features, including increased addressable resolution. I don't think that any Tek-mode upgrades to the LN03+ are available; what you've got is all you're gonna get. :-) I don't know what the software that implements the "print screen" menu option is doing, so I can't say where the problem lies... maybe someone else can shed some light on this. ================================================================================ Note 57.29 Laser Printer Questions 29 of 32 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 25 lines 2-MAY-1989 06:56 -< RAM fonts for LN03s : font memory >- -------------------------------------------------------------------------------- > There are RAM packs that may be plugged into the ln03. Does anyone < > know if these RAM packs will allow larger screen dumps to be printed < > on the ln03? < The RAM cartridges are for extending font memory. To abbreviate two months worth of phone calls to Colorado into one note, we bought an LN03+ (aka LN03s). We discovered italic printing isn't possible with the built-in fonts. Fine; we ordered a "U.S. Legal" font cartridge with italic fonts in it. We now find that if we want the fonts rotated to print in Landscape mode, we need additional memory, which means a RAM. DECdirect confirms this. Of course, they could have told us that when we bought the printer! Also, if you plan to do Tektronix emulation on an LN03+, you may need RAM for complex plots, according to the device interface information for our plotting package. And now, a question of mine own : Anybody know why a plot file in Tektronix format prints OK in landscape mode on an LN03+, but comes out poorly in portrait mode on an LN03 ScriptPrinter? The same plot file, sent to the "LN03s" queue, and to the Tektronix data-type queue for the LN03 ScriptPrinter. The ScriptPrinter changed the orientation, made dashed lines of solid ones, dropped pixels out of the axis labels, etc. ================================================================================ Note 57.30 Laser Printer Questions 30 of 32 EISNER::TINKELMAN "Bob Tinkelman - CCA" 23 lines 2-MAY-1989 17:58 -< Sounds like a bug in VAX ScriptPrinter software >- -------------------------------------------------------------------------------- RE: < Note 57.29 by EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" > > Anybody know why a plot file in Tektronix format prints OK in landscape > mode on an LN03+, but comes out poorly in portrait mode on an LN03 > ScriptPrinter? The same plot file, sent to the "LN03s" queue, and to the > Tektronix data-type queue for the LN03 ScriptPrinter. The ScriptPrinter > changed the orientation, made dashed lines of solid ones, dropped pixels > out of the axis labels, etc. First for my qualifications (or lack thereof) for answering this question: While we have an LN03R ScriptPrinter, we do not use the Tektronix translator. [The need never arose.] The LN03R does not do Tektronix emulation. It sounds like you are using the DEC ScripPrinter software whose print symbiont has an on-the-fly Textronix to PostScript translator. And it sounds like there is a bug in that translator. Maybe you should submit an SPR. However, wrt the orientation issue, I'd go back to the manual first and make sure that DEC claims that PARAM=PAGE_ORIENTATION=PORTRAIT works in conjunction with the Textronix translator. I had thought it was for the ANSI translator only. But, as I said earlier, I've never used the Textronix one. ================================================================================ Note 57.31 Laser Printer Questions 31 of 32 EISNER::COY "Dale E. Coy" 2 lines 2-MAY-1989 22:21 -< The other shoe >- -------------------------------------------------------------------------------- You'll also (sooner or later) come across the bugs in the LN03+ Tektronix emulation. ================================================================================ Note 57.32 Laser Printer Questions 32 of 32 EISNER::HASSINGER "Bob Hassinger" 7 lines 3-MAY-1989 10:51 -< Which version of ScriptPrinter software? >- -------------------------------------------------------------------------------- >< Note 57.29 by EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" > > -< RAM fonts for LN03s : font memory >- Re: problems outputting Tektronix format to the LN03R: Assuming you are using the ScriptPrinter translator, which version of it? V2 was released fairly recently (check the LATEST... conference). It is quite possible it fixed or introduced bugs...  ================================================================================ Note 58.0 Emulex performance 1 reply EISNER::TILLMAN "Brian Tillman, Smiths Industries." 3 lines 30-SEP-1987 16:18 -------------------------------------------------------------------------------- Anyone have experience with Emulex's SMDI disk subsystems, like the XMD drives? If so, how do they work for you? These are the RA plug compatible jobs that can be attached directly to HSCs, UDAs or KDAs. ================================================================================ Note 58.1 Emulex performance 1 of 1 EISNER::KERSHENBAUM "Richard Kershenbaum, The Unive" 35 lines 5-MAY-1988 12:02 -< Emulex SMDI/Sabre IV performance >- -------------------------------------------------------------------------------- Funny that nobody's replied to this for so long....Guess it's up to me. We've been running an SMDI subsystem with the new CDC Sabre IV (9720-850) drives since January. We were something like the second site to install that box, which was a little scary. Since installation, we've had only one real hardware failure, when one of the SMDI boards locked up and wouldn't talk to the HSC. CDC, who maintains our Emulex gear, came in the next day and installed ground straps everwhere, and we haven't seen the problem again. We've had one revectoring of a bad block on one of the four drives -- normal for new HDAs. So, overall, the equipment has proven very reliable so far. In fact, we had enough confidence in the Sabres to put our system disk there. As for performance, the Sabres are considerably faster than either the RA81 or RA82. We obtained a copy of the Digital Review disk benchmark suite and it showed a 25% improvement over the RA81s despite the moderating influence of the HSC. They're impressive in terms of size and power consumption also. The actual drives are about the size of a shoebox; they fit two across in a 19" rack. You can buy a 60" high cabinet with 12 drives, 655mb each. It would take 3 times the floor space for the same amount of storage with RA82s. The price was right, too -- 30% under DEC's discounted bids for an SA482. So....we're pleased with the SMDI/Sabre IV combination. I'd be happy to answer any further questions here, at 913/864-0445, or send to richard@ukanvax on Bitnet.  ================================================================================ Note 59.0 How to update an HSC 10 replies EISNER::DROPPERS "Seton R. Droppers" 13 lines 5-OCT-1987 08:33 -------------------------------------------------------------------------------- I have just gotten USC UTILITY and HSC SYSTEM tapes from field service for my HSC50. Now what do I do with them to make them work in my system? The tapes are 3.50. I have a 2.50 "preliminary" manual, no later manual. Do I read that manual and set parameters as directed, or are there some things I had better keep in mind while I am doing all of this. And, can I update one HSC, fail the system over, and see how it works before I update the other one? Thanks. Seton ================================================================================ Note 59.1 How to update an HSC 1 of 10 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 15 lines 5-OCT-1987 15:54 -< How we did it. >- -------------------------------------------------------------------------------- It's basically pretty simple. Just fault all your disks over to the HSC you are NOT going to update, open the HSC front door, set the HSC offline, reboot the HSC (for a fresh start), type CTRL-Y followed by SHOW SYS, preserving the hard copy from the HSC console, make a couple backup tapes of the release media, mount a copy in drive one, push the init switch while holding the fault switch. Hold the fault switch until you see INIPIO-I Booting... on the console. Wait until the initialization message appears, type CTRL-Y followed by RUN SETSHO, and restore your system parameters from the hard copy you saved earlier. Confirm with SHOW SYS. Enter EXIT. If the HSC says it's rebooting, should it continue, answer yes. Turn the HSC online. Make HSC secure and fault the disks back over. This procedure came from my installation guide for V2.50. I couldn't lay my hands on the 3.50 guide at the moment, so be careful. ================================================================================ Note 59.2 How to update an HSC 2 of 10 EISNER::DROPPERS "Seton R. Droppers" 8 lines 5-OCT-1987 20:17 -< Thanks >- -------------------------------------------------------------------------------- Thanks. Guess I know what I will do for a couple of mornings... P.S. Has anyone else seen how long it takes an HSC50 to recover a shadowed RA82 with a full user load going... And what it does to user response? We had our HSCs ping pong a couple of times one morning, talk about angry user bases... Seton ================================================================================ Note 59.3 How to update an HSC 3 of 10 EISNER::ROCHELEAU "Rod Rocheleau" 20 lines 6-OCT-1987 11:57 -< Don't FAULT disks ... SHUTDOWN HSC >- -------------------------------------------------------------------------------- >> < Note 59.1 by EISNER::TILLMAN "Brian Tillman, Smiths Industries." > >> -< How we did it. >- >> >> It's basically pretty simple. Just fault all your disks over to >> the HSC you are NOT going to update, open the HSC front door, set >> the HSC offline You can do it that way but... just turn the HSC off and all disks will fail-over and will not show mount verification messages on all cpu consoles. This is the way that DEC recommended I do it. Also do not try to run 250 and 350 on a HSC50 pair. DEC does not even recommend that you try 300 and 350. If you have a large group of disks of lot of cpu's find the notes on HSC stuff in HARDWARE and look at the problems patchs for 350 Have fun. ================================================================================ Note 59.4 How to update an HSC 4 of 10 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 9 lines 6-OCT-1987 12:31 -< Two versions, one pair. >- -------------------------------------------------------------------------------- RE: < Note 59.3 by EISNER::ROCHELEAU "Rod Rocheleau" > -< Don't FAULT disks ... SHUTDOWN HSC >- | Also do not try to run 250 and 350 on a HSC50 pair. Of course, to get both HSCs up to current levels, you fault over to one, update the other, fault over to that one and update the first. This is what I was advocating and I don't count it as running both versions simultaneously. ================================================================================ Note 59.5 How to update an HSC 5 of 10 EISNER::DROPPERS "Seton R. Droppers" 23 lines 6-OCT-1987 12:33 -< Comments and Question >- -------------------------------------------------------------------------------- > < Note 59.3 by EISNER::ROCHELEAU "Rod Rocheleau" > > -< Don't FAULT disks ... SHUTDOWN HSC >- > You can do it that way but... just turn the HSC off and all disks > will fail-over and will not show mount verification messages We have volume shadowing on three (2*3=6total) of our disks, when an HSC fails over the new HSC must check and make the data on all disks in the shaddow set the same. To do this the HSC declares one of the shadow set member "most current" and does a copy operation from it to the other members. This procedure can take a long time on a system under load with our three virtual, shaddowed, disks (about one half to one hour, as I recall the last time). While this is occuring you take quite an I/O hit on the disks involved, causing already angry users to wonder about the abilities of the "new multi-million $" computer system... > { Also do not try to run 250 and 350 on a HSC50 pair. DEC does not even > recommend that you try 300 and 350. What about 301 to 350? Seton ================================================================================ Note 59.6 How to update an HSC 6 of 10 EISNER::ROCHELEAU "Rod Rocheleau" 18 lines 6-OCT-1987 16:30 -< 350 on both for FAILOVER to work >- -------------------------------------------------------------------------------- >> < Note 59.5 by EISNER::DROPPERS "Seton R. Droppers" > >> -< Comments and Question >- >> >> > { Also do not try to run 250 and 350 on a HSC50 pair. DEC does not even >> > recommend that you try 300 and 350. >> >> What about 301 to 350? DEC told me that fail-over would not work properly unless both HSC's were at 350. As it turned out 350 doesn't even work properly I dropped the whole cluster twice trying to complete the upgrade. There is a patch to 350 ... changes to 351 that fixes the failover problem on large clusters (we have 12 cpu's) and large number's of disks (we have two string 24 in one and 28 in the other. Rod ================================================================================ Note 59.7 How to update an HSC 7 of 10 EISNER::DROPPERS "Seton R. Droppers" 17 lines 6-OCT-1987 19:56 -< The CHRONIC problems of HSCs >- -------------------------------------------------------------------------------- > DEC told me that fail-over would not work properly unless both > HSC's were at 350. As it turned out 350 doesn't even work properly > I dropped the whole cluster twice trying to complete the upgrade. Well folks, this is getting serious. I have not yet attempted my HSC update, think I will look into this V351 and wait for a real quiet weekend night, then watch for my superviser to log off... Any other words of encouragement??? P.S. Has anybody had any problems updating one HSC, failing over to it, then updating the other one? How many of you have tried to update CHRONIC, and is this going to be a chronic problem? P.P.S. Will VMS4.6 run (without ability to fail a tape mount over) with 301? ================================================================================ Note 59.8 How to update an HSC 8 of 10 EISNER::PRIGOT "Jonathan M. Prigot" 4 lines 7-OCT-1987 10:23 -< V350/1 vs. V300 >- -------------------------------------------------------------------------------- What does V350/1 do that V301 doesn't? If it is worthwhile, who do I beat up to get the new release? FS? SDC? KO? ================================================================================ Note 59.9 How to update an HSC 9 of 10 EISNER::DROPPERS "Seton R. Droppers" 25 lines 7-OCT-1987 12:41 -< What it does, where to get it? >- -------------------------------------------------------------------------------- Chronic 350/1 adds the HSC support for "Limited Support for Dual-Pathed HSC Tape Drives" (Ref. Sec. 2.2.6, pg 2-9 of the VMS 4.6 release notes). In other words, you should not have to re-boot the node you were doing backup on when the HSC crashed any more. To get chronic 350 I talked with the following path: 1) Field Service: They said they had it but could not distribute it, call my account rep. 2) Account Rep: Yes I was supposed to call them, but SPS (Software Product Services) rep should handle, please call... 3) SPS Rep: Yes, Chronic was no longer supported by Field Service, but I needed to talk to HPS (Hardware Product Services) rep... 4) HPS Rep: Still waiting... 5) Field Service: Brought over tapes for 350, both utilities and system, but no manual (Still only have priliminary 250 manual and 300 release notes) 6) DECUServe: You all are helping make up for the lack of manuals and other contacts... ================================================================================ Note 59.10 How to update an HSC 10 of 10 EISNER::ROCHELEAU "Rod Rocheleau" 9 lines 8-OCT-1987 17:38 -< See Latest Release Information 28.* >- -------------------------------------------------------------------------------- < Note 59.9 by EISNER::DROPPERS "Seton R. Droppers" > -< What it does, where to get it? >- Read the 28.* notes in lastest release information. It was/is a place for HSC software updates. I have a note there 28.5 about my experience with getting updates for the HSC's Rod  ================================================================================ Note 60.0 LG01 clone? 2 replies EISNER::NORTON 7 lines 6-OCT-1987 10:05 -------------------------------------------------------------------------------- The recently announced Mannesmann Tally MT660 printer bears a striking (suspicious?) resemblance to DEC's LG01/LG02. Does anyone know anything about this, or have any comments about Tally printers? ================================================================================ Note 60.1 LG01 clone? 1 of 2 EISNER::PRIGOT "Jonathan M. Prigot" 2 lines 6-OCT-1987 10:14 -< DEC printers >- -------------------------------------------------------------------------------- It ought to since the LG01/02 uses a M-T engine. I cannot think of any "pure" DEC line printer. ================================================================================ Note 60.2 LG01 clone? 2 of 2 EISNER::NELSON 12 lines 15-OCT-1987 08:56 -< Look under the Digital label >- -------------------------------------------------------------------------------- < Note 60.1 by EISNER::PRIGOT "Jonathan M. Prigot" > > It ought to since the LG01/02 uses a M-T engine. I cannot think > of any "pure" DEC line printer. The ribbon that came with our LG01 had a Digital stick on label over the Mannesmann Tally label. The printer works OK so far. DEC appeared to have customized the firmware in it to make it compatible with the semi standard DEC escape sequences. Unfortunately, there are enough small differences in the implementation of the sequences to make it a pain.  ================================================================================ Note 61.0 STAT-MUX 4 replies EISNER::HAHN 15 lines 7-OCT-1987 10:47 -------------------------------------------------------------------------------- I am planning on using a STATMUX between a couple of buildings. Our mVAX is in one building and 5-6 users are in another building. Our phone system has both data and voice intergrated in a single instrument. I could have 12 instruments (6 at the mVAX, 6 for users) to "call" into the mVAX, or I could dedicate 2 instruments with STATMUX. The "modem" capability of the phone instrument is 300-19200 asynch or 56kb sync. Any one has suggestions on a vendor(s) of STATMUX? Pierre ================================================================================ Note 61.1 STAT-MUX 1 of 4 EISNER::ROCHELEAU "Rod Rocheleau" 12 lines 8-OCT-1987 16:50 -< Could you use a TDM mux >- -------------------------------------------------------------------------------- < Note 61.0 by EISNER::HAHN > -< STAT-MUX >- I have always had some type of problems with stat mux's, But I have been having very good luck with a TDM mux that we are running between buildings. It does require 2 pair of copper wires ... no equipment ... because it works at about 780kb. It is available in either 8 or 16 port configuration. The vendor is SEQUALINK ... p.s. it rides through lighting storms that nothing else survives. ================================================================================ Note 61.2 STAT-MUX 2 of 4 EISNER::ERSKINE "Chris Erskine" 5 lines 12-OCT-1987 07:18 -< Stat Muxes >- -------------------------------------------------------------------------------- I have worked with both Timeplex and Micom Stat muxes with out any problems. DEC even now has some of its own muxes which I do not know anything about. If you have ethernet, you could even go as far as to try their mux-server or whatever the call the mux/terminal server they have. ================================================================================ Note 61.3 STAT-MUX 3 of 4 EISNER::KENNEDY "Terry Kennedy" 16 lines 12-OCT-1987 21:25 -< Too expensive! >- -------------------------------------------------------------------------------- > I have worked with both Timeplex and Micom Stat muxes with out any... We have several Micom units here. If you can get 2 pair of copper between the locations, all of this is too expensive. You can buy an 8-port mux from places like INMAC for less than the out-of-warranty exchange on a Micom. The Micom units work well if you are in full control of the data link side as well. The minute you invite the phone company to play, it gets weird. There is a bug in the units which only shows up when the leased phone line quality drops - both units think they are told some impossible set-up par- ameters and you have to unplug both ends to regain sanity. Of course, this means you have to re-enter all your setup parameters. They are also not very good at speed conversion, but they *do* pass sync DDCMP just fine if you have the option. Terry ================================================================================ Note 61.4 STAT-MUX 4 of 4 EISNER::TABOR "Bill Tabor" 10 lines 22-OCT-1987 22:07 -< Could you use a Data PBX instead of a Stat MUX >- -------------------------------------------------------------------------------- How far apart are the buildings? If they are close together one of the options you might think of using is a DATA PBX something like the Equinox PBX. I used it at my former job and what we did was put an expansion cabinet in the remote building. This allowed access of up to 64 users in the remote building. The also developed some very good lightning protection for this setup as well. We where getting hit on the average of two to three strikes a week and the last change they made protected us both for the terminals and switch.  ================================================================================ Note 62.0 FCOs for Self/Third Party Maint. 14 replies EISNER::DROPPERS "Seton R. Droppers" 8 lines 15-OCT-1987 12:48 -------------------------------------------------------------------------------- We are moving to third party maintanance of our VAXcluster consiting of 2x8650, 2x785, 6xRA81, 3XRA60, 2xHSC50, 1xSA482, 1xMICROVAXII. The one thing that we have questions about is how to obtain FCOs once DEC is no longer servicing the cluster? Our contract with CDC specifies that they will install any FCO, but we must supply the FCO and parts. What do others in the same boat do? Seton ================================================================================ Note 62.1 FCOs for Self/Third Party Maint. 1 of 14 EISNER::KENNEDY "Terry Kennedy" 18 lines 15-OCT-1987 18:40 -< Consider carefully first >- -------------------------------------------------------------------------------- First, if you don't have it already, get a copy of DEC's Self Maintenance Services catalog. Next, get a subscription to the update fiche. This used to be called DEC-o-log, I don't know if it still is. You may want to carefully consider the economics of this before proceed- ing, though. I would not take any service contract that required me to pro- vide the FCO kits, for the following two reasons: 1) FCO kits are produced in limited numbers, are frequently only avail- able by parts demand quote (taks 1 month to get the quote, 3-6 months to get the parts, must be prepaid), and are allocated by preference to DEC Field Service. 2) The cost of FCO's will possibly outstrip any savings on the contract itself - consider the RA81 swap for 12 RA81's as a small example. Also, there have been FCO kits for the 8600 listed in the SMS book for well over $100,000.00. I don't know what it is for, but that is a *big* lump. ================================================================================ Note 62.2 FCOs for Self/Third Party Maint. 2 of 14 EISNER::HASSINGER "Bob Hassinger" 5 lines 16-OCT-1987 01:09 -< Remember the famous 750 FCO as another example... >- -------------------------------------------------------------------------------- I asked about this when people were trying to sell my boss on going third party. After we reviewed the FCO on our 750 that replaced half the CPU boards things quieted down rather quickly. The third parties we were talking to did not have good answers. Basically they wanted us to take the risk on this I think. ================================================================================ Note 62.3 FCOs for Self/Third Party Maint. 3 of 14 EISNER::PROVOST 49 lines 20-OCT-1987 09:26 -< materials available, but... >- -------------------------------------------------------------------------------- >>> < Note 62.0 by EISNER::DROPPERS "Seton R. Droppers" > >>> -< FCOs for Self/Third Party Maint. >- >>> We are moving to third party maintanance of our VAXcluster consiting >>> of 2x8650, 2x785, 6xRA81, 3XRA60, 2xHSC50, 1xSA482, 1xMICROVAXII. >>> The one thing that we have questions about is how to obtain FCOs >>> once DEC is no longer servicing the cluster? Our contract with >>> CDC specifies that they will install any FCO, but we must supply >>> the FCO and parts. What do others in the same boat do? We encountered the same problem. We sent out 2 RFP's for maintenance, one on equipment here at the Bates Linear Accelerator Center, and one on equipment located at MIT in Cambridge. The Cambridge bid specified as a requirement that the vendor SUPPLY and install the FCOs. CDC conformed to that requirement, as dod some other companies. For the Bates bid, the supplying was not specified in the bid, since we assumed that requiring them to install the FCOs implied they would provide them. The response indicated we would have to supply the FCOs. There were several other items in the CDC contract to which we took exception. CDC retracted all the objectionable items, and agreed to provide the FCO's if and when they could get them. By the time we received this retraction we had already been forced to a decision, so we elected to use McDonnell Douglas. McDonnell Douglas is a major OEM of Digital Equipment, so the have ready access to FCO parts. CDC is at the mercy of Digital's delivery system, which is not as good for outsiders as it is for their OEMs. I am not necessarily recommending Mc D, but wish to point out two things. First, a careful phrasing of your RFP can eliminate, or at least reduce, this kind of problem. Second, since maintenance is a lucretive and highly competitive business, the vendors are often willing to negotiate these points. To more directly answer your question, Digital will provide the FCO materials to anyone who owns the equipment. That includes you. Some of these FCO kits are free of charge. It depends on the severity of the problem, whether saftey is compromised, and whether the FCO includes enhancements to the hardware (e.g. RA81 glue). Obtaining the material does not solve all your problems. Take, for example, a recent DEQNA FCO, which involves a board swap. Assuming you are the type of user who doesn't want to lay hands on your hardware, you must first have CDC de-install the old board. You must then bring or send it to Digital for replacement. You must then have CDC install the replacement. This does little for up-time on your machine. That is not acceptable to us, so CDC backed down... although too late. Tom [:) (with user hat) ================================================================================ Note 62.4 FCOs for Self/Third Party Maint. 4 of 14 EISNER::PROVOST 17 lines 22-OCT-1987 16:59 -< RETRACT McDonnell Douglas comment >- -------------------------------------------------------------------------------- > < Note 62.3 by EISNER::PROVOST > > -< materials available, but... >- Update on our maintenance contract situation: We chose McDonnell Douglas as specified in previous note. After three weeks McD. D. found out they were not going to get the contract for the other system mentioned in the previous note. They then retracted their bid!!! Since paper processing was delayed by the DOE tape machine, they can probably get away with it. I suspect we will now go with CDC with the modified contract... if they'll still have us. Tom [:) (with user hat) ================================================================================ Note 62.5 FCOs for Self/Third Party Maint. 5 of 14 EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 15 lines 22-OCT-1987 21:23 -< Any others with CDC?\ >- -------------------------------------------------------------------------------- Sounds like we are going to go with CDC as well, its nice to know that we are not going to be the ONLY CDC contract in the country. More to the point: If all the DEC customers that are doing their "own" maintenance get together, maybe we can point out to DEC that they can still make "some" money off of us by treating us with respect and setting up a realistic method of distributing FCO's to the non-DEC serviced customers. I am NOT suggesting that DEC should charge for all FCO, or that we would be willing to pay for all FCOs, what I am saying is that I expect my company would be willing to pay a reasonable amount for a guaranteed method of getting FCOs is a reasonable time... Seton ================================================================================ Note 62.6 FCOs for Self/Third Party Maint. 6 of 14 EISNER::KILLEEN "Jeff Killeen" 14 lines 22-OCT-1987 21:47 -< FCO'S - CDC - AND EMULEX >- -------------------------------------------------------------------------------- We have 30 systems under maintenance with CDC. There is another FCO problem. It involves third party products - namely Emulex. DEC is honest when it issues FCO's and will list them as required if it is to correct a design defect. Emulex lists *ALL* of there FCO's as optional - and will only supply them during the first year of ownership. I have stopped buying Emulex products because of this HIGHLY QUESTIONABLE AND TOTALLY DISHONEST practice by Emulex. I you want to check this out ask an Emulex Salesman to give you in writting whats happens after the first year if a problem is found with the controller under a new version of the OS. Now why is this of interest to CDC customers? The CDC contract states they will install all *REQUIRED* FCO's. I had one local office refuse to install a TC02 FCO because Emulex lists all FCO's as optional. Clearly the problem is Emulex's - however it can cause problems with CDC. ================================================================================ Note 62.7 FCOs for Self/Third Party Maint. 7 of 14 EISNER::HASSINGER "Bob Hassinger" 20 lines 23-OCT-1987 00:57 -< CDC and Emulex - #%&*@*%*!!! >- -------------------------------------------------------------------------------- I suspect this is the same CDC that used to have a contract on our Emulex SC7000 controller and CDC 9771 disk. They did a very poor job of supporting it. When there were significant problems we had to wait until they could get a team of three different people assembled from various locations to be on site at the same time. One knew a little about the disk drive, one knew a something (not much) about the controller and the third was the VAX/VMS specialist (i.e. he knew how to print out error logs)! They took about three days to replace the HDA when we had a head crash. Recently we finally got DEC to take the controller and drive under our system contract. Now my regular account rep is responsible for them too. He seems to have the backup in parts and specialist that he needs. When we had another head crash DEC replaced the HDA the same day! By the way, it may be a long time before I buy anything else from Emulex after the way they treated us on the SC7000 - but that's another story... Bob H ================================================================================ Note 62.8 FCOs for Self/Third Party Maint. 8 of 14 EISNER::RICE "Gary Rice" 18 lines 24-OCT-1987 11:15 -< Once (twice, thrice) burned, now shy >- -------------------------------------------------------------------------------- >< Note 62.7 by EISNER::HASSINGER "Bob Hassinger" > > -< CDC and Emulex - #%&*@*%*!!! >- > By the way, it may be a long time before I buy anything else from > Emulex after the way they treated us on the SC7000 - but that's another > story... > Bob H We have "another story" also regarding Emulex service that will preclude any further purchase or use of them as a supplier as well. Too bad. They USED to be a good company. Gary ================================================================================ Note 62.9 FCOs for Self/Third Party Maint. 9 of 14 EISNER::KENNEDY "Terry Kennedy" 5 lines 24-OCT-1987 20:43 -< Shy? Aye! >- -------------------------------------------------------------------------------- > We have "another story" also regarding Emulex service that will preclude > any further purchase or use of them as a supplier as well. Same here - we were bit on both the SC31 and the TC02 products. Perhaps we need a new topic - "Emulex problems and what to do about them". ================================================================================ Note 62.10 FCOs for Self/Third Party Maint. 10 of 14 EISNER::KILLEEN "Jeff Killeen" 4 lines 24-OCT-1987 21:53 -< EMULEX - NEVER AGAIN! >- -------------------------------------------------------------------------------- >Same here - we were bit on both the SC31 and the TC02 products. Perhaps we >need a new topic - "Emulex problems and what to do about them". NO! - We only have an RA81 to store the notes on ;-} ================================================================================ Note 62.11 FCOs for Self/Third Party Maint. 11 of 14 EISNER::CONROY 43 lines 7-JUL-1988 16:32 -< How to have your cake and eat it too >- -------------------------------------------------------------------------------- It's too bad about Emulex - for the longest time they were the greatest! Not that I can say we've had any problems with them lately, but from all the messages I've seen, they must really have taken a turn for the worst. However, I can still say that System Industries is worse. They do not support third party maintenance companies (they want the maintenance themselves) and the throughput of their equipment is less than equivalent Emulex parts. I will never buy SI, and am trying to get rid of what we have! We have CDC maintenance also. The local office tells me that there are three different kinds of FCOs from DEC. The mandatory ones they provide from free to anyone who owns the equipment (or so I've heard). CDC will apply these FCOs at no cost. There are recommended FCOs which deal with performance issues and both CDC and DEC charge for those. There is a third, and less important, category which I don't even remember. I've noticed that you really have to keep up on the maintenance people to make sure that everything is up to spec. They have, though, often informed us of the non-mandatory FCOs. But they part numbers they give us are un-recognizable to DEC-Direct. Does anyone know where to get the proper part numbers (E-Store didn't help) - and once I have those, where do I call to get them? Incidentally, I have been dealing with third-party maintenance companies for a long time and my opinion is that there are enough companies in the business that one could successfully get what one wants by threatening to move to a competitor. Of course, one can go overboard and be unrealistic, but from what I've seen, most people feel they are at the mercy of their maintenance company. It need not be so! You are the customer - YOU are paying THEM. Make demands and threats (and be prepared to back them up). When I came to Timeline four months ago, the situation with CDC here was deplorable. We had a couple meetings, during which I told them point-blank to get their act together or get out. Guess what? They got their act together. I was a little disappointed because I wanted to move to another company, and I'm soft-hearted enough to leave them be if they are doing a satisfactory job. But the important thing is that they did what we told them to (which included buying some equipment - like 11/750s - and using them as hot spares here in Bellevue/Seattle). What I'm trying to say is that you are really in control - so take control! Don't let them push you around. If they don't shape up, get another maintenance company. The CDC people here don't like me, but they want our business so they do what I ask. They realize that I'm still looking for a good excuse to get rid of them... Alan ================================================================================ Note 62.12 FCOs for Self/Third Party Maint. 12 of 14 EISNER::KENNEDY "Terry Kennedy" 24 lines 7-JUL-1988 16:49 -< How to get that info >- -------------------------------------------------------------------------------- The FCO categories are Mandatory, Required, Specification and Improvement, in order of most to least important. The DEC numbers (from the Self-Maintenance Services Catalog) are 603-884-5000 for Part Number Assistance (ex: What is the part number for the keyswitch on an 11/44) and 603-884-5001 for Unlisted Parts. Unlisted parts will ask you for the part number and quantity, and will then prepare a quote and mail it to you. If you want the items, return the quote signed. Beware - this takes a rather long time to get you the parts. By the way, the people at -5000 are the *most* helpful I've ever met at DEC - I've had them drop everything just to get me a number for a 75-cent part. Orderable from DECDirect is 'All-in-All DEC-O-Log', which is an annual fiche subscription of FCO's as they are announced. It is *not* a history (you can't use it to check rev's on your existing boards) but it *does* inform you of FCO's as they come out. I think it's under $500.00 for the initial 1-year sub. Lastly, when you re-negotiate the contract, you might be able to get a better deal. My contract (with a different company) says that they will supply and install *all* FCO's at no cost within 60 days of announce via DEC-O-Log, and that they pay for my DEC-O-Log subscription. THis is very important as anyone with an RA81 not on service found out. When the glue problem developed, the FCO was not Mandatory or Required, so you didn't get it free if you had no contract, or if you had a 'vanilla' third-party contract. ================================================================================ Note 62.13 FCOs for Self/Third Party Maint. 13 of 14 EISNER::CONROY "Alan Conroy" 1 line 18-JUL-1988 14:24 -< Thanks >- -------------------------------------------------------------------------------- Thanks! I'm going to call them right now! ================================================================================ Note 62.14 FCOs for Self/Third Party Maint. 14 of 14 EISNER::MERRIMAN "George Merriman - CCA/NY" 2 lines 2-OCT-1988 17:33 -< for further reading . . . >- -------------------------------------------------------------------------------- See topic 24 in SITE_MANAGEMENT for more discussion of third party hardware maintenance.  ================================================================================ Note 63.0 RA60 Data recovery? 4 replies EISNER::NORTON 8 lines 16-OCT-1987 12:26 -------------------------------------------------------------------------------- An RA60 logged a lot of "Uncorrectable ECC errors", then write-protected itself. When spun down and into mount-verify-timeout status, the subsequent mount elicited "Unrecoverable I/O error in RCT". 1. Does anyone have a method to recover the data? 2. How about a way to reformat the pack, so we can at least re-use the media? ================================================================================ Note 63.1 RA60 Data recovery? 1 of 4 EISNER::NORTON 8 lines 16-OCT-1987 16:29 -< Further developments... >- -------------------------------------------------------------------------------- Later update... Get this - the drive is corrupting the RCT by itself when a pack is spun up, EVEN AFTER all the drive's cards have been swapped. What else is left to change besides the power supply and heads? Ideas? ================================================================================ Note 63.2 RA60 Data recovery? 2 of 4 EISNER::DROPPERS "Seton R. Droppers" 8 lines 17-OCT-1987 10:37 -< The control panel pc card >- -------------------------------------------------------------------------------- > What else is left to change besides the power supply and heads? We had an RA60 that would never ready the heads (never inserted them, thus never finished spinning up). Field service swapped first the CONTOL PANEL card (swapping a ROM when the did), and then the complete drive... Only swapping the drive helped. Seton ================================================================================ Note 63.3 RA60 Data recovery? 3 of 4 EISNER::NAGY "Frank J. Nagy, VAX Guru" 5 lines 17-OCT-1987 13:50 -< RA60 Packs warranty (?) >- -------------------------------------------------------------------------------- To the best of my knowledge, rememberance, etc... If you bought your RA60 packs from Digital, they come with a lifetime warranty. You should be able to send your pack back to Digital for a replacement or reformatting. ================================================================================ Note 63.4 RA60 Data recovery? 4 of 4 EISNER::PRIGOT "Jonathan M. Prigot" 6 lines 19-OCT-1987 09:45 -< Send a cable >- -------------------------------------------------------------------------------- Re:.1 (What else can we change?) The ribbon cables! We had an RA81 go slowly flakey on us. F.S. came in and (all together now:) changed everything from the boards to the HDA with no effect. After she changed the ribbon cable from the HDA's heads to its amp board, the problem went away.  ================================================================================ Note 64.0 RQXD3, but no manual 2 replies EISNER::STAMERJOHN "RW Stamerjohn" 4 lines 16-OCT-1987 13:09 -------------------------------------------------------------------------------- We just bought a RQXD3/RD53 to replace a RQXD1/RD52 on our MicroVAX I. The RQXD3 came without a manual and we cannot get the RX50 to boot. Does anyone have ideas? Would anyone with a manual let me call you to check our jumpers? ================================================================================ Note 64.1 RQXD3, but no manual 1 of 2 EISNER::KILLEEN "Jeff Killeen" 17 lines 16-OCT-1987 13:35 -< QUESTIONS >- -------------------------------------------------------------------------------- I. The RQXD3 came without a manual and we cannot get the RX50 to Need to know more info. 1. BA23 or BA123 config 2. How many RD drives 3. How remote drives 4. Are you using a RQDX1-E Call me if you need quick help 617 478-8098 ================================================================================ Note 64.2 RQXD3, but no manual 2 of 2 EISNER::KENNEDY "Terry Kennedy" 12 lines 16-OCT-1987 20:34 -< It's a bug! >- -------------------------------------------------------------------------------- > The RQXD3 came without a manual and we cannot get the RX50 to ... This is a problem in DEC's packing facility. The same problem happened to me about a month ago. If you ordered it through DEC Direct, you should call them up and tell them to check - the BOM shows a manual should be included. They will take the DEC number of your order and send you one free. However, it will take 3-4 weeks. If you ordered it through a local DEC sales office, call your sales rep and tell him/her that you got a 'short ship'. They then can order the manual for you. I suggest you call Jeff for the information you need (that's what I did).  ================================================================================ Note 65.0 6 MIPS + 6 MIPS = ? 2 replies EISNER::WALLIS "Barry Wallis - Fleetwood Enterprises" 9 lines 16-OCT-1987 16:05 -------------------------------------------------------------------------------- My company currently has a VAX-8600, 19 VAX-11/730s, and 2 VAX-11/750s (everything but the 8600 is scattered in plants around the country). We are looking into centralizing our MIS operation into a single site with a VAX cluster. Given that the cluster is tuned optimally (whatever that means :-}) and that file sharing is also optimal; how much performance (my management thinks in terms of MIPS) will I really get if we cluster two 6 MIPS machines (say VAX-8550s)? ================================================================================ Note 65.1 6 MIPS + 6 MIPS = ? 1 of 2 EISNER::DROPPERS "Seton R. Droppers" 24 lines 17-OCT-1987 10:28 -< Article in VAXcluster Quorum, Aug '87 >- -------------------------------------------------------------------------------- { It turns out that there is an article in the latest "VAXcluster Systems Quorum" (Vol. 3, Issue 1, August 1987, published by DEC Software Services Technical Services Group, you might ask your account rep for a copy) on cluster performance. The article is about 8 pages, including extensive graphs and descriptions of the test environment. I will extract the overview information they gave from the article (From Table 1, page 33): Disclaimer: (From Seton) This information is for provided for reference ONLY. It reflects a specific TEST environment, you should read the complete article for additional information and disclaimers. Number of Nodes Max Throughput Num. Users Rel. Performance 1 35850 240 1.00 2 62010 480 1.73 3 89172 720 2.47 4 111972 880 3.12 The test environment was a transaction processing simulation, throughput was the number of "transaction exchanges" processed systemwide in an hour. ================================================================================ Note 65.2 6 MIPS + 6 MIPS = ? 2 of 2 EISNER::WALLIS "Barry Wallis - Fleetwood Enterprise" 12 lines 20-OCT-1987 10:24 -< 2(8650) + 2(8650) = 3.12(8650) >- -------------------------------------------------------------------------------- > It turns out that there is an article in the latest "VAXcluster Systems > Quorum" (Vol. 3, Issue 1, August 1987, published by DEC Software > Services Technical Services Group, you might ask your account rep for a > copy) on cluster performance. The article is about 8 pages, including > extensive graphs and descriptions of the test environment. I have just finished the article and agree that it suits my needs exceptionally well. Do you (that's the plural and includes anyone out there with cluster experience) think the results can be generalized to most/all of the processors in a VAXcluster regardless of size? In other words, if I cluster 4 VAX-8350s will I get the performance equivalent of approximately 3.12 8350  ================================================================================ Note 66.0 85xx Boot problem 1 reply EISNER::NORTON 34 lines 20-OCT-1987 10:32 -------------------------------------------------------------------------------- Imagine, if you will, your brand new 8530 sitting idle after F.S. checkout and release, because it WON'T BOOT using the supplied BCIBOO.COM file on the console disk. We followed all the comments in the file about what to deposit into which register, and no go. FS hadn't a clue either. Enter one of our people who had installed an 8500 last February. "Oh yeah, the documentation is wrong (again)!" The offending line of the file reads: DEPOSIT R1 00 ! Boot device bus address: ! <3:0>=BI node #, <5:4>=BI # It turns out that an 85xx is half of an 8800, which has BI #'s 0, 1, 2, and 3. The 85xx is the LEFT half, which has BI #'s 2 and 3. So - the first BI on an 85xx is number 2, not 0. Changing the line to read as follows corrected the problem: DEPOSIT R1 20 ! Boot device bus address: ! <3:0>=BI node #, <5:4>=BI # ! (BI #'s start at 2 for 85xx) * FLAME ON * These systems have been shipping for long enough that you'd think this error would have been corrected by now, or at least some sort of cover letter generated, or at least ALL FS branches notified! * FLAME OFF * Once again, thank you for your support. ================================================================================ Note 66.1 85xx Boot problem 1 of 1 EISNER::RENES 11 lines 22-NOV-1987 17:31 -< VAX 8500 -8530 performance myth? >- -------------------------------------------------------------------------------- In a slightly different vein, have you been able to show a 30% increase in performance over an 8500? My benchmarks, run on 750, 780, 785, 8200, 8530 and 8700 machines show the 8530 to be only 3x of an 8200 or 780, and NOT 4x which DEC claims for ALL of it's benchmarks (and some at 4.5X!!). DEC charges abt $60k for an 8500 - 8530 upgrade???? COmments?  ================================================================================ Note 67.0 VT125 Info 5 replies EISNER::MCDOUGALL "Bob McDougall" 8 lines 21-OCT-1987 23:41 -------------------------------------------------------------------------------- Our site recently in-herited a VT125 without documentation. I have a couple questions: 1. It has a swivel base with a male keyboard jack on a 6 inch wire coming out of the base, what is this wire for? 2. Does anyone have the order numbers for VT125 manauls? 3. Better yet does anyone have an extra manual they might be willing to part with? Thanks -Bob ================================================================================ Note 67.1 VT125 Info 1 of 5 EISNER::KENNEDY "Terry Kennedy" 10 lines 22-OCT-1987 02:38 -< Yes, there really is a system... >- -------------------------------------------------------------------------------- > 2. Does anyone have the order numbers for VT125 manauls? Generally, the order numbers for a thingus with a 5-letter model number is EK-thingus-UG for the user's guide, EK-thingus-TM for the tech manual if present. For a four-letter model number (obscene?), it's usually EK-0thingus-UG, although it may sometimes be EK-thingus[blank]-UG or EK-thingusA-UG. For the printset number, call 1-603-884-5000 and tell them you need the printset for a thingus. They can also be used as a last resort for manual part numbers, too. A good reference is the Self-Maintenance Services catalog - call 1-800-DIGITAL and ask for one to be sent to you. ================================================================================ Note 67.2 VT125 Info 2 of 5 EISNER::FRISBIE "Alan E. Frisbie" 51 lines 22-OCT-1987 02:39 -< VT125 Info >- -------------------------------------------------------------------------------- >> Our site recently in-herited a VT125 without documentation. >> I have a couple questions: >> 1. It has a swivel base with a male keyboard jack on a 6 inch >> wire coming out of the base, what is this wire for? The plug (not jack) on the end of the wire plugs into the keyboard jack on the back of the VT-125. The keyboard than plugs into the jack on the right side of the tilt/swivel base (very near the front). This gives you somewhat more freedom of movement with the keyboard. (Remember: JACKS are female connectors, PLUGS are male connectors.) >> 2. Does anyone have the order numbers for VT125 manauls? The following are from page 11-5 of the VT125 User's Guide: EK-VT125-UG VT125 User Guide -- Describes the installation, operation and programming of the VT125 terminal. EK-VT100-J1 VT100 Series Pocket Service Guide -- Describes procedures used to troubleshoot and repair the VT125 to the module level. EK-VT100-TM VT100 Series Technical Manual -- Describes the VT125 terminal to a detailed block level. It provides troubleshooting information for the terminal, but does not contain detailed schematic drawings. EK-VT125-IP VT125 Video Terminal IPB -- Describes a detailed parts breakdown of the VT125 field replacable units. EK-VT125-RC VT125 Terminal Programming Reference Card -- Provides a summary of the VT125 ReGIS escape and control sequences on a pocket size reference card. MP-01053-00 VT125 Field Maintenance Print Set -- Provides a complete set of electrical and mechanical schematic diagrams for the VT125. >> 3. Better yet does anyone have an extra manual they might be >> willing to part with? If DEC will not sell you the User's Guide (EK-VT125-UG), I will loan you mine to copy. Sorry, but I cannot give up my only one. Anyone else have one? Alan Frisbie ================================================================================ Note 67.3 VT125 Info 3 of 5 EISNER::MCDOUGALL "Bob McDougall" 7 lines 4-NOV-1987 18:46 -< VT125 EIA PORT->printer >- -------------------------------------------------------------------------------- I now have a VT125 USER GUIDE: EK-VT125-UG-001. The next trick I'm trying to perform is to drive a DECWRITER IV with the EIA AUXILIARY PORT. The DECWRITER runs at 110 and 300 baud. I've connected it with a DEC BCC14-10 cable and have tried numerous configerations of control sequences and other cables with no success. If everything is configured properly how do you send information to the printer? How do you change the speed on the EIA port. ================================================================================ Note 67.4 VT125 Info 4 of 5 EISNER::DOHERTY "Bob Doherty" 15 lines 4-NOV-1987 22:24 -< Using VT125 Printer port >- -------------------------------------------------------------------------------- < If everything is configured properly how do you send information> < to the printer? How do you change the speed on the EIA port.> The escape sequence to start the printer is [5i and [4i turns it off. The only way to set the speed of the printer port on a VT125 is through the terminal setup. The choice of speeds is quite limited and they are selected by bit patterns and not by the method of setting the rate of the terminal itself. The method of dumping the screen involves REGis commands which I don't remember offhand but which are in the book you cited. Bob. ================================================================================ Note 67.5 VT125 Info 5 of 5 EISNER::ERSKINE "Chris Erskine" 2 lines 5-NOV-1987 06:53 -< Do I need to know there is a printer? >- -------------------------------------------------------------------------------- Does the 125 require the DSR signal like the VT2xx terminals to know that there is a printer connected?  ================================================================================ Note 68.0 Need RA70 help for LAVC 4 replies EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" 19 lines 28-OCT-1987 17:11 -------------------------------------------------------------------------------- I'M IN THE process of piecing together a uVAX II based LAVC. My question concerns the newly announced RA70 disk drives. As I understand it there is a system called the 3602 which is a dual processor (not in the same box) 3600 based system which is configured for use in an LAVC. It uses the dual porting capability of the RA series disks to allow the reboot of the cluster should the boot node fail. Has anyone had any experience with the RA70 used with something other than a 36xx system? It seems like a good idea to me to use it as the boot device on my uVAX II based LAVC. I have access to a few uVAX II's which I'd like to include in my cluster. The kicker is I don't have an computer room environment where this cluster resides. This would preclude the use of the large format RA series of disk drives. I'd also like to keep the disks in the BA123 boxes. I currently use the EMS-380 disk subsystems from EMULEX as my large (380 Mb) drives. What does this device use as a controller? Would a KDA-50 work? ================================================================================ Note 68.1 Need RA70 help for LAVC 1 of 4 EISNER::KILLEEN "Jeff Killeen" 3 lines 28-OCT-1987 17:23 -< KDA50 TO RA70 >- -------------------------------------------------------------------------------- > What does this device use as a controller? Would a KDA-50 work? There are no current cabinet kits in the price list for the BA123 or BA23. ================================================================================ Note 68.2 Need RA70 help for LAVC 2 of 4 EISNER::RICE "Gary Rice" 16 lines 29-OCT-1987 11:00 -< NIX on RA70 >- -------------------------------------------------------------------------------- According to the people I talked to at DECworld about the RA70, it has cooling and power requirements that preclude it's use in the BA23 or BA123 packages. I don't know enough about the "Q5" packaging to comment on the capacity of it's power supply (supplies) to handle the RA70, but it would seem VERY likely that cooling wouldn't be a problem. Regardless, DEC isn't selling the RA70 for MVII's, so you would have to kludge a set of mounting hardware, cables, etc. yourself. Also, it would be questionable whether or not that what you did would be supported by field service. Gary ================================================================================ Note 68.3 Need RA70 help for LAVC 3 of 4 EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" 4 lines 29-OCT-1987 11:55 -< Still need LAVC fail-soft answers >- -------------------------------------------------------------------------------- Well, then, anyone have a suggestion about how to make an LAVC with a uVAX II boot node fail-soft ? And without the computer room requirement? Dual-ported disks seem to be the answer to me, but what disks/controllers could I use ? ================================================================================ Note 68.4 Need RA70 help for LAVC 4 of 4 EISNER::KILLEEN "Jeff Killeen" 5 lines 29-OCT-1987 12:17 -< RIGHT DRIVE - WRONG TIME >- -------------------------------------------------------------------------------- I don't know when you need this system - but it is reasonable to assume at some point in time DEC will come out with an external RA70. Something like an RA70-RA or RA70-DA. I think you had the right drive at the wrong time.  ================================================================================ Note 69.0 DEPCA information sought 7 replies EISNER::TILLMAN "Brian Tillman, Smiths Industries." 4 lines 29-OCT-1987 12:42 -------------------------------------------------------------------------------- Is it REALLY true that a PC with a DEPCA in it can't use extended memory? That seems a little restrictive to me. We've been trying the DEPCA and it works fine except for that. Anyone know any work-arounds? ================================================================================ Note 69.1 DEPCA information sought 1 of 7 EISNER::GARDNER "Tim Gardner" 10 lines 29-OCT-1987 17:51 -< Untrue >- -------------------------------------------------------------------------------- > Is it REALLY true that a PC with a DEPCA in it can't use extended > memory? That seems a little restrictive to me. We've been trying > the DEPCA and it works fine except for that. Anyone know any > work-arounds? Not true; ours works just fine. Do you perhaps mean EXPANDED, rather than EXTENDED memory? If so, there could be an addressing conflict. T. ================================================================================ Note 69.2 DEPCA information sought 2 of 7 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 9 lines 2-NOV-1987 09:46 -< A rose is a rose... >- -------------------------------------------------------------------------------- RE: < Note 69.1 by EISNER::GARDNER "Tim Gardner" > -< Untrue >- | Not true; ours works just fine. Do you perhaps mean EXPANDED, rather | than EXTENDED memory? If so, there could be an addressing conflict. Expanded, extended, what's the difference? All I know is that we have 1 Meg of memory on a PC and can't use more than 640K with the DEPCA card. ================================================================================ Note 69.3 DEPCA information sought 3 of 7 EISNER::GARDNER "Tim Gardner" 56 lines 9-NOV-1987 16:30 -< Rose != Rose != Rose >- -------------------------------------------------------------------------------- >> < Note 69.2 by EISNER::TILLMAN "Brian Tillman, Smiths Industries." > >> -< A rose is a rose... >- >> >>RE: < Note 69.1 by EISNER::GARDNER "Tim Gardner" > >> -< Untrue >- >> >>| Not true; ours works just fine. Do you perhaps mean EXPANDED, rather >>| than EXTENDED memory? If so, there could be an addressing conflict. >> >> Expanded, extended, what's the difference? All I know is that we >> have 1 Meg of memory on a PC and can't use more than 640K with the >> DEPCA card. >> Sorry for the delay in responding, but I spent the week at COMDEX (yech!). Extended memory is the physical memory above 1Mb on a 80286 or 80386 processor (as implemented by IBM). This memory is essentially wasted in an IBM enviornment. IBM chose to map ONLY 640Kb as usable memory. The remaining space up to 1Mb is reserved for I/O devices, ROM, etc. (much like the PDP-11 I/O Page). Any memory above 1Mb is not addressable in a 286 machine simply because the address lines are not enabled when the 286 is emulating an 8086 (or 8088). In addition to the hardware limitations, generic MS-DOS and IBM PC-DOS (yes Virginia, there is a difference) only address 640K directly. In order to use any additional memory (above 1Mb) special programs (i.e. VDISK) switch the CPU into protected mode (which enables full 24 bit addressing) and perform their magic and switch back to real mode. In order to get around all this, Lotus, Microsoft and Intel (LIM) got together and published the Expanded Memory Specification (EMS). This spec has been adopted by a number of companies and a vast array of boards now support LIM-EMS. The secret to EMS is that they use one (or two, I forget which) of the 64K banks of memory between 640K and 1Mb, then page switch around to achieve several megabytes of physical storage. The problem is that any program that wants to take advantage of the memory, must specifically know how to use it; DOS doesn't know its there. To further confuse matters, AST and Quadram (and someone else, I forget who) decided that LIM-EMS just fine for data storage, but useless for programs (quite true, but that is what LIM wanted...), so they defined a superset of LIM-EMS called Enhanced Expanded Memory Specification (EEMS). So we now have THREE memory models, Extended (80286/386), Expanded (LIM-EMS) and Enhanced Expanded (EEMS). Which you choose depends on your application (note that is is possible to have both extended and expanded in the same machine). Now to address your question: If you 1Mb is on the motherboard, it is almost certainly set up as extended memory. If it is on an expansion board, it is anybody's guess. It could even be a combination of regular (<640K), extended and expanded. The problem I mentioned originally (the conflict with the DEPCA) is that expanded memory uses an I/O port, and this may be conflicting with the DEPCA. I hope this helps, but if you need more info, let me know. Tim. ================================================================================ Note 69.4 DEPCA information sought 4 of 7 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 2 lines 10-NOV-1987 09:28 -< So, what do I do about it? >- -------------------------------------------------------------------------------- Whatever the reason, how do I determine why it won't work and how do I fix it? ================================================================================ Note 69.5 DEPCA information sought 5 of 7 EISNER::KENNEDY "Terry Kennedy" 20 lines 10-NOV-1987 19:16 -< Suggestions... >- -------------------------------------------------------------------------------- > Whatever the reason, how do I determine why it won't work and how > do I fix it? There are probably only two areas of possible conflict - I/O address or memory address. The EMS spec states that boards shall be selectable to one of eight I/O addresses (that's how you get 8 Mb with 1Mb boards). Try changing the base address (you'll have to change the driver par- ameters in CONFIG.SYS as well - see the EMS software manual). If that doesn't help, it may be that the Ethernet card is memory-mapped in the same address space as the EMS board. This is probably not the case, because the EMS driver enables/disables the memory when it needs access, which would break EMS but not the Ethernet card. However, the EMS card can be re-configured for different memory addresses as well. Not very many are useful in 'normal' machine configurations, so you will have to experiment. Again, you will also need to change the driver info in CONFIG.SYS P.S. - The EMS specification, Version 4.0, incorporates EEMS and every- body is now happy. If you have an Intel Above Board, they will send you the V4 driver for free. ================================================================================ Note 69.6 DEPCA information sought 6 of 7 EISNER::BURNET "John Burnet" 11 lines 4-OCT-1988 12:58 -< How to find a DEPCA's rev level? >- -------------------------------------------------------------------------------- How can I find out what rev level (e.g., D or E) a DEPCA board is at? I have one sitting in front of me that has two bar-code stickers on it; one sticker has a nine-digit number and the other says "D1". On the reverse (solder) side of the board, above the etched "DEPCA" on the bottom, there is a number (also etched) of the form xx-xxxxx-xx, followed by "B1". I tend to think that the board is at level D1, since in addition to the sticker there's also a "D1" rubber-stamped onto the front of the board. Can anyone say for sure where to find the rev level on this puppy? Thanks. ================================================================================ Note 69.7 DEPCA information sought 7 of 7 EISNER::DELARISCH "Arnold S. De Larisch" 20 lines 4-OCT-1988 13:40 -< DEPCA REV number on Component side of the PCB! >- -------------------------------------------------------------------------------- >> < Note 69.6 by EISNER::BURNET "John Burnet" > >> -< How to find a DEPCA's rev level? >- >> How can I find out what rev level (e.g., D or E) a DEPCA board is >> at? I have one sitting in front of me that has two bar-code >> stickers on it; one sticker has a nine-digit number and the other >> says "D1". On the reverse (solder) side of the board, above the >> etched "DEPCA" on the bottom, there is a number (also etched) of >> the form xx-xxxxx-xx, followed by "B1". Your board is REV D1. According to the new "DEPCA Owners's Manual" the REV level is indicated on the component side of of the PC board just under the BAR Code Serial number (if memory serves me)! The Dn REV boards do not have the 32 Kbyte mode ... so if you've got a PC/AT w/EGA and an above board, you need to 'trade-up' to a rev E1 or latter board! -Arnold  ================================================================================ Note 70.0 Help needed w/ RD53 / RQDX2 combination 1 reply EISNER::KENNEDY "Terry Kennedy" 12 lines 30-OCT-1987 22:19 -------------------------------------------------------------------------------- Does anyone know what the incantation to XXDP is to format an RD53 for an RQDX2 on a Micro PDP-11? The drive was running fine on an RQDX3, but we need to move it to an RQDX2. The RQDX2 is at the current revision level (188 & 189 EPROMs). ZRQB?? says 'nonexistant unit number - pass aborted', but it does go out to the drive (access light flashes). Yes, it's jumpered as DS3, etc. (worked fine in same exact system except with RQDX3 controller) DEC says 'you need "E"'. The person I spoke to implied that he meant ZRQE??, not ZRQB rev E. I do not have either of these in my diagnostic collection. ================================================================================ Note 70.1 Help needed w/ RD53 / RQDX2 combination 1 of 1 EISNER::CETRON 10 lines 7-NOV-1987 18:00 -------------------------------------------------------------------------------- when going from rqdx2 to rqdx3 we had similar problems.... the solution: turn off system, set drive to not ready, boot system (xxdp) letting controller self-test, but NOT query drive (apparently), then start xxdp, then bring drive online to format.... -ed  ================================================================================ Note 71.0 Need help interfacing GPX to plotter 3 replies EISNER::RHODE "Chris Rhode" 12 lines 4-NOV-1987 18:31 -------------------------------------------------------------------------------- We have a group of people at our site that are interested in trying to interface a Tektronix Model 4692 Color Plotter to a VAXstation II GPX so that they can do screen copies of the GPX screen to the plotter. They have already successfully done this for a Masscomp and an IBM PC so they believe it should be possible. Can anybody offer any suggestions? Is this sort of thing really doable? Is there a technical reference manual for the VAXstation GPX graphics system that might be of help to them? advTHANKSance, -Chris ================================================================================ Note 71.1 Need help interfacing GPX to plotter 1 of 3 EISNER::CETRON 6 lines 7-NOV-1987 18:02 -< sixel to plotter convertor >- -------------------------------------------------------------------------------- it can be done if you can get a sixel to 4692 convertor... we use a sixel-to-postscript convertor from utexas and screen dump out to our laserwriters... -ed ================================================================================ Note 71.2 Need help interfacing GPX to plotter 2 of 3 EISNER::RICE "Gary Rice" 8 lines 8-NOV-1987 11:10 -------------------------------------------------------------------------------- > it can be done if you can get a sixel to 4692 convertor... we use > a sixel-to-postscript convertor from utexas and screen dump out > to our laserwriters... Is "utexas" the University of Texas? Is the converter software or hardware? Who can I contact about obtaining such a device/program? Gary ================================================================================ Note 71.3 Need help interfacing GPX to plotter 3 of 3 EISNER::CETRON 6 lines 9-NOV-1987 11:24 -< software only >- -------------------------------------------------------------------------------- convertor is pure software, give me a call and I will pass on the information.... 801-581-6499..... -ed  ================================================================================ Note 72.0 Datascopes and Protocol Analyzers 9 replies EISNER::BYRNE_C "Charlie Byrne - Reuters Info Svcs" 14 lines 6-NOV-1987 12:08 -------------------------------------------------------------------------------- I'm looking into getting a Datascope. Some requirements are: Async, bisynch, DDCMP, future possibly X.25 & SNA Portable Disk or tape storage Handle at least 19.2 Extensive trapping/break facilites Large buffer Filtering Error Injection Naturally as inexpensive as possible I've heard good things about HP4951C and HP4952A. Any comments on these or any other Datascopes? ================================================================================ Note 72.1 Datascopes and Protocol Analyzers 1 of 9 EISNER::PRIGOT "Jonathan M. Prigot" 13 lines 6-NOV-1987 12:43 -< 1 out of 10 recommend... >- -------------------------------------------------------------------------------- I use a Halcyon 801B. It is a nice unit that handles all the protocols that you mentioned with one omission -- Async DECnet which hadn't been released when this box came out. Since Halcyon got bought out soon after that, I don't think they'll be doing anything about it. (I still bring it up with my salesman each time, though.) It doesn't have disk or tape, but it does use battery-backed CMOS RAM for storage. It clocked in at $5K. Since then I have had periodic access to a Digilog unit (I forget the number, but I believe it is their entry-level product). It seems to meet all of your requirements, and would be my box of choice at this time. I think that the async DDCMP requirement will weed out a lot of boxes. ("DECnet? What's that?") ================================================================================ Note 72.2 Datascopes and Protocol Analyzers 2 of 9 EISNER::TINKELMAN "Bob Tinkelman - CCA" 12 lines 6-NOV-1987 16:03 -< Digilog and Asynch DECnet >- -------------------------------------------------------------------------------- > Since then I have had periodic access to a Digilog unit (I forget > the number, but I believe it is their entry-level product). It seems to > meet all of your requirements, and would be my box of choice at this > time. I think that the async DDCMP requirement will weed out a lot of > boxes. ("DECnet? What's that?") Digilog's entry level product is the Digilog 200. We bought one of the very early 600s several years ago. You are correct about asynch DDCMP. While the various "setup menus" seem to allow you to specify ASYNCH quite independently from DDCMP, specifying them both yeilds nothing useful. Of course, if anyone reading this has a Digilog which does decode asynch DDCMP, I'd be very interested to hear that. ================================================================================ Note 72.3 Datascopes and Protocol Analyzers 3 of 9 EISNER::LEDERMAN "Bart Z. Lederman" 28 lines 7-NOV-1987 18:41 -< Look into FELINE and PC >- -------------------------------------------------------------------------------- We have just about every kind of datascope ever invented here. The one I like best is, of all things, a card that fits into a (gasp!) IBM PC or equivalent. There are two: one is the FELINE unit made by Fredricks in Maryland. I've forgotten the name of the other, but it wasn't very good when we tested it (though they tell us they've improved it since then). The FELINE monitors a number of kinds of data lines, including X.25, BISYNC, HDLC, dumb ASCII or EBCDIC, etc. It decodes X.25 packets better than most other units I've seen. What really makes it usefull is the amount of data it can store: basically, as much as you can fit on your hard disk, and/or backup to floppies. You can then "replay" the session, moving forwards and backwards on the screen with arrow keys more easily than with otehr monitors I've seen. Also, when other monitors can store data, they usually do it on some non-standard tape cartridge, not standard 5 1/4" floppies. I've even been able to upload the data file to the VAX and process it there. The unit can be pre-programmed to trigger on events, will search for pattersns, and can be set up so that commands come in over a serial line so you can set it up and control it from a remote location. Besides all this, it costs a lot less than any other monitor I've seen that does anywhere near the amount of work. Somewhere between $1000 and $1500 for the unit, plus a PC (nowadays, probably under $1000). Other units that can decode high level protocols and store and analyze them run $15,000 to $50,000. ================================================================================ Note 72.4 Datascopes and Protocol Analyzers 4 of 9 EISNER::FRISBIE "Alan E. Frisbie" 12 lines 7-NOV-1987 23:14 -< Very simple requirements >- -------------------------------------------------------------------------------- Will the FELINE (or similar) unit record/display the relative character timings of two independent async data streams? I already know *what* data is being transmitted in each direction, I need to know *when* a response occurs. Also, will it get freaked out by a Break signal? The last one I used wouldn't re-sync for several characters. Most of my needs are for simple async ASCII, with DDCMP and Kermit desirable. Alan ================================================================================ Note 72.5 Datascopes and Protocol Analyzers 5 of 9 EISNER::LEDERMAN "Bart Z. Lederman" 8 lines 9-NOV-1987 08:57 -< It will to bi-directional data >- -------------------------------------------------------------------------------- The FELINE will record characters in both directions on the same line (i.e., send and receive) and will properly time-stamp every single character. It will display either direction by itself, or both streams together (on adjacent lines). Most of the other datascopes I've seen will display both streams in adjacent lines. I do not recall it's being upset by Break, but I'm not sure I ever really tried it. I've used it mostly on synchronous lines with external clock, but it does work on async lines with internal clock. ================================================================================ Note 72.6 Datascopes and Protocol Analyzers 6 of 9 EISNER::HANRAHAN "Jamie Hanrahan" 24 lines 10-NOV-1987 13:38 -< The HP 4853A isn't bad... >- -------------------------------------------------------------------------------- The nicest monitor we use here is an HP 4953A. I believe it's obsolete technology these days but it's been very useful. Bisync, HDLC, SDLC, X.25, DDCMP (async and sync), X.75 (???), and character async/sync. Yes, Alan, it has programmable timers and counters. You can set it to watch for a packet going one way, start a timer and bump a counter, watch for the ACK or NAK going the other way, and stop the timer when it's seen. After as many transactions as you like you look at the accumulated timers and counters and divide to get the average. (No way to record min and max, though.) It also records modem control signal states along with the data, and so can show you waht the modem lines were doing when the device was (or wasn't) sending or receiving. I have yet to see one that can run in a sort of "asynchronous mode"-- not as in async vs. sync, but wherein dead time on the line would be recorded in the memory. Then you could do more in the way of stimulus/response timing analysis. A fancier model -- I believe it's the HP 4951A -- doesn't know DDCMP but is programmable in a form of BASIC. You could program it to understand DDCMP, Kermit, International Morse code, or just about anything else. But it didn't have a "log idle time" option either. ================================================================================ Note 72.7 Datascopes and Protocol Analyzers 7 of 9 EISNER::BYRNE_C "Live, from Freeport, New York" 9 lines 17-NOV-1987 16:17 -< Yes Virginia, there is a Fredricks, Maryland >- -------------------------------------------------------------------------------- > < Note 72.3 by EISNER::LEDERMAN "Bart Z. Lederman" > > -< Look into FELINE and PC >- > unit made by Fredricks in Maryland. I've forgotten the name of Operator has no listing for a Feline company in Fredricks, Maryland. I must assume that I misinterpreted and the company name is Fredricks. Do you know where in Maryland I can reach them? ================================================================================ Note 72.8 Datascopes and Protocol Analyzers 8 of 9 EISNER::ROBITAILLE "Mike Robitaille" 10 lines 17-NOV-1987 18:45 -< < Could it be.... > >- -------------------------------------------------------------------------------- I have an instruction manual on my lap as I type this note for a real ancient piece of gear which is dated May 1977. It is written by: Fredrick Electronics Corporation Hayward Road, Post Office Box 502 Fredrick, Maryland 21701 (301) 662-5901 They *might* be the same guys...Hope this helps and not confuses... Mike R. ================================================================================ Note 72.9 Datascopes and Protocol Analyzers 9 of 9 EISNER::LEDERMAN "Bart Z. Lederman" 8 lines 18-NOV-1987 07:46 -< Address >- -------------------------------------------------------------------------------- The address of the company that makes the FELINE (according to my owner's manual) is: Frederick Engineering, Inc. 54 Cessna Court Gaithersburg, Maryland 20879 Don't have a telephone number.  ================================================================================ Note 73.0 DHU-11 problems and fixes 6 replies EISNER::LEDERMAN "Bart Z. Lederman" 63 lines 9-NOV-1987 09:13 -------------------------------------------------------------------------------- We recently added a DHU-11 to our 785, and it took Field Service over a day and a half to do it. We found out a few things. 1. If you are using a KMS-11 as a PSI interface, you must deliberately put it outside of the normal CSR and Vector assignments. I knew this, because it is clearly stated in the PSI release notes, and it had been properly assigned before. I told the field service engineer this when I printed out the current assignments in the morning, but he chose not to believe it for about 5 hours, untill someone in his support center convinced him of it. 2. SYSGEN has a CONFIGURE routine which is supposed to calculate the proper CSR and Vector assignments for a list of devices. As noted above, if you have a KMS-11 DON'T include it in the list when you type it in. 3. If you have a setup like mine (UNA Unibus Adaptor, DMF-32 multipurpose board, and DHU-11), the SYSGEN routine will give you the wrong Vector assignment for the DHU-11. It puts the Vector 10 past the DMF-32, when it should be 20 past. Example: it calculated the DHU-11 vector at 340, but when the board was installed at that vector and the system booted, AUTOCONFIGURE said that the board was at 350 (which was where SYSGEN wanted it to be, not where it actually was). The fix was to go back and change the Vector on the board to 350, after which things worked correctly. A lot of time was wasted trying to find out what was wrong with several new DHU-11s which would cause them to report the Vecor incorrectly, when in fact nothing was wrong except wishful thinking on the part of SYSGEN. 4. Modem control on the DHU-11 is different than on the first two lines of the DMF-32, even though all of the manuals says they work the same. Specifically, if the DHU-11 does not see carrier (pin-8) go high and stay high, then the line will disconnect. If you have a long login timeout set (something like TTY_TIMEOUT), then you can actually log into VMS and start to work before the line is dropped. The clue is that, when you do a SHOW TERM, it is still set to "No Remote" even though you are on a modem controlled line: if the line connected properly, it should always say "Remote" when you are logged in and "No Remote" when disconnected. The DMF-32 apparently does not check for carrier: if pins 5 and 6 are high it accepts the connection. The DHU-11 is actually more "correct", but it had me going in circles for a while because lines which worked on the DMF-32 didn't work on the DHU-11, and none of our lines going to old DH-11s on our PDP-11s showed this problem. Since the port selector we use doesn't supply pin 8, I just tied 5, 6, and 8 together, and this keeps the DHU-11 happy. Most "real" modems do supply pin 8, so you might not see this problem on a DHU-11, though you might see the reverse problem on the DMF-32 if it doesn't disconnect on a true loss of carrier. 4. Until we start an FCO conference, the following information was on my shipping document entitled ECO STATUS }iSub Unit C.S. REV. ETCH CK-DKU11-AD b1 H3029 A H325 B1 C BC05L-10 U exactly as you see it, including the lower case letter ================================================================================ Note 73.1 DHU-11 problems and fixes 1 of 6 EISNER::LEDERMAN "Bart Z. Lederman" 9 lines 9-NOV-1987 09:16 -< P.S. >- -------------------------------------------------------------------------------- }i}r Two things I forgot. 1. At first, field service told me that telephone support had told them that the DHU-11 can't be AUTOCONFIGUREd because of a bug in SYSGEN. Don't you believe it. Once the CSR and Vector assignments were finally correct, it AUTOCONFIGUREd exactly as it should, and has continued to do so for a week. 2. The ECO STATUS was d~Rlted 03-JUN-87. ================================================================================ Note 73.2 DHU-11 problems and fixes 2 of 6 EISNER::BYRNE_C "Charlie Byrne" 13 lines 17-JAN-1989 11:38 -< Need Asynch DMA board - DHU11 OK? >- -------------------------------------------------------------------------------- Speaking of the DHU11, I need an serial asynch DMA board for a PDP-11/70 Unibus running under M+. The DHU11 appears to fit the bill. However, the write up in the DEC Store says something odd (actualy the entire write-up is odd): "* Direct Memory Access on transmit relieves your CPU of the overhead associated with character interrupt devices" Nothing is mentioned about DMA on Rx. What's the story? Also, are there alternatives in 3rd party-land (e.g. Simpact). ================================================================================ Note 73.3 DHU-11 problems and fixes 3 of 6 EISNER::MERRIMAN "George Merriman - CCA/NY" 35 lines 17-JAN-1989 16:16 -< DHx11 input and output operation >- -------------------------------------------------------------------------------- > "* Direct Memory Access on transmit relieves your CPU of the overhead > associated with character interrupt devices" As I remember, the DHx11 devices do DMA transfers on transmit but use a character interrupt silo on receive. If you think about it, this makes sense for two reasons: 1. When driving a terminal the receive data rate is usually quite slow; and 2. The DH would need to know about the receive terminating character, etc. to do DMA on input. As for the silo, I remember from the good old days of RSX that the DEC driver makes little use of it unless the DH is being driven hard. What happens is that the DH signals interrupt whae a character (from ANY channel) gets to the bottom (output) of the silo. When the driver gets around to servicing the interrupt it attempts to unload all of the characters from the silo at interrupt (or maybe fork) level, so if any characters get backed up in the silo it saves the OS a few interrupt service cycles. One way to improve the performance of the DH family in non-terminal applications is to modify the driver to tell the DH to interrupt only after some N characters are in the silo so the interrupt routine always services multiple characters. You have to put some kind of time-out in the driver so it can check to see that the last few characters don't get stuck in the silo. > Also, are there alternatives in 3rd party-land (e.g. Simpact). A Simpact ICP can be programmed to understand a particular application protocol and detect terminating characters, etc. on input. They don't exactly do DMA; the memory on the board shows up in the CPU memory space. The on-board processor is the T-11 (last time I looked) and it uses the same serial line driver as the KMV. They also have some canned software for the board that can be used for all sorts of strange things. ================================================================================ Note 73.4 DHU-11 problems and fixes 4 of 6 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 98 lines 17-JAN-1989 16:40 -< DHx class options for 11s >- -------------------------------------------------------------------------------- > I need an serial asynch DMA board for a PDP-11/70 Unibus > running under M+. >... > Nothing is mentioned about DMA on Rx. What's the story? >... > Also, are there alternatives in 3rd party-land (e.g. Simpact). The classic port use was to type a little, and get a lot back. The little one typed had to be watched character by character for such things as escapes and various control characters. DMA makes sense on output, but not on input. It is often argued that short bursts of DMA can cost more CPU than simply taking the character interrupts for each character output, but the DMA WINS with big buffers. (do a show term here on Decuserve, and you will find you are NOT DMA!, I set it in my login.com file, but am not sure if I should...) Multi line MUXes have fairly deep input silos and one input interrupt will handle all characters in the silo from any lines active. There are MANY 3rd party DH units, and probably some DHU units, and I think at least one brand (Dilog?) has microcode for both and a dip switch declares which emulation to use. The DHU, unlike the DH (I think I am right about this), has an output silo that one can preload, and this is preferable to setting up DMA for short bursts. I doubt the M+ TT driver uses this, but if you see adds for DHV lookalikes that say DHU functionality, this is what the difference is. I think DEC's DHQ now does it too, to bring the Q units up to the U ones smarts. Emulex, Dilog, and ABLE all make DHs. An Emulex CS11 will do 64 ports (4 unit emulation) with full modem control with 1 hex U bus card and only one 34 wire cable to up to four remote 16 port panels up to 50' away. The remote panels are 'smart', and that 34 wire cable is actually a microprocessor bus. The input silo is twice as deep as DEC's, but a dip switch will limit it while you run diags... This one card (CS11) can be: DH, DMF32, DV, depending on emulation proms, and the same remote panels (where most of the cost $ are) are usable on several other controller cards they sell. The CS32 controller card can drive 8 of these CP34 panels for 132 ports, and DMF32 ASYNC ports is the only emulation. Note that these DMF32 versions have modem control on ALL ports, unlike D*C. The CS04 Q bus card does 64 ports of DHV emulation with the same panels. I mention the other emulations in answer to your 11/70 question, because what you purchase NOW you can recycle onto a VAX later for small $ change-over. If you wanted DMF32 for a vax and wanted to go > 50' to the panels there are even nicer options, but I'll stick to the 11/70. Able has cards that go to their Easyway twisted pair net and out to remote terminal boxes or even via a bridge to CASE/Rixon global class statmuxes at their ARQ2 inter node link level, and I assume Able does several DHx emulations. Dilog has a Q bus card that does many ports of DHV (several emulations), and comes out in T1 format and the distribution panels are Micom ones that use T1. This connection can also go directly into a big Micom switch and thence to the port panels. I havn't heard of a U version of that card, but you could ask, or maybe you plan an upgrade to an 11/83?... N.B. M+'s TTdriver MUST move all output to its own buffers for DMA, as one can't afford UMRs to map everyone's buffers. Years ago Bob Tinkelman had repeatedly suggested (at symposia sessions) to DEC that they ought to NOT bother with the move for Q bus systems, and I believe the TTdriver now is so optimised for certain cases, probably where the task is also non-checkpointable. Emulex and Dilog and Able also make emulations that just do one 16 port 1:1 emulation. You did not mention D/N, but I should point out that the DH is NOT supported as a D/N device, but the DHU is. I hope, but do not know, that D/N uses it in DMA mode where possible, or at least uses the output silo effectively. If your application is D/N, DZs are CHEAP (used) and M+ D/N DOES support DMA use of DZs!! You need a KMC11-A (the older 1k, NOT the newer -b 4k version) microprocessor to nursemaid several DZs then called KDZs in D/N terminology. Is there some reason you do not want to or can't use terminal servers? Ask George Merriman about why terminal servers turned out to be better than custom Simpact cards in another wire service environment that may be similar to yours. There may be some more general info on DHx class muxes buried in the HOME BREW DS500 thread here, but I have just repeated what came to mind as relevant. ================================================================================ Note 73.5 DHU-11 problems and fixes 5 of 6 EISNER::HARVEY "Jack Harvey" 5 lines 17-JAN-1989 23:52 -< Using a Read-only Driver? >- -------------------------------------------------------------------------------- Charlie Byrne asking about DHU's on RSX11M+ makes my skin crawl. If he's using it for an application I recall, it's *all* input, with no flow control, and zero output. If it's PQ (or offspring) hang onto your hat, Charlie. Be prepared to become a silo farmer. ================================================================================ Note 73.6 DHU-11 problems and fixes 6 of 6 EISNER::BYRNE_C "Charlie Byrne" 15 lines 18-JAN-1989 20:52 -< Thanks for help, decision will be made soon >- -------------------------------------------------------------------------------- > < Note 73.5 by EISNER::HARVEY "Jack Harvey" > > -< Using a Read-only Driver? >- > Charlie Byrne asking about DHU's on RSX11M+ makes my skin > crawl. If he's using it for an application I recall, > it's *all* input, with no flow control, and zero output. What a memory Jack, when did you last work on CC: driver (a custom Reuter M/M+ driver) - 1981?! It's actually is not quite *that* bad, but, yes, it will be mostly input from another computer, not terminals. There are a a number of reasons why we aren't using synchronous devices which I will not go into here. Thanks to George and Barton for their valuable assistance. Will post results if they are of interest.  ================================================================================ Note 74.0 BA23 power switch needed 18 replies EISNER::KENNEDY "Terry Kennedy" 18 lines 17-NOV-1987 00:21 -------------------------------------------------------------------------------- Help! The power on/off switch on my BA23 died. I called DEC to get the part number so I could order a new one (actually, two to be safe), but I was told 'This part cannot be sold'. If it could, it would cost $18. I then called Corporate Customer Relations (the nice people who are supposed to help with these things). After a day of cogitation, their best answer was 'Call Field Service to come out and install a new one'. I explained that I didn't really want to pay $175.00 for a FS droid to deliver an $18.00 part, but the message didn't get across. I am now looking for a switch on the open market. The problem is that it is an oddball made specially for DEC in Europe. The oddness has to do with that it is illuminated, but the BA23 has a universal power supply. Suffice to say that only a real DEC switch will work. (Right now I have the wires jumpered permanently 'ON'). Anybody know who else to beat up on in DEC? Or, better still, happen to have a switch or two laying around? ================================================================================ Note 74.1 BA23 power switch needed 1 of 18 EISNER::KILLEEN "Jeff Killeen" 5 lines 17-NOV-1987 01:06 -< HAVE YOU TRIED THIS? >- -------------------------------------------------------------------------------- > I explained that I didn't really want to pay $175.00 for a FS droid to > deliver an $18.00 part, but the message didn't get across. What about depot repair? ================================================================================ Note 74.2 BA23 power switch needed 2 of 18 EISNER::KENNEDY "Terry Kennedy" 6 lines 17-NOV-1987 01:09 -< No good >- -------------------------------------------------------------------------------- > What about depot repair? That involves sending the whole BA23 in (if they would take it), but they won't take it because it is a problem with a cabinet component. Besides, it would cost more and we'd be without the Newsletter system. ================================================================================ Note 74.3 BA23 power switch needed 3 of 18 EISNER::NORTON 2 lines 18-NOV-1987 00:38 -< Junkyards? >- -------------------------------------------------------------------------------- Maybe one of the brokers, like Newman or Midwest Systems, or a repair service like ESS has a junk chassis they can grab the switch from. ================================================================================ Note 74.4 BA23 power switch needed 4 of 18 EISNER::FRISBIE "Alan E. Frisbie" 8 lines 18-NOV-1987 02:19 -< Another possible source >- -------------------------------------------------------------------------------- >> Maybe one of the brokers, like Newman or Midwest Systems, or a >> repair service like ESS has a junk chassis they can grab the switch from. I suggest that you try Digital Basics first. They are a lot smaller and tend to deal a bit better with unusual requests. Alan ================================================================================ Note 74.5 BA23 power switch needed 5 of 18 EISNER::KENNEDY "Terry Kennedy" 10 lines 18-NOV-1987 21:04 -< Happy ending >- -------------------------------------------------------------------------------- This sorry saga has a happy conclusion. After enough hopping up and down on my part, Corporate Customer Relations told the local branch F.S. office to come out and install the switch for *free*! When the unit manager called me, I said that was much to much to expect them to do and I'd be glad enough to get the switch. So I am going out to the branch tomorrow to pick it up. Now, I hope it is the right one... See, at least someone in there *does* care about customers. ================================================================================ Note 74.6 BA23 power switch needed 6 of 18 EISNER::HANRAHAN "Jamie Hanrahan" 10 lines 19-NOV-1987 18:25 -< NIH strikes again (no, not Nat'l Inst. of Health) >- -------------------------------------------------------------------------------- > The problem is that [the switch] is an > oddball made specially for DEC in Europe. The oddness has to do with > that it is illuminated, but the BA23 has a universal power supply. > Suffice it to say that only a real DEC switch will work. I know this is off the topic (and probably belongs in SOAPBOX), but this seems to me to sum up much of what is wrong with DEC these days. Can you say "Not Invented Here"? I knew you could... If it was only things like switches, it wouldn't matter so much, but... ================================================================================ Note 74.7 BA23 power switch needed 7 of 18 EISNER::KENNEDY "Terry Kennedy" 4 lines 20-NOV-1987 05:59 -< The saga continues... >- -------------------------------------------------------------------------------- Well, I picked up the switch today. It was the right part. So much for the good news... The replacement switch was defective! So, considering what I'd been through just to get it, I disassembled both the old and the new switches and concocted one that works. Took about 3 hours... ================================================================================ Note 74.8 BA23 power switch needed 8 of 18 EISNER::KENNEDY "Terry Kennedy" 4 lines 22-DEC-1987 15:16 -< BA23-UC has it now! >- -------------------------------------------------------------------------------- My BA23-UC cabinet kit arrived today, and surprise! - it includes a whole new power switch assembly. So, if anyone needs a switch, that's how to get one... ================================================================================ Note 74.9 BA23 power switch needed 9 of 18 EISNER::FRISBIE "Alan E. Frisbie" 5 lines 22-DEC-1987 16:17 -< What is it? >- -------------------------------------------------------------------------------- >> My BA23-UC cabinet kit arrived today What is a BA23-UC cabinet kit? Alan ================================================================================ Note 74.10 BA23 power switch needed 10 of 18 EISNER::KILLEEN "Jeff Killeen" 4 lines 22-DEC-1987 16:40 -< BA23-UC >- -------------------------------------------------------------------------------- The BA23-UC changes your BA23 box from the old style 11/73 four button. It allows you mount two RDxx drives in a BA23 box. ================================================================================ Note 74.11 BA23 power switch needed 11 of 18 EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" 2 lines 22-DEC-1987 17:08 -< How about uVAX II's ? >- -------------------------------------------------------------------------------- Does this only apply to 11's or can one use this BA23-UC in uVAX's, too ? ================================================================================ Note 74.12 BA23 power switch needed 12 of 18 EISNER::KENNEDY "Terry Kennedy" 11 lines 23-DEC-1987 04:38 -< Should work... >- -------------------------------------------------------------------------------- > Does this only apply to 11's or can one use this BA23-UC in uVAX's, > too ? Physically, any BA23 box (23 only, not 123, etc.) can have the upgrade. As of the last time I looked, VMS did not claim to support the RD3x drives in this cabinet. No reason why it wouldn't work, though... When I install the kit (tomorrow) I will relate my experiences. I sus- pect I may have a problem because I have an external RD53 as well. There are 3 jumpers on the new panel (undocumented, of course). I think 2 of them specify whether the unit in question is internal or external. ================================================================================ Note 74.13 BA23 power switch needed 13 of 18 EISNER::KILLEEN "Jeff Killeen" 34 lines 23-DEC-1987 08:25 -< RTFM >- -------------------------------------------------------------------------------- > There are 3 jumpers on the new panel (undocumented, of course). I think > 2 of them specify whether the unit in question is internal or external. When you order your MICRO-11 with documentation you find these things out. LTC DIP SWITCH UNIT SW 1 SW 2 ON DISABLE BEVNT ENABLE RESTART SWITCH OFF ENABLE BEVNT DISABLE RESTART SWITCH JUMPERS W1, W2, W3 W1 = RD0 WRITE PROTECT SWITCH W2 = RD1 WRITE PROTECT SWITCH !--! ! ! !--! ! ! ! ! !--! ------- ------- 2 C 1 2 C 1 INTERNAL EXTERNAL W3 = DC OK LIGHT CONTROL ! !--! !--! ! ! ! ! ! ! ! ------- ------- 2 C 1 2 C 1 STANDARD EXPANSION BA23 BOX BA23 BOX ================================================================================ Note 74.14 BA23 power switch needed 14 of 18 EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" 3 lines 23-DEC-1987 13:01 -< Where is BA23-UC described ??? >- -------------------------------------------------------------------------------- WHere did you find information on ordering the BA23-UC ? I've looked in DECDirect, the price book, and in my Digital Reference Service manuals and can't find a thing mentioned about BA23-UC. ================================================================================ Note 74.15 BA23 power switch needed 15 of 18 EISNER::KILLEEN "Jeff Killeen" 4 lines 23-DEC-1987 13:11 -< NOTE 40.18 >- -------------------------------------------------------------------------------- > -< Where is BA23-UC described ??? >- NOTE 40.18 IN THIS CONFERENCE ================================================================================ Note 74.16 BA23 power switch needed 16 of 18 EISNER::KENNEDY "Terry Kennedy" 11 lines 24-DEC-1987 01:14 -< Who do *you* get your manuals from? >- -------------------------------------------------------------------------------- >> There are 3 jumpers on the new panel (undocumented, of course). I think >> 2 of them specify whether the unit in question is internal or external. > When you order your MICRO-11 with documentation you find these things out. I ordered the Micro-11 tech manual LAST MONTH (EK-MIC11-TM) and it didn't say anything. Neither did the 'preliminary' BA23-UC Installation Guide. By the way, when installed 'as is', the panel duplicates the controls and indicators on the external RD5x-D box. EIther the main system or the ex- pander box can un-ready or write-protect the external drive. Neat! ================================================================================ Note 74.17 BA23 power switch needed 17 of 18 EISNER::KILLEEN "Jeff Killeen" 16 lines 24-DEC-1987 09:26 -< BUY A SYSTEM - GET A MANUAL SET >- -------------------------------------------------------------------------------- > I ordered the Micro-11 tech manual LAST MONTH (EK-MIC11-TM) and it didn't > say anything. Neither did the 'preliminary' BA23-UC Installation Guide. There is a MICRO-11 documentation set that comes with the systems if you order the "-xA" variants instead of the "-x2" or -x3" variants. It includes the following.... EK-MIC11-TM-002 M11 TECH MANUAL AZ-GPTAA-MC M11/53 SUPPLEMENT (THATS WHERE THE POWER SWITCH STUFF IS) AZ-FI13A-MC M11 OWNERS MANUAL EK-006AA-AD-001 M11 HARDWARE ADDENDUM DEC is always changing the manuals in this set so we order them with every system. ================================================================================ Note 74.18 BA23 power switch needed 18 of 18 EISNER::KENNEDY "Terry Kennedy" 6 lines 24-DEC-1987 20:48 -< Hole in the logic... >- -------------------------------------------------------------------------------- > There is a MICRO-11 documentation set that comes with the systems if you > order the "-xA" variants instead of the "-x2" or -x3" variants. There is a slight logic flaw in DEC's thinking here. By definition, if I order the BA23-UC, I have an *old* system and don't have the 11/53 adden- dum. Oh well.  ================================================================================ Note 75.0 IRIS/VAX INTERCONNECTION 6 replies EISNER::NORTON 10 lines 18-NOV-1987 19:48 -------------------------------------------------------------------------------- We have a Silicon Graphics IRIS workstation (a Unix box) that the users now tell me needs to do file transfers to/from our VAXcluster. What's the best way to do this, given I don't know beans about Unix. We already have a Kermit distribution. Is it easy to get one (which one?) of the Unix Kermits bootstrapped into the IRIS? Or - there seems to be an Ethernet transceiver port on the back of the IRIS. How about TCP/IP on our Ethernet? Is one of the third-party packages markedly better than the others? ================================================================================ Note 75.1 IRIS/VAX INTERCONNECTION 1 of 6 EISNER::RICE "Gary Rice" 38 lines 19-NOV-1987 09:55 -< KERMIT is OK, but . . . >- -------------------------------------------------------------------------------- > We have a Silicon Graphics IRIS workstation (a Unix box) that the > users now tell me needs to do file transfers to/from our VAXcluster. > ... - there seems to be an Ethernet transceiver port on the back > of the IRIS. How about TCP/IP on our Ethernet? Is one > of the third-party packages markedly better than the others? We have a similar requirement that will get $$$ attached to it next year. My investigations to date indicate that the Ethernet approach while certainly a LOT more expensive than KERMIT is probably best. SG transmits TCP/IP as you indicated. The Ethernet cable will handle that just fine and co-exist with DECnet, but they will ignore each other. To get TCP understood on your VAXen, you will need to go to a 3rd party house for hardware & software. I have looked at several vendors. I will mention 2 here, one good, one bad. The BAD one I mention because they happen to be the one that DEC has formed a Cooperative Marketing Program agreement with. That being the Wollongong Group. I don't have any comments about the quality of their stuff since I haven't used it yet. However, I have found them to have a "holier-than-thou" attitude when it comes to service and product availability. A vendor that seems to be a lot more interested in providing us with a solution to the connectivity problem is Excelan. Their Southern California location is at: 4 Marigold Irvine, CA 92714 Phone: (714)733-1500 They are not real big yet, so you may not have a local office. Gary ================================================================================ Note 75.2 IRIS/VAX INTERCONNECTION 2 of 6 EISNER::ERSKINE "Chris Erskine" 13 lines 19-NOV-1987 10:30 -< Excelan seems to be trying to help the customer >- -------------------------------------------------------------------------------- >A vendor that seems to be a lot more interested in providing us with >a solution to the connectivity problem is Excelan. I installed their software in a Micro Vax II without any problem. After installation it seemed to come right up and do what it claimed. This was done a consulting job for a proposal show and tell session which included a number of different systems and vendors being tied together. The customer felt that the show and tell did what they were looking for in talking between systems. When I had questions for the Excelan hardware/software, they were very helpful and help discuss some workarounds for nonstandard function requests which was desired by the customer. ================================================================================ Note 75.3 IRIS/VAX INTERCONNECTION 3 of 6 EISNER::CETRON 23 lines 19-NOV-1987 14:09 -< tcp/ip is best >- -------------------------------------------------------------------------------- as the local vms networking guru here (tcp/ip and decnet) I have a few odds and ends to mutter about. 1. excelan IS very helpful, but the flip side is that their packages are hardware/software mixes so every machine to tcp/ip needs a board. 2. wollongong's new version is supposed to be very good, and there service IS improving (since they had acquired such a bad reputation I think they must have almosts gone under and I think that that has scared them.) I spent over two hours with a vp on the phone after writing some really scathing articles on the usenet about them. mark them as improving.... 3. what we actually run, is the cmu tcp/ip from cmu.. top notch stuff, cost is only for distribution (100 bucks or so). Is basically a site license and runs very well. Down side - it requires a bliss compiler (but the source comes free) and there is NO support. but with the sources and our bliss compiler, we have no problem -then again, I've only had to make 1 patch.... -ed ================================================================================ Note 75.4 IRIS/VAX INTERCONNECTION 4 of 6 EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" 5 lines 20-NOV-1987 13:11 -< Green code with lumps. >- -------------------------------------------------------------------------------- BLISS... geez, I've never heard of a product more inappropriately name. You should see some of the Tektronix TCP/IP code. Reminds me of the EXORCIST. ================================================================================ Note 75.5 IRIS/VAX INTERCONNECTION 5 of 6 EISNER::PROVOST 7 lines 20-NOV-1987 13:25 -< Ancient etouq >- -------------------------------------------------------------------------------- > < Note 75.4 by EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" > > -< Green code with lumps. >- > BLISS... geez, BLISS is IGNORANCE ================================================================================ Note 75.6 IRIS/VAX INTERCONNECTION 6 of 6 EISNER::CETRON 11 lines 23-NOV-1987 14:14 -< ...green code>?> >- -------------------------------------------------------------------------------- Bob, Before cmu rewrote the tek code, I use to modify it daily....in fact, I even wrote a fortran front end to the BLISS..... using the lse set to bliss is even got real good at it.. but the tek code!!! green code nothing - it was emesis yellow, with green lumps.... -ed  ================================================================================ Note 76.0 HSC50 lost its NAME 2 replies EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 25 lines 19-NOV-1987 12:07 -------------------------------------------------------------------------------- I had an interesting experience with our dual HSC50s this morning. Seems when DEC F/S last "fixed" one of the HSCs for us, by swapping either the K.ci or P.ioc (I think the P.ioc), they must have swapped the tape with its partner. When I brought a node up we had two nodes with the same ID and name. It has really helped my confidence with DEC F/S to see this. Not only that but the tape in the other HSC did not have ID, NAME, or ALLOCATION CLASS info. I fixed it all up and now it all seems to be working. Lesson: 1) When DEC fools with your system assume they do not know what they are doing, unless you know them well. 2) Learn how to manage your HSCs. Question: When a SET is done on the HSC does this write the information to the tape? Can this information (NAME, ID, ALLOCATION CLASS) be written to tape? Thanks Seton ================================================================================ Note 76.1 HSC50 lost its NAME 1 of 2 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 45 lines 19-NOV-1987 16:19 -< One possible scenario. >- -------------------------------------------------------------------------------- Quoting from the "HSC User Guide": Some SET commands change the HSC in such a way that reboot is necessary for the commands to take effect. For example, the command: SETSHO> set name alamo sets the HSC System Name to ALAMO. The system name is a parameter that may not change while the HSC is running... The HSC must reboot in order for the command to take effect. If any command you enter in SETSHO creates changes that warrent a reboot of the HSC, SETSHO [prompts you if you want the reboot.] If you you type CTRL-Y or CTRL-C to the prompt, the SETSHO program returns to the KMON prompt without making changes. . . . Almost all of the permanent modifications to HSC parameters made by SETSHO are to the System Configuration Table (SCT). The SCT is a file on the HSC load medium called SCT.INI. When you change load media, remember that will also change the SCT.INI file. When you use a different medium, the permanent parameters will be different. . . . When we send you a new load medium, we send it with a set of standardized values for each of the parameters. . . . If you ever care to go back to the default parameter settings, you can hold in the FAULT button while pressing the INIT button when rebooting your HSC. Continue to hold FAULT in until you see the INIPIO-I Booting... message appear on the console. My feeling is that this last paragraph describes what is happening to you. Your DEC service techie is rebooting your HSC for you when s/he is done with this default value boot. There is an HSC utility, COPY, that allows you to make a copy of the load medium. See chapter 10 in the HSC guide. ================================================================================ Note 76.2 HSC50 lost its NAME 2 of 2 EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 14 lines 19-NOV-1987 20:43 -< Thanks... And what we learned... >- -------------------------------------------------------------------------------- Sounds good. (Must admit I worked from the release notes, just to get things back to normal.) I certainly plan on setting up the proper backups, etc. This whole thing is a result of poor attention on DECs part during the system installation. We did not have the best team doing the cluster installation. When we asked about the HSCs they said "They are all set, don't worry about them...", and we didn't. We were naive, bordering on dumb... My point in posting the note is to remind myself, and others, not to think of your HSCs as black (digital grey?) boxes that work and F/S fixes them... Learn about them, hug them, treat them nicely... Seton  ================================================================================ Note 77.0 VAX 8530 performance myth 1 reply EISNER::RENES 21 lines 22-NOV-1987 17:36 -------------------------------------------------------------------------------- WE have recently installed a VAX 8530 system. I made the mistake (haha) of benchmarking it against the following processors: 750 780 uV2000, 8200, 785, 8700 My comparitive results show that the VAX 8530 is ONLY 3x the 8200 (or 780) and *NOT* 4x, as DEC advertises. Their sales update indicates that the VAX 8530 is 4x an 8200 in all their benchmarks and 4.5x in some! My benchmark was a simple integer arithmetic example, using a lib$show_timer to show elapsed CPU. This test was run on 3 different VAX 8530 processors at two different sites. Any comments? ================================================================================ Note 77.1 VAX 8530 performance myth 1 of 1 EISNER::RENES 11 lines 24-NOV-1987 16:12 -< ..more info.. >- -------------------------------------------------------------------------------- I guess I'll add the first reply, to wit: I have now discovered that: 1) All Nautilus machines (85nn,87nn,88nn) now share the exact same microcode (4 files). I can only presume that 8500 machines used butchered up microcode to erode performance? 2) Character data benchmarks indicate that all these machines perform about the same (they are roughly 7x a 785!!).  ================================================================================ Note 78.0 360K Floppies on VaxMates 5 replies EISNER::DOHERTY "Bob Doherty" 19 lines 28-NOV-1987 22:47 -------------------------------------------------------------------------------- We have several VAXMates being served by a 750 and have had trouble transporting floppies from the VAXMates to PC clones even when the disks in question had been formatted(at 360k) by the destination PCs. Has anyone had any experience adding 360k floppies to VAXMates? We have a 'pizza box' on one of the machines with both slots free. What kind of controller is necessary for the 360K floppy? Will there be an adverse interaction with the DEC driver used for the internal VAXMate's floppy? We have a fair amount of experience with Unibus, Massbus, and QBus peripherals but next to none with PC peripherals. What should we be looking for? Thanks for any help. Bob. ================================================================================ Note 78.1 360K Floppies on VaxMates 1 of 5 EISNER::KENNEDY "Terry Kennedy" 22 lines 29-NOV-1987 05:37 -< Tech info, possible alternative >- -------------------------------------------------------------------------------- First, I presume Vaxmates use an IBM-compatible 1.2 Mb diskette? Given that, the problem is that in order to fit 80 tracks on the same space as the 40 on a 360K diskette, the tracks must be narrow- er. They are actually only about 40% as wide as a 360K track. (the extra 10% is for a guard band between tracks. Therfore, when you write a 360K disk in a 1.2M drive, you're writing a narrow strip down the center (hopefully) of the 360 track. The Vaxmate will read this fine, but the PC still sees the old data around the narrow strip, as well as the strip itself. As for putting a 360K drive in a Vaxmate, I don't know as I don't have one. At the Anaheim Symposium I'll try to poke into one and let you know when I get back. Another thing you could do is to put 1.2M drives in your PC's - one product I know of to do this is the 'Compaticard' from Micro Sumltions. Itr replaces your regular floppy controller in the PC and lets you attach 8" (RX01-style) drives, 3 1/2" 720K, 3 1/2" 1.4M, and 5 1/4" 1.2M drives. Depending on the number of PC's you need to interchange with, this might be the easiest (and 'sup- ported') solution. ================================================================================ Note 78.2 360K Floppies on VaxMates 2 of 5 EISNER::HAHN 4 lines 29-NOV-1987 12:40 -< Need More information >- -------------------------------------------------------------------------------- Can you give telephone number/address for Micro Simulation. Would like to get the literature on the above product. Pierre ================================================================================ Note 78.3 360K Floppies on VaxMates 3 of 5 EISNER::KENNEDY "Terry Kennedy" 7 lines 30-NOV-1987 21:02 -< Micro Solutions' address >- -------------------------------------------------------------------------------- Here is the information: Micro Solutions, Inc. 132 West Lincoln Hwy. DeKalb, Illinois 60115 (815) 756-3411 ================================================================================ Note 78.4 360K Floppies on VaxMates 4 of 5 EISNER::HORN "Larry Horn" 9 lines 1-DEC-1987 01:47 -< bulk erase | format /4 >- -------------------------------------------------------------------------------- One other thing to try -- bulk erase the PC diskette or use a virgin diskette, then format it on the PC. It could be that there is a lot of residual 'noise' on the diskette from long-term use. You might also try "FORMAT A:/4" on the VAXmate on a bulk-erased diskette. (See appendix H of the VAXmate User's Guide -- it mentions some problems with changing the density of diskettes and recommends bulk erasing them, saying that going from high to low density will likely REQUIRE bulk erasing.) ================================================================================ Note 78.5 360K Floppies on VaxMates 5 of 5 EISNER::STRIEGEL "Alan Striegel" 9 lines 29-NOV-1988 21:49 -< Only format on high-density drive >- -------------------------------------------------------------------------------- The best method is to use a bulk-erased floppy and use the FORMAT /4 command on the VAXmate. You should never try to write with a high-density drive to a diskette that has been formatted with a low density drive. There is also a software product being advertised to perform the proper track width writing with a high-density drive - this program is called COPYAT2PC, but I can't recall whose package it is. Alan  ================================================================================ Note 79.0 IBM (!) PC configurations 31 replies EISNER::BYRNE_C "From scenic Long Island" 30 lines 3-DEC-1987 17:00 -------------------------------------------------------------------------------- I'm not sure where to put this note, but here it is, for now at least: I need to buy an IBM-PC compatible system for use here at work. Main reason is all project managers here must use a project management package called PMW. I will also use it for GEM (for pictures in my TeX documents), Catalyst (on-line ADA Language ref manual) and God knows what else. I know very little about PC's. Here is what I am considering: 1) a 386 PC with 20MB disk, 1 floppy, 1 parallel port, 2 serial ports, EGA controller 2) Monitor (AMDEK 710? NEC Multisync?) 3) Epson FX - 286 printer 4) Bus mouse 5) PC sidecar, CRT float device, extended cables. Cost is not a primary concern. I would greatly appreciate any comments particularly on Monitor and Printer, because there are millions out there and I just don't think I can do a proper evaluation job in less than several years. Also, what would be the advantages of getting DECnet DOS to connect to our main development network? And there's some other package (PCSA, I think) that's supposed to be good for file servicing. I wish I could spend this money on a Vaxstation, but we apparently are stuck, corporate-wide, with this non-DEC Project Management tool. Thanks in advance. ================================================================================ Note 79.1 IBM (!) PC configurations 1 of 31 EISNER::FRISBIE "Alan E. Frisbie" 19 lines 3-DEC-1987 18:29 -< Printer == Toshiba! >- -------------------------------------------------------------------------------- >> I would greatly appreciate any comments >> particularly on Monitor and Printer, because there are millions I have been using a Toshiba P1351 printer for four years now and am EXTREMELY happy with it. There are newer models available now, but I still consider their quality to be absolutely tops. After four years of moderate use, the only repairs have been: A new carriage drive motor (high speed skips became erratic) A new print head after a paper jam occured 15 minutes into a 2 hour print job (I made the mistake of leaving the office). I can only recall two or three paper jams in four years, and none of the others caused any damage. Alan ================================================================================ Note 79.2 IBM (!) PC configurations 2 of 31 EISNER::KENNEDY "Terry Kennedy" 45 lines 3-DEC-1987 20:58 -< 5000-word essay... >- -------------------------------------------------------------------------------- Having built several hundred PC 'clones' for the place I work for, here is my advice, in the same order you used: 1) a) If you feel you need a 386 PC, be *sure* that the vendor has tested it with the OS/2 software. That will catch many minor compatibility problems you won't notice right away with DOS. b) Why cripple a machine like that with a 20 Meg drive? I sug- gest the AT70I from Storage Dimensions - it is 70Mb for only $825.00 (by the way, it's the same drive as the DEC RD53) c) Two floppies is a must - one 1.2 Mb and one 360K - otherwise you'll have trouble making disks for regular PC's d) Two serial ports or 1 serial port and an internal modem? If you need a modem, try the Everex EV-940 - it's 2400 baud, only $189.00, and highly reliable (no MNP, though) e) EGA controller - avoid the Taiwanese at all costs here - most of them won't work at 386 speeds. Two I know of that do work with a 386 are the Video 7 Vega Deluxe (overpriced at $375, and the software isn't the best), and the Everex Micro Enhancer ($175.00 with a parallel port on it, also bad software) 2) Monitor - The MultiSync is now a 'so-so' unit, as better ones have become available. Almost all MultiSyncs suffer from a pincushioned display, and none will do 132 columns with the mentioned EGA's with out some knob-twiddling (picture goes off the right edge). Take a look at some of the new Taxan units - many come bundled with a Par- adise Auto-switch II EGA card. 3) Printer - get something that your software 'knows' about. The P1351 is good, as was pointed out earlier. Personally, I like the NEC P6/P7 family (narrow/wide carriage, respectively) Whatever printer you get, plan on plenty of extra ribbons. 4) Just be sure you have enough slots and interrupts left. 5) Some systems don't like being tilted 90`. Keyboard can be extended with a cable from Curtis (~$25.00). Do *not* extend the monitor cable - you WILL have problems - trust me. 6) Tape backup - you need one! Personally, I like the Wangtek 60Mb mechanism - comes under many names & prices from most vendors - drop me a MAIL message if you want particular suggestions. 7) DECnet-DOS - If you're not on a VAXmate, woof! The terminal emulator really wants to run on a Vaxmate. If you don't nead heavy file access, just terminal emulation, check out version 2.30 of MS-Kermit (free). ================================================================================ Note 79.3 IBM (!) PC configurations 3 of 31 EISNER::GARDNER "Tim Gardner" 54 lines 4-DEC-1987 07:44 -< Comments (lots) >- -------------------------------------------------------------------------------- RE: < Note 79.0 by EISNER::BYRNE_C "From scenic Long Island" > > 1) a 386 PC with 20MB disk, 1 floppy, 1 parallel port, > 2 serial ports, EGA controller > 2) Monitor (AMDEK 710? NEC Multisync?) > 3) Epson FX - 286 printer > 4) Bus mouse > 5) PC sidecar, CRT float device, extended cables. Some general comments: a) as noted elsewhere, a 20Mb drive isn't big enough. I don't know that I'd reccommend a 70Mb drive unless your doing software development for the PC, but certainly a 40Mb drive is essential. b) The NEC Multisync is pretty good, the Multisync II is better. I personally like the Mitsubishi Diamond Scan (higher res that NEC). If your going to use an EGA graphics card, why not just by an EGA monitor? You will spend less, and usually get a slightly crisper image. The Amdek 722 is highly recommended. c) The Epson FX series is excellent. From my role as the sole support person in my company (supporting 60 or so PC as well as all the DEC stuff), I always recommend Epson. Any Epson. They rarely require repairs, can be serviced when it does become necessary, and virtually all software knows how to speak to them. d) If you haven't already purchased the system unit, make sure you get a box that does NOT use the Intel 512Kb motherboard. Many problems including no 80387 support (it has a socket, but its broke). Intel could only get 512Kb on the motherboard, you will need 640Kb (or 1Mb). The boxes from Everex are very high on my list. See ads from Eltech, Club American Technologies, etc in mags like PC Week, Byte or PC Magazine. e) If you want graphics, don't buy anything less than EGA. CGA sucks rocks, big time. VGA is nice, but still pretty expensive, and not much software can use it in other than EGA mode. Avoid buying any graphics card from IBM; the clones usually cost only 1/3 the IBM price, and usually can do emulation of several different cards. Vega 7 from Video 7 (or from Quadram) or one of the cards from Everex is recommended. f) DECnet-DOS. Great if you have Ethernet, but pretty slow over serial lines. I use it every day as the connection between my AT and the PDPs and VAXes. (I don't even have a terminal, the VT102 emulator in DECnet-DOS is quite sufficient). g) If you like LK201 (VT220) keyboards, get the 101 key "enhanced" keyboard, If you prefer VT100 keyboards, get the "old style" 84 key keyboard. I could go on like this for some time, but this covers the highlights. Send me mail if you would like to talk further. Tim. ================================================================================ Note 79.4 IBM (!) PC configurations 4 of 31 EISNER::OSTROSKY "Jim "Crash" Ostrosky" 27 lines 8-DEC-1987 07:33 -< other PC notes >- -------------------------------------------------------------------------------- RE: < Note 79.0 by EISNER::BYRNE_C "From scenic Long Island" > > 1) a 386 PC with 20MB disk, 1 floppy, 1 parallel port, > 2 serial ports, EGA controller > 2) Monitor (AMDEK 710? NEC Multisync?) > 3) Epson FX - 286 printer > 4) Bus mouse > 5) PC sidecar, CRT float device, extended cables. a) Disk drive - as noted elsewhere. I won't add anything. b) EGA is a must c) Monitor; the AMDEK 722 is a real nice monitor; I wish I had bought one for my home system. As I recall, it had an almost flat screen, small footprint, and a pushbutton on the front for various color modes. d) Printer; yes, make sure your software knows the printer *OR* make sure the printer you buy emulates something the software knows. I just bought a NEC P2200 narrow width 24 pin printer that emulates EPSON. I am hoping that in a year some of the WP software will recognize it. But also consider sharing a laser printer among several PC's (ala multiport spooler black box). HP Laserjets seem to multiply around here. 5) and if money isn't a problem, I'll give you MY shopping list and have you send it to me for Christmas... ================================================================================ Note 79.5 IBM (!) PC configurations 5 of 31 EISNER::NAGY "Frank J. Nagy, VAX Guru" 15 lines 16-DEC-1987 08:18 -< Try the Sony MultiScan monitor >- -------------------------------------------------------------------------------- The PC users in my department (Electrical & Electronics) are very happy with AST Premium-286s (AST now makes a 386 machine), with 45 MB Micropolis drives (yes, you will need more than 20 MB), with Video-7 Vega Deluxe EGA controller, Sony MultiScan monitors and serial LogiTech mice. Video-7 now has a new display board which is both EGA and VGA compatible, the Sony monitor is capable of handling the VGA output in (I think) 800x600 mode; the NEC MultiSync could not handle this display format. We also use the Microm Ethernet boards and DECnet-DOS. I think that if we had to do it over again, we would seriously consider the Digital Network Integration kit (at least the software and Ethernet board). ================================================================================ Note 79.6 IBM (!) PC configurations 6 of 31 EISNER::BYRNE_C "Can you say Hauppauge?" 5 lines 16-DEC-1987 15:44 -< Keep cards and letters coming >- -------------------------------------------------------------------------------- Thanks for all these replies to date. I am investigating. More replies are welcome. Someone has suggested a particluar printer that has both parallel and serial ports, thus can connect to PC and VTxxx. Is this a significant factor? ================================================================================ Note 79.7 IBM (!) PC configurations 7 of 31 EISNER::BYRNE_C "Charlie Byrne, Hauppague NY" 15 lines 6-JAN-1988 22:03 -< Bill of Goods >- -------------------------------------------------------------------------------- It looks like this is what is in the plans: NEC Powermate II (286) w/ 40MB hard and 1.2MB floppy NEC Multisync II Monitor 2400 baud internal modem w/ MNP (not sure of manufacturer) Quadram EGA board (w/ free mouse thrown in) Epson FX-1000 (serial/parallel/near letter quality) DEC Network Integration Kit (DECnet DOS, Ether board, various other stuff, plus the LK250 keyboard) One of the reasons for the NEC stuff is that I believe my company has some kind of volume deal with them. Thanks for comments, if you think I'm making a mistake somewhere, let me know. Else will let you know how it's going in a few months. ================================================================================ Note 79.8 IBM (!) PC configurations 8 of 31 EISNER::HAHN 6 lines 6-JAN-1988 22:46 -< Would opt for External Modem >- -------------------------------------------------------------------------------- Beware of internal MODEMS, because of local storm in our area I know of 2 mother boards that where zapped thru the telephone line still connected to the internal modem, even though power was disconnected. Other PCs with external modems all survived. Pierre ================================================================================ Note 79.9 IBM (!) PC configurations 9 of 31 EISNER::KENNEDY "Terry Kennedy" 3 lines 7-JAN-1988 19:25 -< 360K floppy for interchange... >- -------------------------------------------------------------------------------- > NEC Powermate II (286) w/ 40MB hard and 1.2MB floppy If you are planning on writing disks that others with XT-class machines will want to read, you should [also] get a 360K floppy... ================================================================================ Note 79.10 IBM (!) PC configurations 10 of 31 EISNER::BYRNE_C "Charlie Byrne, Hauppague NY" 9 lines 7-JAN-1988 22:26 -< Can use DECnet for floppy interchange? >- -------------------------------------------------------------------------------- > < Note 79.9 by EISNER::KENNEDY "Terry Kennedy" > > -< 360K floppy for interchange... >- > >> NEC Powermate II (286) w/ 40MB hard and 1.2MB floppy > If you are planning on writing disks that others with XT-class machines >will want to read, you should [also] get a 360K floppy... If they are also on our net via DECnet-DOS, though, I could just do the transfer via the net. I assume? ================================================================================ Note 79.11 IBM (!) PC configurations 11 of 31 EISNER::KENNEDY "Terry Kennedy" 8 lines 8-JAN-1988 02:36 -< Well... >- -------------------------------------------------------------------------------- > If they are also on our net via DECnet-DOS, though, I could > just do the transfer via the net. I assume? True, if everyone you want to exchange data with is on your net. However, if you need to make disks for others not on your net in 360K format, you'll have to tie up someone else's machine to write the disk. Since 360's are selling for $65. or so, it doesn't really make sense to not have one [in my opinion]. ================================================================================ Note 79.12 IBM (!) PC configurations 12 of 31 EISNER::ERSKINE "Chris Erskine" 15 lines 11-JAN-1988 06:55 -< I Second the 360k Disc >- -------------------------------------------------------------------------------- I agree with the 360k floppy. I have been developing some code on 2 different machines and worked it by - Work on XT - make disc - Work on AT - make disc - Go to another AT - convert HD disc to 360k. I am planning on buying a personal PC and have decided that I want an AT class machine with both 1.2k and 360k flakey disc drives. When I can afford it, I will put a tape drive on to support backups from the hard drive. It does take some work on both machines to transfer using DECnet. You have to run FAL on the remote machine to transfer the file. ================================================================================ Note 79.13 IBM (!) PC configurations 13 of 31 EISNER::HANRAHAN "Jamie Hanrahan" 26 lines 11-JAN-1988 14:58 -< Don't have a 360K drive? Try bulk erasing >- -------------------------------------------------------------------------------- If you have only 1.2M drives, and you have trouble writing floppies in 360K mode (for reading on 360K drives; there's generally no problem reading them on a 1.2, but what's the point of that?), try bulk-erasing them first; then format in 360K mode on the 1.2 drive, then write. Any AC line-operated bulk eraser, such as is sold for audio cassettes, will work fine. (Do this at a considerable distance from any mag media with data you want to keep. Theroetically, two or three feet is more than sufficient, as magnetic field strength from such devices drops off VERY fast with distance. I keep my bulk eraser at the opposite corner of the room, about ten feet from my media.) Explanation: The track written by a 1.2 head is less than half the width of that written by a 360K head. If your floppy was previously formatted on a 360K drive, then you write on it with a 1.2 head, the "edges" of the 360K tracks don't get completely erased. When you read such a floppy with a 360K head it reads the full track width, including the 1.2-width good data and the edges of the old 360K data, and so the signal is garbled. Similarly, if the floppy had been used in 1.2 mode, formatting it at 360K in a 1.2 drive simply overwrites *every other* 1.2 track (the even-numbered ones, I think); the rest don't get touched. When a 360K head tries to read the disk it picks up the good data on the even tracks plus the edges of the adjacent odd-numbered tracks. Garble again. By bulk erasing the disk first you ensure that what you write with the 1.2 head is all there is to read. ================================================================================ Note 79.14 IBM (!) PC configurations 14 of 31 EISNER::LEDERMAN "Bart Z. Lederman" 8 lines 11-JAN-1988 20:35 -< I also find it works >- -------------------------------------------------------------------------------- For what it's worth, I've also been able to exchange between the two kinds of floppies by making sure the floppy is blank and formatting it on the right kind of machine. It's even possible to go from IBM PC to Rainbow RX50 that way (though DEC frowns upon using anything but RX50 diskettes for any length of time in an RX50 drive). It's best to use this technique just for transfer, and to copy the data onto the proper kind/format diskette as soon as possible if the data must be used a lot on the target machine. ================================================================================ Note 79.15 IBM (!) PC configurations 15 of 31 EISNER::ERSKINE "Chris Erskine" 6 lines 12-JAN-1988 08:46 -< Reformat >- -------------------------------------------------------------------------------- > Similarly, if the floppy had been used in 1.2 mode, formatting it at 360K > in a 1.2 drive simply overwrites *every other* 1.2 track (the even-numbered > ones, I think); the rest don't get touched. When a 360K head tries to read After formatting a disk at 1.2M, I have not been able to reformat it at 340K on either a 1.2 or 3600 drive. ================================================================================ Note 79.16 IBM (!) PC configurations 16 of 31 EISNER::HAHN 4 lines 12-JAN-1988 09:04 -< ERASE all first >- -------------------------------------------------------------------------------- If you degauss it (tape bulk eraser works fine) it will be able to be reformatted. Pierre ================================================================================ Note 79.17 IBM (!) PC configurations 17 of 31 EISNER::KENNEDY "Terry Kennedy" 13 lines 12-JAN-1988 22:54 -< Not worth the trouble... >- -------------------------------------------------------------------------------- > After formatting a disk at 1.2M, I have not been able to reformat > it at 340K on either a 1.2 or 3600 drive. You shouldn't be able to format a 1.2 at 360 even *before* formatting it at 1.2 - 1.2's use a radically different oxide and the drive has to be told to run in 'scorch mode' to write on them. Regarding using bulk-erased diskettes, sure, but there are still severe problems (such as the 360's centerline being between two 1.2 tracks), so people with out-of-alignment 360's may have problems reading your disks. Also, you have to remember which pile of disks is bulk-erased. I contend that it isn't worth the trouble... ================================================================================ Note 79.18 IBM (!) PC configurations 18 of 31 EISNER::BYRNE_C "Charlie Byrne, Hauppague NY" 7 lines 13-JAN-1988 20:40 -< Not WORTH the trouble sounds right >- -------------------------------------------------------------------------------- >< Note 79.17 by EISNER::KENNEDY "Terry Kennedy" > > -< Not worth the trouble... >- >Also, you have to remember which pile of disks is bulk-erased. I contend >that it isn't worth the trouble... I think I agree. Me thinks me will spend the lousy 60 bucks or whatever and get a 1.2 and a 360 drive. ================================================================================ Note 79.19 IBM (!) PC configurations 19 of 31 EISNER::ERSKINE "Chris Erskine" 5 lines 14-JAN-1988 06:47 -< I always can find them >- -------------------------------------------------------------------------------- >Also, you have to remember which pile of disks is bulk-erased. I contend >that it isn't worth the trouble... That is easy. The bulk erased disks are the ones that have your latest backup on them when the hard drive crashes. ================================================================================ Note 79.20 IBM (!) PC configurations 20 of 31 EISNER::NORTON 18 lines 28-JAN-1988 11:02 -< DEC/PC comparison questions >- -------------------------------------------------------------------------------- On another track - how do I compare IBM PC's to things like 11/23's for evaluation purposes? You can get mux boards for PC's but how do you program under MS-DOS to get multiple terminals to work? Its a given with RSX, but MS-DOS is supposedly single-user. Do you have to set up a PC with one big program that scans the ports for characters and tries to look multitasking? And for that matter, how would an AT perform with 16 ports, four of them terminals with barcode wands, four scale inputs, one printer, and two of them LAN connections? (I know - it depends on the application!) I can rate this sort of thing if we say its RSX, but how does a PC with MS-DOS compare? HELP \bill ================================================================================ Note 79.21 IBM (!) PC configurations 21 of 31 EISNER::GARDNER "Tim Gardner" 41 lines 28-JAN-1988 15:18 -< The P means Personal >- -------------------------------------------------------------------------------- Re: < Note 79.20 by EISNER::NORTON > > On another track - how do I compare IBM PC's to things like 11/23's > for evaluation purposes? You can get mux boards for PC's but how > do you program under MS-DOS to get multiple terminals to work? > Its a given with RSX, but MS-DOS is supposedly single-user. Do > you have to set up a PC with one big program that scans the ports > for characters and tries to look multitasking? You compare PCs with 11/23s the same way you compare apples and oranges. Both are computers (fruits), both have appealing features, both have similar uses, but they are different things. A PC is a ONE user system. Sure, you can put in a 4, 8 or 16 port adapter but don't expect to run MS/PC-DOS. The board makers say you can, and in fact DOS will boot, but almost no PC software will work over a dumb terminal. Most software expects a PC video board that it can talk to directly, something that just doesn't exist if you try to use a serial port as a terminal. Also note that DOS will still run the CPU in real mode, thus allowing user programs to trash other users' programs. You must run the machine in protected mode (or virtual mode on a '386) to prevent even accidental trashing of other users and/or the system itself. One possible answer to the DOS limitations is The Software Link's PC-MOS. It is a protected mode operating system that runs on 286 and 386 systems that does a very good job of emulating DOS. It allows mutitasking and multiuser operations and supports up to 25 users (TSL admits that even a 16Mhz 386 system is practically limited to about 8 or 9 users). Another possible answer is Xenix. Xexix is still practically limited to two or three users on an 8Mhz '286 (plan on at least 1Mb of memory for the system and another 1Mb for each simutaneous user). A 80386 system will outperform a 80286 system any day, but you are still severely limited by the bus bandwidth and specifically the disk throughput (especially under Unix, where EVERYTHING requires several disk reads). In general, remember that P in PC means Personal. tg ================================================================================ Note 79.22 IBM (!) PC configurations 22 of 31 EISNER::ERSKINE "Chris Erskine" 5 lines 29-JAN-1988 07:08 -< = RT-11SJ >- -------------------------------------------------------------------------------- Another way to look at it is PC-DOS is like the single job RT-11 OS. If you want to right the application, you can talk to as many ports as you want but it is single tasking. Depending on the system, the CPU may have much more power than the 11/23 but PD-DOS is (as I say in print) a monitor, not an OS. ================================================================================ Note 79.23 IBM (!) PC configurations 23 of 31 EISNER::GARDNER "Tim Gardner" 14 lines 1-FEB-1988 08:09 -< Agree!! >- -------------------------------------------------------------------------------- > -< = RT-11SJ >- > > Another way to look at it is PC-DOS is like the single job RT-11 > OS. If you want to right the application, you can talk to as many > ports as you want but it is single tasking. Depending on the system, > the CPU may have much more power than the 11/23 but PD-DOS is (as > I say in print) a monitor, not an OS. > An excellent analogy. I agree wholeheartedly (especially with the last sentence). tg ================================================================================ Note 79.24 IBM (!) PC configurations 24 of 31 EISNER::ERSKINE "Chris Erskine" 16 lines 2-FEB-1988 08:58 -< Auto proof reader needed here >- -------------------------------------------------------------------------------- > -< = RT-11SJ >- > > Another way to look at it is PC-DOS is like the single job RT-11 > OS. If you want to right the application, you can talk to as many write > ports as you want but it is single tasking. Depending on the system, > the CPU may have much more power than the 11/23 but PD-DOS is (as PC-DOS > I say in print) a monitor, not an OS. saw I hate to see myself quoted because it allows me to see my craze typo's. Spell check finds spelling but what I really need is a system to correct my mind to put down what I really meant. ================================================================================ Note 79.25 IBM (!) PC configurations 25 of 31 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 17 lines 17-MAY-1988 14:15 -< PC Power Up/Down reduce MTBF >- -------------------------------------------------------------------------------- Now that I finally have my PC system: No one in their right mind would power down a mid sized or large computer system every night (say a Vax with a bunch of disk drives). However, folks seem to have no such reservations about powering PC's up and down as if they were a refrigirator light, even those equipped with a hard disk. Me thinks I should leave my system on, just as I would a large system (of course it is at work so I have power protection plus no power bills to pay). Does any one know what the best practice is? What will reduce the MTBF more: leaving the Winnie running 24 hours a day or turning it on and off every day? I tend to think that if you have the option, you should maintain the system in a steady state (i.e. On) as long as possible. What is the proper procedure? ================================================================================ Note 79.26 IBM (!) PC configurations 26 of 31 EISNER::KENNEDY "Terry Kennedy" 7 lines 18-MAY-1988 16:24 -< Tradeoff - cycling power vs. heat buildup >- -------------------------------------------------------------------------------- > What is the proper procedure? If (and this is a big if) your system has adequate cooling, it is probably better to leave it on. That's what I do with all my office PC's. The only problem is that the standard PC fan is not big enough to cool a loaded system, so you need to get a bigger one, otherwise you accelerate component aging via heat. ================================================================================ Note 79.27 IBM (!) PC configurations 27 of 31 EISNER::MERRIMAN "George Merriman - CCA/NY" 3 lines 20-MAY-1988 21:40 -< Don't forget the temp cycle >- -------------------------------------------------------------------------------- On the other hand, if the beast is actually running hot cycling the power might exacerbate the temperature cycle problem, which is a real killer for some types of electronic components. ================================================================================ Note 79.28 IBM (!) PC configurations 28 of 31 EISNER::FULLER_S "Stuart Fuller" 19 lines 22-MAY-1988 01:36 -< Leave it on! >- -------------------------------------------------------------------------------- Re: .25 Let me put it this way, I've had my Rainbow-100 for about 4 years now. It has been powered down probably 5 times since I received it, typically while my family and I have taken out of town trips. On 2 of these 5 power cycles, I've had trouble of one form or another. One time, the disk controller board went bad. I don't remember what the other problem was. Both cases, however, occured upon powering up the system. Also, for what it's worth, when do incandescent light bulbs tend to burn out? Not while they're on, but rather when you first turn them on. Draw your own conclusions. Stu ================================================================================ Note 79.29 IBM (!) PC configurations 29 of 31 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 10 lines 23-MAY-1988 14:13 -< If it's going to fail, do it now >- -------------------------------------------------------------------------------- >> Also, for what it's worth, when do incandescent light bulbs tend >> to burn out? Not while they're on, but rather when you first turn >> them on. For this reason, I recommend that people cycle the power a few times any time that Field Service is due to do any work on the machine. Better to have it fail when FS is there, rather than at some critical time. Alan ================================================================================ Note 79.30 IBM (!) PC configurations 30 of 31 EISNER::WICHMAN "Jon Wichman" 6 lines 15-NOV-1988 16:30 -< Powering on and off >- -------------------------------------------------------------------------------- > Cycling power Might be good to do it now while it's under warranty. If something's marginal and is going to break sooner or later. Now, while it's under warranty and hasn't got a ton of stuff on it's a good time... ================================================================================ Note 79.31 IBM (!) PC configurations 31 of 31 EISNER::BYRNE_C "Charlie Byrne" 10 lines 15-NOV-1988 22:00 -< I power PC on/off appx 1.5 times daily >- -------------------------------------------------------------------------------- I've taken to using the PC appx. once or twice a day max, usually to dial into DECUServe from work, especially if I need to up- or -down-line load something (the other times I dial in on a Microcom AX/2400 and VT220 from home). I've had zero problems so far; even though the power on/off cycle does generally speaking have significant detrimental effects upon electronic components, I can't see leaving the PC on 24 hours a day when I only use it perhaps ½ hour/day. Obviously this is just one persons experience.  ================================================================================ Note 80.0 TU78 on VAX 8550? 6 replies EISNER::MCDOUGALL "Bob McDougall" 11 lines 8-DEC-1987 15:09 -------------------------------------------------------------------------------- We are are running VMS V4.6 on a VAX 11/780 with 14 meg, 2 RM05's, 2 RA81's, and 104 user terminals. We are currently considering a major upgrade to a VAX 8550. We have a TU78 tape drive which we have been told will not run on the Q-Bus. DEC Sales has said we could anticipate a 50% decrease in transfer rate with a Q-Bus drive and are suggesting that we replace our 1 drive with 2. 1. Is this really necessary? 2. Are their any adapters available to run a TU78 off an 8550? 3. Are their any drives in development :-) we should wait for? ================================================================================ Note 80.1 TU78 on VAX 8550? 1 of 6 EISNER::ERSKINE "Chris Erskine" 10 lines 9-DEC-1987 08:09 -< Q-Bus on 8550??? >- -------------------------------------------------------------------------------- Where is the Q-Bus in this configuration? The 8550 is a BI bus machine. The TU78 uses a MASSBUS controller which is not on the BI bus. I do not know if you could go BI to UNIBUS to MASSBUS to get it to work or not. It looks like right now the only tape the DEC supports on the BI bus is the TU81. The other way of going is to upgrade the TU78 to a TA78 and put it on an HSC. If your upgrade is going to include an HSC, this would not be a bad way to go but if you are not adding an HSC to start with, it will cost you. ================================================================================ Note 80.2 TU78 on VAX 8550? 2 of 6 EISNER::KENNEDY "Terry Kennedy" 14 lines 9-DEC-1987 20:43 -< Suggest TU79 on HSC >- -------------------------------------------------------------------------------- > The other way of going is to upgrade the TU78 to a TA78 and put > it on an HSC. If your upgrade is going to include an HSC, this would > not be a bad way to go but if you are not adding an HSC to start > with, it will cost you. The above is quite true, but if I were in this situation I'd look at writing off /selling the TU78 and going for a '79 (TA flavor). This is based solely on the reliability of the 78 vs. the 79. The TU81 achieves its performance by streaming, which implies a *long* re-positioning cycle if the CPU can't keep the tape fed. The TU78/79's, on the other hand, achieve at least the same speed, if not 25% or so faster, but do not have the re-positioning delay. As they say, 'your actual mileage may vary'. ================================================================================ Note 80.3 TU78 on VAX 8550? 3 of 6 EISNER::MCDOUGALL "Bob McDougall" 5 lines 9-DEC-1987 22:58 -< TU78 on BI bus? >- -------------------------------------------------------------------------------- > -< Q-Bus on 8550??? >- > > Where is the Q-Bus in this configuration? The 8550 is a BI bus machine. Ooops, my mistake, I don't know where I got Q-Bus from, I meant BI bus, Thanks... ================================================================================ Note 80.4 TU78 on VAX 8550? 4 of 6 EISNER::PRIGOT "Jonathan M. Prigot" 4 lines 10-DEC-1987 14:22 -< TU81E-BA >- -------------------------------------------------------------------------------- While perusing my DECdirect catalogue, I came across something called the TU81-PLUS which if ordered as TU81E-BA will hang directly on the BI. (Reference page 56) It is a 1600/6250 streaming drive, not as fast as your TU78, but a lot cheaper than TA81+HSC, etc. ================================================================================ Note 80.5 TU78 on VAX 8550? 5 of 6 EISNER::PROVOST 10 lines 14-DEC-1987 13:25 -< 3rd party? >- -------------------------------------------------------------------------------- If your 8550 includes a UNIBUS, there are many third party systems available. We use Storage Technology 1953 drives on AVIV controllers. These provide 125 ips 6250 BPI. You can find a wide range of capability at a wider range of costs. There are several sources for comparison information, including DATA PRO REPORTS, HARDCOPY, and DIGITAL REVIEW. Any high-speed device on a UNIBUS on a BI will cause degradation in performance, but BI configurations are severely limited in quality and quantity. Tom ================================================================================ Note 80.6 TU78 on VAX 8550? 6 of 6 EISNER::RENES 9 lines 26-DEC-1987 17:07 -< TU81+ beats TU78 >- -------------------------------------------------------------------------------- We have an 8530 with a TU81+ and in most cases it outperformed our TU78 on the 785 we had. This is primarily due to the fact that the 8530 is much faster and can supply the TU81 enough data to keep it streaming. Since you are going from a 78n to an 8550, you should notice quite an improvement even with a little ole TU81+.  ================================================================================ Note 81.0 VAXmate clock/battery 1 reply EISNER::HORN "Larry Horn" 17 lines 10-DEC-1987 04:27 -------------------------------------------------------------------------------- We've recently installed a few VAXmates and have noticed that the battery-backed clock doesn't always work correctly. The extended self-test will occasionally not find the battery/clock (the "Li" line of the configuration is "--"). On power-up, the time gets set correctly, however on the first soft reset (Ctl/Alt/Del), the time is almost never correct (usually xx-xx-1980). On subsequent soft resets the time apparently randomly is correct or back in 1980. Not all the machines exhibit this problem, but of the ones that do, they are consistent about it. Have any of you seen this before? The suggestion from DEC to contact Field Service will be taken as a last resort -- they are good guys, but as usual with things we get, we are their first site to have VAXmates. [An aside: I HATE to watch FS learn on our equipment!] So if the problem is familiar perhaps you can give me some references to help us find a speedly solution. ================================================================================ Note 81.1 VAXmate clock/battery 1 of 1 EISNER::KENNEDY "Terry Kennedy" 5 lines 12-DEC-1987 17:59 -< National Semi chip? >- -------------------------------------------------------------------------------- Take a look and see if the clock chip is made made by National Semiconductor. I think the part # is MM58?167 or so. I have found any of these chips made in '86 or later to be out of spec. If you have one, replace it with an '84 or '85 date code. I can send you one to try if you like.  ================================================================================ Note 82.0 Interfacing an optical scanner 1 reply EISNER::RICE "Gary Rice" 12 lines 13-DEC-1987 09:55 -------------------------------------------------------------------------------- I am interested in using a Thunderscan optical pattern scanner (made for the MACintosh) in my LA75. I plan to write whatever S/W is necessary to use the device, but I need to know what the signals are that come from the unit as well as how to power it. The people at Thunderware (the manufacturer) are very nice and horribly tight lipped about the pin outs. How should I proceed? Gary ================================================================================ Note 82.1 Interfacing an optical scanner 1 of 1 EISNER::CETRON 8 lines 6-JAN-1988 19:41 -< 'scope it Mr. Scott >- -------------------------------------------------------------------------------- yuck, ick, blech.... I hated our thunderscan... but anyway, I do have a mac disk with the software if that would help, but I would suggest the brute force approach: hook up an oscilliscope and move the print head by hand.... -ed  ================================================================================ Note 83.0 RD53 Problems 8 replies EISNER::MERUSI 15 lines 16-DEC-1987 11:05 -------------------------------------------------------------------------------- RD53 PROBLEMS Has anybody had any problems with RD53's on a uVAX? We have had some inconvenient experiences with two RD53's, both exhibiting the exact same kind of problem. In both situations, as weird as this seems, the error log directory gets blown away! Is there a bad spot on both these disks? In addition, various other aspects of the system begain to fail, Can't even run SHUTDOWM, because VMS can't access NMLSHR ... can't link ofr disk access failure reasons. They just want to replace RD53 unit again but I say there may be something else. Also, does anybody happen to know another about the nature of the formatting action performed by the RD53 diagnostic program? Obviously, there is no Digital service on this system .... appreciate any input, thanx. ================================================================================ Note 83.1 RD53 Problems 1 of 8 EISNER::FRISBIE "Alan E. Frisbie" 16 lines 16-DEC-1987 11:56 -< Drive, controller, software or ??? >- -------------------------------------------------------------------------------- >> Has anybody had any problems with RD53's on a uVAX? We have had >> some inconvenient experiences with two RD53's, both exhibiting >> the exact same kind of problem. From what you describe, it does not sound like an RD53 disk drive problem. Without more detailed information, I would not want to guess what the real cause is. >> In both situations, as weird as >> this seems, the error log directory gets blown away! In what sense is it "blown away"? Are the files in Sys$ErrorLog deleted? Is the directory file itself corrupted? What are the exact error messages you receive? Alan ================================================================================ Note 83.2 RD53 Problems 2 of 8 EISNER::MERUSI 35 lines 16-DEC-1987 15:50 -< More info on RD53 Problem >- -------------------------------------------------------------------------------- < Note 83.1 by EISNER::FRISBIE "Alan E. Frisbie" > -< Drive, controller, software or ??? >- >> Has anybody had any problems with RD53's on a uVAX? We have had >> some inconvenient experiences with two RD53's, both exhibiting >> the exact same kind of problem. From what you describe, it does not sound like an RD53 disk drive problem. Without more detailed information, I would not want to guess what the real cause is. >> In both situations, as weird as >> this seems, the error log directory gets blown away! In what sense is it "blown away"? Are the files in Sys$ErrorLog deleted? Is the directory file itself corrupted? What are the exact error messages you receive? Alan By "blown away" I mean it is gone ..... SYSERR in SYS0 disappears, apparently, ERRLOG.SYS is still around and you can eventually get it back with ANAL/DISK etc. Unfortunately, there is still no data in the log file that might provide you with a hint as to what's going on. Other situations that we've seen are "bad directory format" errors, these will usually occur in whatever area evidentally happens to have activity taking place in it whenever this business happens. The uVAX system itself has 2mb, 1 RD53 and an IEQ-11A. When everything runs it runs great .... but for some reason, this strange event occurs. We always know when it happens because the disk becomes write locked for a short period of time, then it appears to recover and I realize that the error log directory is gone. I know this sounds outrageous but it has happened twice. ERRFMT is not even running after this happens. ================================================================================ Note 83.3 RD53 Problems 3 of 8 EISNER::ERSKINE "Chris Erskine" 4 lines 17-DEC-1987 07:57 -< Controller >- -------------------------------------------------------------------------------- If this is occurring on the same system with both RD53 drives, I would look at the controller. It sounds like you have tried analyzing the disk for errors and it is clean. If your software is not corrupt, then the controller could be mangling the data on write. ================================================================================ Note 83.4 RD53 Problems 4 of 8 EISNER::NAGY "Frank J. Nagy, VAX Guru" 6 lines 17-DEC-1987 08:58 -< Probably controller, and afterwards... >- -------------------------------------------------------------------------------- I also think you should look at the controller. As a side note, after the controller is fixed you should either restore your disks from a *clean* BACKUP or ANALYZE/DISK/REPAIR. We once had a failure in an RM03 disk subsystem (loose Massbus cable) which caused the file system to get clobbered; so just fixing the hardware is not enough. ================================================================================ Note 83.5 RD53 Problems 5 of 8 EISNER::KENNEDY "Terry Kennedy" 5 lines 18-DEC-1987 02:12 -< Some ideas... >- -------------------------------------------------------------------------------- 1) Take a look and see if you have an RQDX2 (as opposed to a 3) in the system. Early uVII's shipped with 2's (I got one). This is a definite no-no in a MicroVAX II. Also, if you have an RQDX3, confirm that the ROMS are 23-243E5 and 23-244E5. As far as I know, these are the latest ROMS. ================================================================================ Note 83.6 RD53 Problems 6 of 8 EISNER::BURNET "John Burnet" 16 lines 3-OCT-1989 22:44 -< RD53/RQDX2 in a MicroVAX II? >- -------------------------------------------------------------------------------- > < Note 83.5 by EISNER::KENNEDY "Terry Kennedy" > > > 1) Take a look and see if you have an RQDX2 (as opposed to a 3) in > the system. Early uVII's shipped with 2's (I got one). This is a > definite no-no in a MicroVAX II. Why is this? We have an old RD53 and RQDX2 that we were thinking of adding to a MicroVAX II we'll be putting together from used parts. The main drive on the system will probably be an Eagle, so we don't really need the RD53, but we thought we'd toss it into the stew since we have no other use for it and the resale value is practically nil. Aside from poor performance relative to the RQDX3, what's wrong with the RD53/RQDX2 combination in a MicroVAX II? ================================================================================ Note 83.7 RD53 Problems 7 of 8 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 15 lines 4-OCT-1989 00:00 -< RQDX2/uV2 ok here... >- -------------------------------------------------------------------------------- > > < Note 83.5 by EISNER::KENNEDY "Terry Kennedy" > > > > > 1) Take a look and see if you have an RQDX2 (as opposed to a 3) in > > the system. Early uVII's shipped with 2's (I got one). This is a > > definite no-no in a MicroVAX II. > > Why is this? ... > > Aside from poor performance relative to the RQDX3, what's wrong with > the RD53/RQDX2 combination in a MicroVAX II? We've been running our RD53/RQDX2 combo (two drives, want to add a third) for YEARS. To us it's not worth upgrading to an RQDX3, as the system is just a test bed (development is elsewhere). ================================================================================ Note 83.8 RD53 Problems 8 of 8 EISNER::KENNEDY "Terry Kennedy" 12 lines 4-OCT-1989 03:40 -< FS memo. >- -------------------------------------------------------------------------------- > Aside from poor performance relative to the RQDX3, what's wrong with > the RD53/RQDX2 combination in a MicroVAX II? Sorry I can't give chapter and verse, but FS had a memo which said "change the RQDX2 to a 3 if the customer complains or you have excessive errors or there is a problem with the RQDX2 not passing NPR grant." It also want on to say that the "RQDX3 is required for a future release of product VMS". This was in the 4.2 timeframe, by the way, so they probably meant V4.4. If you're not having problems, fine, but (at least locally) FS won't support 'em.  ================================================================================ Note 84.0 VAX Processor timings 7 replies EISNER::BYRNE_C "Can you say Hauppauge?" 13 lines 16-DEC-1987 11:49 -------------------------------------------------------------------------------- Can anyone explain these timings. 1 million times in a macro loop, 16 bytes were moved. CPU timing was via LIB$INIT_TIMER etc... 1 MOVC3 2.0 uSec 2 MOVQ 0.9 4 MOVL 4.8 8 MOVW 3.7 16 MOVB 7.8 Why do 4 MOVL's take longer than 8 MOVW's? This is on an 8530, V4.5 Same test done on both uVII and 3600, things come out in the "expected order". ================================================================================ Note 84.1 VAX Processor timings 1 of 7 EISNER::FRISBIE "Alan E. Frisbie" 13 lines 16-DEC-1987 12:01 -< Faster than a speeding turtle! >- -------------------------------------------------------------------------------- >> 1 MOVC3 2.0 uSec >> 2 MOVQ 0.9 >> 4 MOVL 4.8 >> 8 MOVW 3.7 >> 16 MOVB 7.8 >> >> Why do 4 MOVL's take longer than 8 MOVW's? This is just a guess, but I suspect that your buffer was not aligned on a longword boundary. That doesn't explain why a MOVQ is so fast, however. Alan ================================================================================ Note 84.2 VAX Processor timings 2 of 7 EISNER::BYRNE_C "Can you say Hauppauge?" 9 lines 16-DEC-1987 15:41 -< Hmmm Very strange... >- -------------------------------------------------------------------------------- RE: < Note 84.1 by EISNER::FRISBIE "Alan E. Frisbie" > > -< Faster than a speeding turtle! >- > >This is just a guess, but I suspect that your buffer was >not aligned on a longword boundary. That doesn't explain >why a MOVQ is so fast, however. > Nope, it was not only on a lw boundary, but page aligned. ================================================================================ Note 84.3 VAX Processor timings 3 of 7 EISNER::ERSKINE "Chris Erskine" 2 lines 17-DEC-1987 08:00 -< Background? >- -------------------------------------------------------------------------------- What else was happening on the system. This includes background items like LAT or DECnet? ================================================================================ Note 84.4 VAX Processor timings 4 of 7 EISNER::BYRNE_C "Charlie Byrne, Hauppague NY" 16 lines 17-DEC-1987 11:00 -< Processor was loaded, but... >- -------------------------------------------------------------------------------- > RE:< Note 84.3 by EISNER::ERSKINE "Chris Erskine" > > -< Background? >- > > What else was happening on the system. This includes background > items like LAT or DECnet? Lots of stuff was happening on the system. Maybe 60 processes. But since this is cpu, not elapsed, time, is that relevant? These tests were run many times with very similar results, so I don't think it's due to some spike in bus or memory traffic or something like that. I'd be interested if anyone else with an 85xx would run these. The code is obviously simple. ================================================================================ Note 84.5 VAX Processor timings 5 of 7 EISNER::ERSKINE "Chris Erskine" 2 lines 17-DEC-1987 15:30 -< Pay for everything >- -------------------------------------------------------------------------------- Since your process gets charged for context switches ..., yes it does matter. ================================================================================ Note 84.6 VAX Processor timings 6 of 7 EISNER::RENES 13 lines 26-DEC-1987 17:13 -< 8530 performance issue >- -------------------------------------------------------------------------------- In general, we have noticed a "lack" of performance of our 8530 compared to DEC's published performance ratings. i.e, 8530's seem to be about 3x an 8200, rather than 4x, as DEC advertises. We have talked to DEC's 8530 product group... no answer yet! I have talked to at least 3 other sites that have had similar bench mark results on this processor. ...more details to follow as the situation develops... ================================================================================ Note 84.7 VAX Processor timings 7 of 7 EISNER::MERRIMAN "George Merriman - CCA/NY" 15 lines 11-JAN-1988 22:39 -< 8550 results >- -------------------------------------------------------------------------------- > I'd be interested if anyone else with an 85xx would > run these. The code is obviously simple. I ran the test on a quiet 8550 running VMS 4.5. Here are the results: MOVC3: 2.29 microseconds 16 MOVB: 4.73 " 8 MOVW: 2.83 " 4 MOVL: 1.17 " 2 MOVQ: 0.86 " These are average results of 10 interleaved runs of each test program.  ================================================================================ Note 85.0 PC Lap-top's 1 reply EISNER::STAMERJOHN "RW Stamerjohn" 5 lines 17-DEC-1987 13:37 -------------------------------------------------------------------------------- I am looking into a lap-top (PC) for myself. Are there any recomendations? Are there any to be avoided? Besides standard features, I am interest in hooking up Ethernet boards (possible via external box). ================================================================================ Note 85.1 PC Lap-top's 1 of 1 EISNER::HAHN 6 lines 17-DEC-1987 17:05 -< ex >- -------------------------------------------------------------------------------- We have both Toshiba 1100 and NEC Multispeed. They both work fine, if you get the NEC it is worthwhile to get it with Backlit Screen. If you are a WPSPLUS user you should be able to get EBI's WPS PC to run on it. Pierre  ================================================================================ Note 86.0 Need Q-bus help... 8 replies EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" 8 lines 17-DEC-1987 18:50 -------------------------------------------------------------------------------- We build our own enclosures for PDP 11/73's here in our group. The problem I'm currently having is I'm not getting any signals passed between 2 H9270-Q backplanes (this is a disk expansion box we're trying to build). We're using Power One power supplies, but the problem isn't there. I just can't boot from the disk on the expansion box. Anyone have any ideas ? ================================================================================ Note 86.1 Need Q-bus help... 1 of 8 EISNER::KENNEDY "Terry Kennedy" 6 lines 18-DEC-1987 02:18 -< Thoughts... >- -------------------------------------------------------------------------------- 1) I presume you are passing all the signals except the power lines? 2) Grant cards (G7273?) in all empty slots in 1st card cage? 3) Termination properly applied (1st slot in 1st cage, last slot in 2nd?) More information would help... Terry ================================================================================ Note 86.2 Need Q-bus help... 2 of 8 EISNER::HAHN 3 lines 18-DEC-1987 13:33 -< Correct ROM ? >- -------------------------------------------------------------------------------- Does the system boot with the disk on the first cage? If not you may not have the correct Boot ROM. ================================================================================ Note 86.3 Need Q-bus help... 3 of 8 EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" 23 lines 18-DEC-1987 13:36 -< Here's the list of components... >- -------------------------------------------------------------------------------- Here's what I have: 11/73 in first slot (240 ohm term) M9405-yb in eighth slot (240 ohm term) M9404 in 9th slot (no term) M9405-ya in 16th slot (120 ohm term) All other slots filled or have grant cards installed. The H9270-Q has a total of 8 slots. There is a cable connecting the M9405-YB and the M9404. We need to know what signals the M9404/5 pass. Any other ideas?? ================================================================================ Note 86.4 Need Q-bus help... 4 of 8 EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" 1 line 22-DEC-1987 13:55 -< Still need help... >- -------------------------------------------------------------------------------- Yes, the system boots off the first cage with no problems. ================================================================================ Note 86.5 Need Q-bus help... 5 of 8 EISNER::MCGLINCHEY "SANCHO! My armor! My TECO Macros" 8 lines 31-DEC-1987 15:52 -< Dear Alan.... >- -------------------------------------------------------------------------------- Bob - I'm glad you posted this here, as I have come up with no suggestions for you. (I hope Frisbie tunes in) Jim ================================================================================ Note 86.6 Need Q-bus help... 6 of 8 EISNER::FRISBIE "Alan E. Frisbie" 5 lines 2-JAN-1988 16:53 -< Tuned in (Zeeewoooop!) >- -------------------------------------------------------------------------------- >> (I hope Frisbie tunes in) Give me a call at (213) 256-2575 and I will try to help. Alan ================================================================================ Note 86.7 Need Q-bus help... 7 of 8 EISNER::CETRON 12 lines 6-JAN-1988 19:48 -< this is a toughy >- -------------------------------------------------------------------------------- check the ltc lines, I remember some controllers wanted the clock signals also, and some inter-box cables/boards didn't pass them. also, try new cables, we added an expansion box to a 750 and after 6 days work someone suggested the cables (we had tried a replacement with a new one but now tried one that worked elsewhere) and poof, all was well..... -ed ================================================================================ Note 86.8 Need Q-bus help... 8 of 8 EISNER::FRISBIE "Alan E. Frisbie" 5 lines 6-JAN-1988 21:42 -< Problem being worked offline >- -------------------------------------------------------------------------------- We have been working this problem over the phone and have found several problems so far. I will post a note when all of the bugs have been firmly squashed. Alan  ================================================================================ Note 87.0 tape drive hangs system 17 replies EISNER::OSUDAR "John Osudar" 29 lines 21-DEC-1987 15:28 -------------------------------------------------------------------------------- We're having a problem with one of our favorite pieces of DEC hardware, namely our TU77 tape drive. DEC Field Circus hasn't been able to do much for us (maybe because the problem doesn't happen often enough to be considered "reproducible" -- just often enough to be highly annoying!) We are wondering whether anyone else has seen this type of behavior on their system with a TU77: Basically, here's what happens: while in the middle of a tape reel, it may become necessary to take manual control of the tape drive. (This has happened under several circumstances. In one case, the drive lost vacuum in the middle of a reel, so we had to re-load it; in another case whose details I've largely forgotten, a rewind had to be initiated manually.) It often happens (but not always) that the system goes into a strange state while the tape rewinds -- for all intents and purposes it appears to be completely hung. Once the tape is done, the system comes back to life, including remembering some (if not all) of what may have been typed (but not echoed) while the system appeared to be hung. Of course, by the time the system wakes up, a lot of things have timed out -- including network connections and LAVC (it's a lot of fun watching all your LAVC nodes "voluntarily exit the VAXcluster" at the same time!) Other than applying band-aids to the effects (e.g. increasing RECNXINTERVAL) we haven't been able to deal with the situation in any positive way. Nobody seems to know why this is happening, or what we can do about it (although someone did mention something about tuning the TU77 with a sledgehammer :-) Any ideas (other than getting rid of the TU77)? ================================================================================ Note 87.1 tape drive hangs system 1 of 17 EISNER::KENNEDY "Terry Kennedy" 10 lines 21-DEC-1987 18:18 -< 1 - FS adjustment, 2 - SPR >- -------------------------------------------------------------------------------- Mid-reel unloads are a problem on 77's and 78's. There is an adjustment (hard) that Field Service can do to correct this. I believe it is called 'park loops' or some such. Off-lineing a tape drive in the middle of an operation is unpleasant at best. The driver should return 'fatal drive error' to VMS. If it does not, I suspect you've got a buggy driver. Feel free to SPR it (and don't forget to put it in SPR_Directory). ================================================================================ Note 87.2 tape drive hangs system 2 of 17 EISNER::PRIGOT "Jonathan M. Prigot" 4 lines 22-DEC-1987 09:15 -< MASSbus adapter too >- -------------------------------------------------------------------------------- Since you say that the system hangs, I would also look for some strange interaction between the TU77 and the DW7x0 MASSbus adapter. Since the DW7x0 sits on the CPU backplane, I would think it a key player in the problem. ================================================================================ Note 87.3 tape drive hangs system 3 of 17 EISNER::HASSINGER "Bob Hassinger" 15 lines 22-DEC-1987 11:36 -< Sounds a little like something we saw once... >- -------------------------------------------------------------------------------- It was a long time ago - can't remember details - but I do recall a situation involving tapes (TU77 in fact) that caused a serious system problem. Maybe what little I can recall will give someone an idea: I think in this case it had to do with a user who was doing a large write of a tape unexpectedly needing to mount a second reel. As I recall there was a problem getting a message out that caused the system to more or less "hang". It may have been that the message was supposed to go to OPA0: via OPCOM but OPCOM was in a state where it's output was not going to any terminals (a bug in the way we were using stuff from KMSKIT as I recall). I just can't remember the details now but basically it was problems getting a message out to a terminal or maybe it was going through the error logging system. Bob H ================================================================================ Note 87.4 tape drive hangs system 4 of 17 EISNER::FRISBIE "Alan E. Frisbie" 7 lines 22-DEC-1987 12:44 -< From my dusty memories... >- -------------------------------------------------------------------------------- I remember TU77 tape drives hanging the system as far back as VMS v2.x. As I recall, it usually happened when I would manually rewind the tape at the same time the system was trying to dismount it. My memory is very vague, but I seem to recall it in VMS v3.x also. Alan ================================================================================ Note 87.5 tape drive hangs system 5 of 17 EISNER::HANRAHAN "Jamie Hanrahan" 9 lines 22-DEC-1987 16:13 -< I remember it too... >- -------------------------------------------------------------------------------- I also remember this happening with TE16s (which use the same Massbus controller -- TM03). The explanation I got went something like this: When you take the drive offline, the VAX is interrupted, and the driver tries to read the drive status, but it can't because it's offline; it continues trying until you put the drive online again. Or something like that. This happened with two different TE16s and two different TM03s (all four combinations). I think it's a design flaw in the driver rather than a hardware problem. ================================================================================ Note 87.6 tape drive hangs system 6 of 17 EISNER::PROVOST 32 lines 22-DEC-1987 16:25 -< FCO & PCO addressed this problem >- -------------------------------------------------------------------------------- Going back to the early days, there were problems. If one pushed off-line on a TU77 and manually rewound it, the system would hang until rewind complete and drive set on-line. Variations occurred over time. These variations included times when the VAX would hang anytime the TU77 was rewinding, and times when the system didn't come back unless the TU77 was put back on-line before rewind complete. Fixes included software patch or new release, and hardware FCO to the controller. I believe the driver was awaiting the rewind complete interupt, which of course never came. Since the driver was at interupt priority, the system was on coffee-break. Once the hardware was all updated, I think the software was unfixed. This is because the software fix would not have been necessary if the hardware worked. If this is all true, then one of three things may be true: 1. Your hardware has not been updated. 2. Your software is still on VMS 2.0 or there abouts. 3. Both were updated, but the hardware is broken. This last case is most likely. If it is true, the area to look for the broken hardware is the area affected by the FCO to eliminate the rewind interrupt problem. Have your FS man check it. Your might also check to see if the VAX hangs anytime you do a manual rewind on a mounted tape. (There was even a time when manual rewind was the work-around for system hangs on programmed rewind!) Lastly, someone mentioned the DW interface. I believe your drive is connected via MASSbus, not UNIBUS. ================================================================================ Note 87.7 tape drive hangs system 7 of 17 EISNER::PRIGOT "Jonathan M. Prigot" 4 lines 23-DEC-1987 09:28 -< My failing memory >- -------------------------------------------------------------------------------- (All my MASSbus knowlege left when we traded in our 11/750 for an 8250.) Provost is correct, a DW interface is Unibus, the part I was thinking of was an RH7x0 MASSbus adapter. (I don't see any included in my diagram of an 86x0. Do they use an RH780 or what?) ================================================================================ Note 87.8 tape drive hangs system 8 of 17 EISNER::PROVOST 8 lines 23-DEC-1987 09:42 -< No MASSbus on BI >- -------------------------------------------------------------------------------- > was thinking of was an RH7x0 MASSbus adapter. (I don't see any included > in my diagram of an 86x0. Do they use an RH780 or what?) There is no MASSbus adapter for BI! If you must interface MASSbus equipment to BI machines, there is a product which interfaces MASSbus to UNIBUS, and there is a UNIBUS adapter for a BI, although it is is the slowest UNIBUS DEC ever made. Tom ================================================================================ Note 87.9 tape drive hangs system 9 of 17 EISNER::KILLEEN "Jeff Killeen" 2 lines 23-DEC-1987 10:37 -< KLUDGE >- -------------------------------------------------------------------------------- MASSBUS to UNIBUS adapter then a UNIBUS to BI adapter? Amazing it runs at all! ================================================================================ Note 87.10 tape drive hangs system 10 of 17 EISNER::PRIGOT "Jonathan M. Prigot" 5 lines 23-DEC-1987 15:58 -< 86xx BI? >- -------------------------------------------------------------------------------- At the risk of diverting this note too much, is the 86xx a BI machine? I see no mention of BI gear for it. SBI, to be sure, but no BI. In any case, the 86xx does use RH780s attached to the SBI adapter(s). By the way, what flavor VAX is the drive connected to? ================================================================================ Note 87.11 tape drive hangs system 11 of 17 EISNER::OSUDAR "John Osudar" 34 lines 23-DEC-1987 16:39 -< reply to other replies >- -------------------------------------------------------------------------------- The problem I described in the original note occurs on an 11/785 system; the TU77 is on the fourth Massbus adapter, which it shares with a TU78. The TU77 and controller have been FCO'ed to, as far as we know, current rev levels. We had the midreel offlining problem before, too -- there apparently was a bug in the driver, which I believe has been fixed. (Not to say that was the ONLY bug in the driver! :-) The problem is that DEC can't make up their minds whether this is a hardware problem or a software problem. All we know is that it IS a problem... OPCOM and OPA0: don't have anything to do with this particular problem, although console problems have been known to cause system hangs under other circumstances. We are running VMS V4.6, so I don't think obsolete software is the answer. The explanation proposed by Jamie (driver waiting at high IPL for something to happen that DOESN'T happen until rewind completes) sounds plausible; it certainly agrees with observed behavior. The possibility of broken hardware (in the area affected by the rewind interrupt FCO), as suggested by Tom, also sounds plausible. I am hoping that our Field Service people know enough to figure out what to check -- but that may be expecting too much. I will try manually rewinding a mounted tape (some other time, when there aren't 60 users logged in!) (Finally, as the tail end of the replies concluded, the 86xx are not BI machines, and do use RH780's. Remember that a lot of DEC internal stuff refers to the 86xx's as "790's"!) ================================================================================ Note 87.12 tape drive hangs system 12 of 17 EISNER::HANRAHAN "Jamie Hanrahan" 41 lines 23-DEC-1987 18:15 -< Suggested research technique >- -------------------------------------------------------------------------------- > The explanation proposed by Jamie (driver waiting at high IPL for > something to happen that DOESN'T happen until rewind completes) > sounds plausible; it certainly agrees with observed behavior. Here's a way to find out. Next time the system gets into this state, go to your system console and hit ^P and type H for halt. Observe the displayed PC. Type C for continue, then H for halt again, observe PC, repeat until the rewind is almost done. Then let the machine run and do SET TERMINAL PROGRAMMABLE [disclaimer: I've never touched an 86xx; I'm assuming that its console command language is like the 780's; perform appropriate translations if not]. If you have time, display registers R3, R4, and R5 and the PSL (which includes the hardware IPL value in bits 16 through 20) also. (I'm sure the young-uns among you -- yeah, you! The ones who've never touched a real switches-and-lights console -- are chuckling at the thought of using such a crude technique to find out anything. Trust me; it works. When VMS gets into a tight loop at high IPL it's almost always a VERY tight loop, and this trick has pointed me to the solutions for several different problems.) When the system comes back, get into SYSGEN and enter: SYSGEN> SHOW/DEVICE=MT This will give you the start and end address of the MT driver, which bosses the TM03 and all its children. If most of the displayed PCs are within those bounds, you have a clue. Now go look in the driver source to see what it's doing. You can also do the following in SDA: SDA> EXAMINE/INSTRUCTION pc_val where pc_val is one of the PC values you got from the console. SDA will, in addition to disassembling the instruction, interpret the instruction's address as an offset from the nearest system global symbol or driver start address, assuming that the offset is in the range 0 through FFF (hex). Magtape drivers are large beasts, though, so this may not be helpful. ================================================================================ Note 87.13 tape drive hangs system 13 of 17 EISNER::PROVOST 6 lines 28-DEC-1987 10:29 -< CRASH might help >- -------------------------------------------------------------------------------- Similarly, you could force a crash during rewind while the system is hung. Be sure to stuff FF into the priority of the PSL. (Easier to stuff FFFFFFFF, or -1) Analyze/crash should then give you the state of the driver. This has the disadvantage of killing everyone off, but it has the advantage of giving you plenty of time to analyze the state of the system. Tom ================================================================================ Note 87.14 tape drive hangs system 14 of 17 EISNER::FULLER_S "Stuart Fuller" 4 lines 27-APR-1988 17:17 -< Still broke? >- -------------------------------------------------------------------------------- Well, it's been 4 months. Any resolution to this hanging problem? Stu ================================================================================ Note 87.15 tape drive hangs system 15 of 17 EISNER::OSUDAR "John Osudar" 5 lines 27-APR-1988 18:10 -------------------------------------------------------------------------------- Nope, no solution (unless you want to consider avoiding the use of the TU77 as a solution! ;-{) And, since we're looking at trading in our MASSBUS junk for some more modern hardware, we may never get a solution. (Plus, we no longer have DEC hardware maintenance. Hmmm... maybe now we CAN get it fixed...) ================================================================================ Note 87.16 tape drive hangs system 16 of 17 EISNER::FULLER_S "Stuart Fuller" 7 lines 28-APR-1988 14:31 -------------------------------------------------------------------------------- There was an FCO to the TM03 (tape formatter) that cured problems with the system hanging during an unsolicited rewind. The FCO was to the M8909-YA module, and was called FCO M8908-YA R006. If you'll check the M8909-YA module,{on the back of the handles, you should see a revision letter stamp. It should be 'K' or higher. This FCO was required for VMS V3.0 (it's an _old_ FCO!). ================================================================================ Note 87.17 tape drive hangs system 17 of 17 EISNER::BRUCKER "Ken Brucker" 24 lines 2-MAY-1988 13:06 -< We have seen this too >- -------------------------------------------------------------------------------- We have had this problem twice so far in the last couple of months. The first occurance was under vms 4.6 on a 780. We have a TK50 connected via TUK-50 controller on the Unibus. (Yes Ma a TK50 on a 780) We have exerienced a number of problems the worst of which was when the ACP went into a loop (or whatever it did, we still have an SPR outstanding on it) and "hung" the system. We could work but only *V*E*R*Y* *S*L*O*W*L*Y*. The usual problem is having the device hang and the only solution we have come up with is rebooting the 780 to get things working again. Then this weekend we had a TA78 connected to an HSC50 (v350 software) give us a headache. Our 8650 (vms 4.7) went out to lunch. I'll be sending the SPR off for it today. And about ancient history....we used to have a TE16 connected to the massbus in a 780, that would cause the 780 to repeatedly lose and regain connection to the cluster when the simple operation of placing the drive ONLINE was performed. This had FS going crazy, and until I showed our FE that I could do it with extremely high reliability, he was unwilling to believe that it was happening. But, in our favor(?), that worked into a good justification for new HSC connected tape drives. We never did get the problem resolved, just made it go away.  ================================================================================ Note 88.0 VR260 documentation needed 12 replies EISNER::FRISBIE "Alan E. Frisbie" 14 lines 22-DEC-1987 22:10 -------------------------------------------------------------------------------- Does anyone have a print set and/or maintenance manual for the VR260 monitor (for VAXstations) that I can borrow for a few days? My VR260 is displaying a rastor that is varying between a half and a tenth the normal height. It should be pretty easy to find the problem, but I don't like to get into a piece of equipment without the documentation. This is not a critical problem, since I have a backup monitor here. (Remember what started the Swap_Meet discussion?) Alan Frisbie (213) 256-2575 ================================================================================ Note 88.1 VR260 documentation needed 1 of 12 EISNER::MCCORMICK_B "Bob McCormick" 9 lines 28-SEP-1988 22:02 -< Indirect answer... >- -------------------------------------------------------------------------------- I looked around and couldn't find what you need. BTW, when DEC performs alignment on these monitors it has to be done in some kind of faraday cage/field (memory serves me that it was a aluminum cube frame about 8 feet on a side wound with wire) so that when the metal flap/door is open to make the adjustments you get the correct effect when its all put back together. I think you will only be able to make rough adjustments in the field... ================================================================================ Note 88.2 VR260 documentation needed 2 of 12 EISNER::MAYHEW "Bill Mayhew" 32 lines 2-MAY-1989 17:28 -< VR260 needs tuning >- -------------------------------------------------------------------------------- Well, I finally got off my butt and got my VR260 "fixed", sort of. (See 336.4, roughly, for a general problem statement). The shrinking and growing raster got a lot worse in recent days such that starting about 5 minutes after power up, it would *rhythmically* shrink and grow for a half-hour or more. A great argument for an early-morning coffee break, but otherwise not too useful ;-} ... Anyway, when DEC responded (at 4pm Tues to a 9:20am Monday call -- they seem to have personnel-allocation problems), they swapped out the deflection card and now I have another, slightly less irritating problem. The screen image is distorted; characters on the left are wider than those on the right, the image "bows" somewhat at both the left and right extremes. I was able to persuade the tech to spend about 10 minutes or so trying to adjust it to meet my standards but he eventually gave up and ordered another board; given the LARGE number of adjustments that seem to be possible on this thing, I'm not sure that approach isn't simply deferring the inevitable. He claims the deflection boards are supposed to come through pre-adjusted. This one, when plugged in, gave us a raster that was about half its normal height, so it needed *some* adjustment from the word "go". I have done a fair amount of adjusting of video equipment (esp. consumer TV and dumb-terminal monitors) and understand the concepts, but have no VR260-specific experience. Is alignment difficulty common on this monitor? What chances do I have of getting it "right" again? (This is partly a question of "how far can I push Field Service to fix this before they can *justifiably* say 'that's as good as it gets'?") Experiences/war stories welcome. ================================================================================ Note 88.3 VR260 documentation needed 3 of 12 EISNER::MADISON "Matt Madison" 11 lines 2-MAY-1989 21:09 -< "squished" is typical >- -------------------------------------------------------------------------------- > He claims the deflection boards are supposed to come through > pre-adjusted. This one, when plugged in, gave us a raster that > was about half its normal height, so it needed *some* adjustment > from the word "go". That's typical. I don't think I've seen one deflection board come in that didn't need some height adjustment (at least), and we've gone through dozens. We once got brave and tried fiddling with some of the other adjustments, without much success, partly (I think) because we didn't let the new boards "burn in" long enough before "adjusting" them. ================================================================================ Note 88.4 VR260 documentation needed 4 of 12 EISNER::KENNEDY "Terry Kennedy" 9 lines 2-MAY-1989 22:28 -< A fix is coming soon... >- -------------------------------------------------------------------------------- > ... and we've gone through dozens. I picked up two VR260's for the express purpose of blowing them up to see what goes wrong with them. Obviously, something on that board is sensitive to voltage/current surges (under-engineered). When I get back from the Sym- posium (and moving the week after), I'll tear into them and post the info for all you daring people so you can fix 'em yourselves. That way, it makes more sense to spend time tweaking the other adjustments if you know you'll never have to swap the board again. ================================================================================ Note 88.5 VR260 documentation needed 5 of 12 EISNER::MCCORMICK_B "Bob McCormick" 13 lines 4-MAY-1989 18:54 -< VR260 adjustment was difficult >- -------------------------------------------------------------------------------- Bill... I don't know where I posted it ... but yes, alignment of the VR260 is something that is (was?!) very difficult. These were made in Albuquerque where they had an aprox 6 foot cube aluminum frame wound with wire. The CRT is aligned face down with a video camera aimed at it in the center of the cube ... the wires have current flowing to help compensate for some problem with the covers off/open during adjustments. This info was true when the VR260 was first released (VAXstation I & II QVSS) and I was doing work in the ABQ plant. NB Any attempts I tried with VR260 adjustments in the field could never come close to what was done in mfg with that setup. Good luck. ================================================================================ Note 88.6 VR260 documentation needed 6 of 12 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 31 lines 5-MAY-1989 00:48 -< The only VR260 info I could find >- -------------------------------------------------------------------------------- >> Is alignment difficulty common on this monitor? What chances >> do I have of getting it "right" again? (This is partly a >> question of "how far can I push Field Service to fix this >> before they can *justifiably* say 'that's as good as it >> gets'?") The only source of alignment/repair information I have found for the VR260 is in the technical manual that came with the VAXstation. This is the nine inch high cloth-covered binder in the box-like case. Mine is titled "VAXstation II Technical Manual, BA23 Enclosure", order number AZ-GLFAB-MN. The VR260 info is at the end of chapter 6, "FRU Removal and Replacement". It isn't much, but it is better than nothing (but not very). I have managed to tweak mine to a reasonable state, except for a slight rotation of the rastor. Since the yoke appears to be glued to the neck of the CRT, I am reluctant to twist it very hard. The only remaining problem is an intermittant jiggle of the entire raster. It is only about a scan line or three in magnitude, just enough to make it annoying. It lasts for an hour or two and then goes away for a week or so. I leave mine on 24 hours a day except when I will be away for more than two or three days. In 3.5 years I have had two deflection boards go bad, so I keep a spare monitor. Despite the problems, I would be utterly lost without it. Having 55-line edit windows has really spoiled me. ================================================================================ Note 88.7 VR260 documentation needed 7 of 12 EISNER::BYRNE_C "Charlie Byrne" 6 lines 5-MAY-1989 00:56 -< To save others time... >- -------------------------------------------------------------------------------- > < Note 88.5 by EISNER::MCCORMICK_B "Bob McCormick" > > -< VR260 adjustment was difficult >- > I don't know where I posted it ... $SEARCH DECUSERVE_ALL_TITLES: VR260 only came up with HARDWARE_HELP 88 and 336 ================================================================================ Note 88.8 VR260 documentation needed 8 of 12 EISNER::MAYHEW "Bill Mayhew" 27 lines 5-MAY-1989 13:59 -< VR260 saga continues >- -------------------------------------------------------------------------------- Thanks for the doc reference, Alan. No such book comes with the 2000, so I'll have to order one. I'm beginning to get nervous about relying on DEC Field Service for things that can't be solved by a simple swap. On Wednesday, after being unable to produce a reasonable facsimile of a square raster after replacing the deflection board, DEC swapped in a new VR260, watched for a few minutes, we agreed it was "reasonably" square, and they went away. Unfortunately once it was on for a few hours, the raster behaved more like a rubber sheet stretched at its corners... MUCH worse than the original. Power cycles and cool-downs indicate that it is a heat-related problem. Thursday they came back with two technicians and tweaked everything within sight, and called Atlanta, which recommended replacing the system board (this is on a VS2000). I don't quite buy that... and buy it even less, since Thursday night on three occasions the screen just "popped" off for about two seconds and then popped back on again. Only the screen was affected; I didn't have the presence of mind to check the power LED on the front, but it's not due to a loose power cord. On the basis of that I was able to persuade them today that another VR260 is required, and they are ordering one. Maybe Monday. Meanwhile I already mailed back my Field Service survey with the question about "how satisfied are you with the technical competence of the engineer" and "...time to repair" questions . ================================================================================ Note 88.9 VR260 documentation needed 9 of 12 EISNER::MAYHEW "Bill Mayhew" 12 lines 15-MAY-1989 16:18 -< ECO coming? >- -------------------------------------------------------------------------------- After 1 deflection board, 2 whole VR260-swaps, and several hours of DEC technicians tweaking things, plus 2 weeks elapsed time since my first call, I still have a monitor that is (now) stable, but that does not offer straight lines in either the horizontal or vertical direction. Friday I got a call from the local F/S unit manager who states that there is an ECO which she is trying to get, and she is in contact with Engineering about the problem. I don't know if this is related to the failure-on-power-cycle problem in 336.* or not, but as soon as I get the info on the ECO (and see that it works!) I will pass it on. ================================================================================ Note 88.10 VR260 documentation needed 10 of 12 EISNER::MAYHEW "Bill Mayhew" 9 lines 2-JUN-1989 00:15 -< If you can't fix it... >- -------------------------------------------------------------------------------- On or about May 23, I spoke with the F/S unit manager again, who now states that the fix for the VR260 is called a VR262. ;-} I intend to hold DEC's feet to the fire until I get *something* that works. What I can't fathom is why, all of a sudden, all VR260s DEC seems to be able to ship are exhibiting this "gull-wing" distortion problem. Allegedly, the VR262 will be available this month (June '89). ================================================================================ Note 88.11 VR260 documentation needed 11 of 12 EISNER::MAYHEW "Bill Mayhew" 27 lines 22-JUN-1989 15:49 -< VR262 initial impressions >- -------------------------------------------------------------------------------- I am now the (proud?) owner of a new VR262. There are trade-offs... The "gull-wing" distortion problem on the VR260 (which would cause lines of text across the screen to be non-straight, in a gull-wing pattern) is gone, but the corners of the display have some distortion; they are "pushed in" about 1/16" from where they would be if you drew truly straight lines around the display area. Since F/S has to come back tomorrow to replace the RD32 in this VS2000, we're going to let it sit and burn in for a while before starting to tinker with it. Other attributes of the VR262 are that it now comes with a "champagne beige" bezel right around the screen (the overall case is still off-white). Also I'm told that there is no tilt-swivel base for the VR262 available. I didn't have one on the VR260, so this didn't bother me too much. Finally, the VR262 has the "paper white" phosphor. Overall the image looks sharper but this may be an issue of better focus adjustment, more than anything else. Time will tell if my eyes like the new phosphor better, or not. Evidently this is one of the, if not _the_, first VR262s to be installed in this area; Field Service was anxious enough to get it in here that they scheduled a special trip to install it, even though the same guy is coming tomorrow to swap the RD. ================================================================================ Note 88.12 VR260 documentation needed 12 of 12 EISNER::MAYHEW "Bill Mayhew" 34 lines 9-AUG-1989 12:12 -< VR262 isn't a good solution, yet >- -------------------------------------------------------------------------------- I thought I had posted some of this before, but... I've now gone through 2 (or is it 3?) VR262s since the previous note was posted. Problems include spontaneous power-downs for a fraction of a second (though this behavior was *primarily* confined to one unit, it has been seen on the other(s)), and distortion in the image. Basically, anything within about a 1" border around the edge of the VAXstation's "background" display is distorted: horizontal lines bend up or down, depending on whether you're looking at a line that's below or above the center-line, respectively; vertical lines aren't straight either (though not as bad as horizontal). DEC brought a support engineer from the Desktop Systems Group of CSSE in Westford out to my site to look at the problem. He had an interesting curved-plexiglass (I'm guessing) thing that matches the curvature of the face of the CRT, with some lines etched into it that allows him to do some alignment tests... but unfortunately the VS2000 doesn't provide the circle/crosshatch display he needed to really do it. He did verify that the problem exists, and we looked at it on 2 monitors. Another one was ordered, and tested at the local F/S office, and it too is distorted. DEC engineering (supposedly "the top guy") from Albuquerque has been flown in to look at the problem and they concur that a problem exists, but they believe it is a manufacturing and/or set-up problem, not an engineering problem. They claim they're going back to review all the set-up procedures and find a fix, and I'm told I'll get a "good" monitor within a month. So... progress is being made, inexorably slowly. If this were a critical problem, I'd hope DEC could respond more quickly. As it is, I'm just losing all sense of what's straight and what isn't (Martha, why does Peter Jennings look like a Picasso??).  ================================================================================ Note 89.0 VAX <==> Apple Laserwriter 10 replies EISNER::RICE "Gary Rice" 5 lines 6-JAN-1988 14:19 -------------------------------------------------------------------------------- I would like to connect an APPLE Laserwriter Plus to a VAX via an RS-232 terminal line. Can it be done? What do I need in the way of hardware to accomplish that? Gary ================================================================================ Note 89.1 VAX <==> Apple Laserwriter 1 of 10 EISNER::CETRON 6 lines 6-JAN-1988 19:56 -< easy as pie (apple that is) >- -------------------------------------------------------------------------------- no hardware required at all, matter of fact, I recall it is even documented in the Hardcopy UIS description in the vaxstation manuals. -ed ================================================================================ Note 89.2 VAX <==> Apple Laserwriter 2 of 10 EISNER::NAGY "Frank J. Nagy, VAX Guru" 2 lines 6-JAN-1988 20:43 -< Required hardware :-) >- -------------------------------------------------------------------------------- One piece of hardware is certainly required: an RS232 cable! :-) ================================================================================ Note 89.3 VAX <==> Apple Laserwriter 3 of 10 EISNER::HYDE "When all else fails,hug your teddybear" 7 lines 7-JAN-1988 07:16 -< Cable and queue >- -------------------------------------------------------------------------------- The DEC cable is a BCC05, your basic null modem cable. That and a basic printer queue setup and your off and running. I just talked a customer through the whole process on the phone yesterday. mark ================================================================================ Note 89.4 VAX <==> Apple Laserwriter 4 of 10 EISNER::RICE "Gary Rice" 9 lines 7-JAN-1988 11:43 -------------------------------------------------------------------------------- > -< Cable and queue >- > The DEC cable is a BCC05, your basic null modem cable. > That and a basic printer queue setup and your off and running. Thanks for all the encouragement. I was initially nervous because the MAC interface says that Apple-talk hardware is required. Gary ================================================================================ Note 89.5 VAX <==> Apple Laserwriter 5 of 10 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 9 lines 19-APR-1988 18:51 -< Look at PacerPrint >- -------------------------------------------------------------------------------- Connecting the printer is the easy part. What you probably really want to do (or at least what we wanted to do) was to be able to use the printer for printing from all of our applications not just those that blat out PostScript. As it happens, a company called Pacer Software sells a modified VMS print symbiont that will do on the fly translations of text files, automatically recognize a PostScript file (if it has a %!PS_ADOBE as rec.1), translate several different graphics protocols to PostScript etc. We have about 40 PostScript printers and a pretty happy with the results. ================================================================================ Note 89.6 VAX <==> Apple Laserwriter 6 of 10 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 6 lines 20-APR-1988 09:38 -< Use of setup modules can help, too. >- -------------------------------------------------------------------------------- We don't even use a modified symbiont. We just have setup modules for the forms we use that are Postscript programs directing the printer to handle the stream as we wish. For example, our Postscript printers handle straight ASCII files because our setup module is a program to do so. We also have a form that doesn't specify a setup module so that person can print a Postscript file directly. ================================================================================ Note 89.7 VAX <==> Apple Laserwriter 7 of 10 EISNER::HASSINGER "Bob Hassinger" 1 line 20-APR-1988 11:19 -< It would be VERY nice of you would make them available... >- -------------------------------------------------------------------------------- ================================================================================ Note 89.8 VAX <==> Apple Laserwriter 8 of 10 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 5 lines 20-APR-1988 15:45 -< OOOooooo NNooooo! It's against POLICY! >- -------------------------------------------------------------------------------- RE: < Note 89.7 by EISNER::HASSINGER "Bob Hassinger" > -< It would be VERY nice of you would make them available... >- I'd like to, but Killeen would get on my butt because they're programs! :-) ================================================================================ Note 89.9 VAX <==> Apple Laserwriter 9 of 10 EISNER::HASSINGER "Bob Hassinger" 1 line 20-APR-1988 16:20 -< But newsletters, Library and SIG tapes are all OK... >- -------------------------------------------------------------------------------- ================================================================================ Note 89.10 VAX <==> Apple Laserwriter 10 of 10 EISNER::BRYANT "Geoff Bryant" 6 lines 14-APR-1989 09:19 -< Any other ways to PRINT to Apple LaserWriter? >- -------------------------------------------------------------------------------- We have a customer who wants to solve the very problem mentioned in this thread - setting things up to PRINT/QUE=Apple_Laserwriter. Did the setup modules from .6 make their way onto a SIG tape? Is there another product besides PacerPrint mentioned in .5?  ================================================================================ Note 90.0 TK50, TK70 & THE VAXstation 3200 9 replies EISNER::KILLEEN "Jeff Killeen" 1 line 7-JAN-1988 12:23 -------------------------------------------------------------------------------- Does anyone know why the TK70 works with the 3200 but the TK50 does *NOT*? ================================================================================ Note 90.1 TK50, TK70 & THE VAXstation 3200 1 of 9 EISNER::HANRAHAN "Jamie Hanrahan" 14 lines 8-JAN-1988 21:16 -< Maybe it will. >- -------------------------------------------------------------------------------- In Note 90.0, EISNER::KILLEEN ("Jeff Killeen") writes: > Does anyone know why the TK70 works with the 3200 but the TK50 does *NOT*? Are you sure it doesn't work? Or is it just that DEC *claims* it doesn't work? I ask because I know of a site that got a 3600 on (very short-term) loan; while they had it, they decided to plug a DMV11 (which DEC claims won't work) into the Qbus, hang an analyzer on the bus, and see just what the problem was (they make Qbus hardware and wanted to avoid whatever the DMV's problem was). Lo and behold, they were unable to make the DMV fail. Is it that the diagnostics aren't ready? Or that DEC simply doesn't have DMVs with the funny PC-like bracket ready yet? Or is this a "marketing" decision...? ================================================================================ Note 90.2 TK50, TK70 & THE VAXstation 3200 2 of 9 EISNER::KILLEEN "Jeff Killeen" 6 lines 8-JAN-1988 22:37 -< SOUND LIKE IT DOES NOT WORK >- -------------------------------------------------------------------------------- DEC just sent out a note on RONNIE that it could not be done and will be fixed in the future. Remember the 3200 is in a standard BA23 box just like an MVII. It does *NOT* need the funny handles. ================================================================================ Note 90.3 TK50, TK70 & THE VAXstation 3200 3 of 9 EISNER::ERSKINE "Chris Erskine" 2 lines 11-JAN-1988 06:46 -< Marketing strikes again >- -------------------------------------------------------------------------------- Or is this the standard Marketing ploy? Do not support this device and we can sell more of the new ones. ================================================================================ Note 90.4 TK50, TK70 & THE VAXstation 3200 4 of 9 EISNER::KILLEEN "Jeff Killeen" 14 lines 19-JAN-1988 11:08 -< FYI >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! TK50 NOT SUPPORTED AS OPTION ON VAXSTATION 3200 - 07-January-1988 The TK50 tape drive cannot be supported as an option on the VAXstation 3200 at this time. Customers requiring immediate shipment of a VAXstation 3200 with tape media must purchase the new TK70-AA tape drive and the TQK70-AA controller module. ================================================================================ Note 90.5 TK50, TK70 & THE VAXstation 3200 5 of 9 EISNER::KOZAM 15 lines 19-JAN-1988 20:57 -< Does 'can not' = 'will not'{ >- -------------------------------------------------------------------------------- >.___.___.___.___.___.___.___. >! ! ! ! ! ! ! ! >| d | i | g | i | t | a | l | N e w s >!___!___!___!___!___!___!___! > >The TK50 tape drive cannot be supported as an option on the VAXstation >3200 at this time. Why? Should we read the above line as 'cannot' or as 'will not'? There is a BIG difference in regard to the implications for other controllers. Marc Kozam ================================================================================ Note 90.6 TK50, TK70 & THE VAXstation 3200 6 of 9 EISNER::HAHN 64 lines 31-MAY-1988 23:08 -< Bad batch of TK50 >- -------------------------------------------------------------------------------- The following was received at SUNY at Stony Brokk; use as you wish; it is not an endorsement from me, but a caution to users of TK50 tapes. Pierre From: IN%"UHAC010%vaxa.rhbnc.ac.uk@NSS.Cs.Ucl.AC.UK" 28-MAY-1988 06:15 To: INFO-VAX <@NSS.Cs.Ucl.AC.UK:INFO-VAX@kl.sri.com> Subj: Return-path: "UHAC010%vaxa.rhbnc.ac.uk" <@KL.SRI.COM:UHAC010%vaxa.rhbnc.ac.uk@NSS.Cs.Ucl.AC.UK> Received: from KL.SRI.COM by ccmail.sunysb.edu via TCP; Sat May 28 06:15 EDT Received: from NSS.Cs.Ucl.AC.UK by KL.SRI.COM with TCP; Tue 24 May 88 06:08:23-PDT Received: from vaxa.rhbnc.ac.uk by NSS.Cs.Ucl.AC.UK via Janet with NIFTP id aa04940; 24 May 88 12:04 BST Date: 24-MAY-1988 12:03:20 GMT From: UHAC010%vaxa.rhbnc.ac.uk@NSS.Cs.Ucl.AC.UK To: INFO-VAX <@NSS.Cs.Ucl.AC.UK:INFO-VAX@kl.sri.com> Warning to all TK50 users in the UK (and maybe stateside too): We have recently had problems with TK50 drives hanging (yes I know it happens all the time). The engineer told us that there is a bad batch of tapes around and we had thirteen of them. The bad tapes slough off oxide into your drive... The bad tapes are identified from the two numbers printed (usually in purple ink) on the bottom of each tape (ie the side with the screwheads showing). Typically: 92802 17102 - -- If you have an eight in the underlined position on the first number, and a ten in the underlined position in the second number then you have a problem. Note that sometimes the second number has an extra digit in the LSD - you should look in the third and fourth positions from the right. If you contact DEC they will swap the tapes for good ones. TK50 tapes also clean the drive. If you have had a bad tape in your drive it will be dirty, and this may make a good tape appear faulty. However, after a few runs through, the good tape should clean the drive up and all will be well. ------------------------------------------------------------------------------ Adrian Johnstone, Mail: Royal Holloway & Bedford New College, Egham, Surrey, TW20 0EX, England Phone: 0784 39025 Fax: 0784 37520 Telex: 935504 RHC Email: Janet: A.JOHNSTONE@uk.ac.rhbnc.vaxa Arpa: A.JOHNSTONE%vaxa.rhbnc.ac.uk@cs.ucl.ac.uk Bitnet/NetNorth/Earn: A.JOHNSTONE@vaxa.rhbnc.ac.uk ------------------------------------------------------------------------------ ================================================================================ Note 90.7 TK50, TK70 & THE VAXstation 3200 7 of 9 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 12 lines 21-SEP-1988 16:05 -< Look good to FS ! >- -------------------------------------------------------------------------------- After reading all this about TK50's and 3200's not being "buddies" I've sweating bullet as this is the combo I had order several months ago. The word from FS is that "sure it works as long as you have the latest the latest rev" (J or K I think). Well, the installer just left and was able to boot and run the full diagnostics on the system so things look good ! I'll be loading up VMS in the next day or so and I'll let you know what VMS thinks about TK50's on a 3200 ! ================================================================================ Note 90.8 TK50, TK70 & THE VAXstation 3200 8 of 9 EISNER::CHILD "John Child" 16 lines 28-SEP-1989 16:30 -< They might work now >- -------------------------------------------------------------------------------- Just to end the dicussion once and for all, it does NOT work (at least with whatever rev I have) - it's just fine for small numbers of files, but if you try to, say, back up an entire RD54, it goes along just fine for a while, and then you get a "fatal controller error" (on a standalone backup, the drive actually takes itself offline). The word I heard from Colorado 6 months ago was that there's a "signal problem" between the TK50 and the 3200, because the 3200 is "too fast" for the controller. FS refused to put it under contract at that time - but just the other day they said that they would - there must be a new controller rev out, or something - I'll let you know. In the meantime, has anyone got any ideas about ways of "slowing down" the system so I can do a full backup? One of the system manglers here thought I might be able to tweak some caching parameters... ================================================================================ Note 90.9 TK50, TK70 & THE VAXstation 3200 9 of 9 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 3 lines 29-SEP-1989 11:23 -------------------------------------------------------------------------------- Gee, we've had a TK50 on our VAXstation 3200 for almost a year now with now problems. We haven't done too many SA BACKUPS but they all have worked. Sorry don't know rev numbers on the hardware.  ================================================================================ Note 91.0 What Hardware Documents Do You use? 3 replies EISNER::KOZAM 23 lines 9-JAN-1988 03:39 -------------------------------------------------------------------------------- Since DEC no longer ships much documentation with their products, I'm currently looking for hardware docs for the types of things found on MicroPDP and Microvax systems. Things like RD53, RD54, RX50, TK50, BA23, BA123, DHV11, DEQNA, RQDX2, RQDX3, TQK50. DEC offers many documents - installation manuals, user guides, programming guides, technical manuals, maintenance manuals, pocket service guides, handbooks, reference manuals, illustrated parts breakdowns, and print sets. I want to get a fairly complete set of hardware docs, but can't spend a fortune. In the past, about 99% of all the information I ever needed to install and program a device was contained in the appropriate handbook, so I didn't generally bother buying all the device-specific manuals. I purchased print sets as they were needed for board repairs. The latest handbooks I have no longer contain the programming and connector configuration information I sometimes need. I primarily need installation and technical manuals, but want to be as complete as possible, so I'm considering ordering the Maintenance Documentation Service Library (MDS). How complete is it? Does the QBUS set include uVax? At $ 700, it could be quite economical. What do YOU use most? (If you've got part numbers, that would help greatly.) ================================================================================ Note 91.1 What Hardware Documents Do You use? 1 of 3 EISNER::KENNEDY "Terry Kennedy" 66 lines 9-JAN-1988 05:28 -< Here goes... >- -------------------------------------------------------------------------------- > Since DEC no longer ships much documentation with their > products, I'm currently looking for hardware docs for the types of > things found on MicroPDP and Microvax systems. Things like RD53, > RD54, RX50, TK50, BA23, BA123, DHV11, DEQNA, RQDX2, RQDX3, TQK50. Well, since someone asked, I'll open my can of worms... General information: DEC should ship a user's guide (when one exists, part number EK-xxxxx-UG), or when one does not, a technical manual (part number EK-xxxxx-TM) for each of these items. If they did not, you've been stiffed, and should call and ask. Given that, these doc's can be pretty useless for anything beyond simple installation. RD53: The user's guide says essentially 'this is a disk drive - don't drop it - have fun'. The actual disk drive is a Micropolis model 1325 or 1335 (depending on it's age) THey are in Chartsworth, CA and will sell you a full service book quite inexpensively. Also, their swap-out rate on dead drives is about 1/3 that of DEC's. RD54: About the same as the RD53, except that it's a Maxtor brand drive. RX50: Don't bother - if it breaks get a RX33 clone. The DEC drive is a TEAC FD-55GFV-17-U. This is the same drive used in PC/AT's. You can get one for about $110. A company called JDR Microdevices in San Jose, CA has the complete service manual for about $30. (yes, they also sell the drives). TK50: A) You don't want to program it. B) if it breaks and it's a non-obvious problem, DECmailer it. It will cost $600, but you'll get a new mechanism with all the bugs worked out. BA23: Don't buy the Field Maintenance Print set - It's $80., but it's only four pages long! BA123: Don't know - don't have one DHV11: This one is a winner. Both the -TM and the MP- parts are full of good information. Perhaps this is because the board is discontinued? DEQNA: THe only book I've ever seen for it is the -UG. I think that pro- gramming information is available, but I have never seen the -TM. RQDXN: The -UG is basicly an installation guide. Programming information is available by purchasing license QP-905. Last I looked, that was a $xxx, xxx.xx price tag. Not for me... RQDXE: (RQDXn expander) The -UG is a perfect example of how NOT to write a manual. There are references to non-existant figures, the tables are WRONG, etc. Just plug it in and hope (works as shipped in most normal cases). In a well-designed board, where no component is overstressed, the probable failure points are the chips with the highest density. In most of the boards above, those are proprietary DEC chips which they will not sell to you. Further, many of these boards are driven by local software, which makes them very hard to troubleshoot. Therefore, the MP- printsets are not as valuable as they once were. > I primarily need installation and technical manuals, but want > to be as complete as possible, so I'm considering ordering the Maintenance > Documentation Service Library (MDS). How complete is it? Does the > QBUS set include uVax? At $ 700, it could be quite economical. I don't know, but I would be interested in finding out whatever you discover. Terry ================================================================================ Note 91.2 What Hardware Documents Do You use? 2 of 3 EISNER::KILLEEN "Jeff Killeen" 7 lines 9-JAN-1988 08:03 -< SEE 6.30 >- -------------------------------------------------------------------------------- Check out note 6.30 in this conference. The Z????? doc set has changed. Order the manuals not the doc set. You need the 11/53 manual even if you don't have a 11/53 since it covers all hardware released around the time the 11/53 was introduced. ================================================================================ Note 91.3 What Hardware Documents Do You use? 3 of 3 EISNER::KOZAM 36 lines 17-JAN-1988 00:51 -< All about Maintenance Documentation Service >- -------------------------------------------------------------------------------- I called the Self-Maintenance Program Information Line (603-884-5000) for information about the Maintenance Documentation Service Library microfiche. On the plus side: 1. It is complete - including user guides, technical manuals, programming manuals, installation guides, and FCO information - everything except print sets. 2. You can buy just what you need. The library is divided into Diagnostic Listings and Hardware Documentation. Within each of these two categories are Hardcopy Terminals, Video Terminals, PDP-11, Q-BUS, UNIBUS, Low-End VAX, VAX, and PDP-8. 3. An optional yearly subscription keeps your library up to date. 4. Product compatability matrix (I don't think it is available elsewhere). On the minus side: 1. Unless you really need many of these manuals, it is cheaper to buy hardcopy. 2. You can't make notes or corrections(!) in the "margins" of microfiche. 3. You need to carry a portable microfiche reader. 4. If you need FCO information (like firmware changes), DEC-O-LOG is about half the price and a year of updates in included. Prices are in the Self-Maintenance Services Catalog under the Spares Listing (Class MD parts). It is on page A-110 of the June, 1987 to December, 1987 catalog. Before ordering, be sure of what you will be getting. For instance, the Q-BUS set is only for Q-BUS microvax products, not Q-BUS PDP. I've decided to buy hardcopy. Marc Kozam  ================================================================================ Note 92.0 What's in that box (RD32)? 4 replies EISNER::NAGY "Frank J. Nagy, VAX Guru" 7 lines 14-JAN-1988 17:52 -------------------------------------------------------------------------------- Does any who has an RD32 taken things apart to see what is really in that box? What type of disk drive is it? Why? Well, we have some 1/2 height ST506 interface drives with 40-some-odd megabytes of storage left from a defunct project and are interested in trying these in a new incarnation as pseudo-RD32s in our VAXStation-2000s. Anyone know what type of drives (whose?) Digital is selling as the RD32? ================================================================================ Note 92.1 What's in that box (RD32)? 1 of 4 EISNER::KENNEDY "Terry Kennedy" 7 lines 15-JAN-1988 01:33 -< ST-251 >- -------------------------------------------------------------------------------- Seagate ST-251. The PROM may be different, though. Try (with RQDX3) the formatter with 'use down-line-load' = yes. You should get a table of allowable drive models. Enter the code number for a RD32. Failing in this, contact me offline (MAIL) for the drive modifications necessary for many DEC-alike drives. P.S. - XCOM - would posting this information here be deemed appropriate? ================================================================================ Note 92.2 What's in that box (RD32)? 2 of 4 EISNER::KILLEEN "Jeff Killeen" 4 lines 15-JAN-1988 01:37 -< GO FOR IT! >- -------------------------------------------------------------------------------- > P.S. - XCOM - would posting this information here be deemed appropriate? It is OK as long as you aren't selling drive mod kits. ================================================================================ Note 92.3 What's in that box (RD32)? 3 of 4 EISNER::NAGY "Frank J. Nagy, VAX Guru" 10 lines 4-FEB-1988 07:50 -< Seagate ST251 as RD32 via select jumper >- -------------------------------------------------------------------------------- Thanks to advice from Terry Kennedy, we now have an RD32 lookalike (Seagate ST251) in our VAXStation-2000. Once we moved the select jumper from position 1 to position 3 (as seen from the back of the drive with the circuit board on the bottom), the VS-2000 recognized the drive and happily formatted and verified it as an RD32. We have moved the page and swap files from the RA81 (served by a MicroVAX-II) to the local RD32 and realized a very significant and very noticeable improvement in the performance of the VS-2000. Now to put all those ST251s in storage to work (they came out of PCs and some defunct VME systems) thanks to DECUServe. ================================================================================ Note 92.4 What's in that box (RD32)? 4 of 4 EISNER::NAGY "Frank J. Nagy, VAX Guru" 6 lines 18-JUN-1988 12:06 -< Seagate ST251-1s work as RD32s >- -------------------------------------------------------------------------------- We just received a shipment of Seagate ST-251-1 drives (same as ST251 except that the -1 has a 28 msec average access time compared to the 40 msec time of the original ST251). We have successfully installed one drive into a VAXStation-2000 (replacing the ST251 which was in there) with NO problems. We hve not done any timing tests to verify this but the system does seem peppier now.  ================================================================================ Note 93.0 Low-end disk devices - The Digital difference 35 replies EISNER::KENNEDY "Terry Kennedy" 22 lines 16-JAN-1988 01:14 -------------------------------------------------------------------------------- The disk drives used in Digital's low-end systems are generally slightly- or un-modified versions of devices available on the open market. DEC's markup appears to average about 800% of the best price you can get on the open market. For some, that is not an unreasonable amount to pay for the security of an 'all-DEC' system. For others (such as myself) it is entirely unnecessary. There are several companies who will do you the 'favor' of sell- ing you a DEC 'equal' drive for *only* a 400% mark-up. Since the PC 'explosion', prices on these equivalent units are very low due to the cut-throat pricing of the marketplace. I therefore set out to determine what the differences (if any) between the DEC device and the generic device are. This series of notes will present my discoveries. Please feel free to use this information in any way you choose, but remember that neither I, Digital, nor the vendor you buy the equipment from can be responsible for problems you may en- counter. Having said that, let me also say that I will do my best to help you overcome any problems, and that all of the vendors I have used offered to give me a refund if the unit did not work in my DEC system. ================================================================================ Note 93.1 Low-end disk devices - The Digital difference 1 of 35 EISNER::KENNEDY "Terry Kennedy" 20 lines 16-JAN-1988 01:28 -< RD52 >- -------------------------------------------------------------------------------- The DEC RD52 comes in two flavors - the more common one is the Quantum Q540 drive. The second-source drive, an Atasi AT3046, was added due to shortages of the Q540 when DEC's demand was highest. These drives will work on RQDX1, 2, or 3 controllers at the current revision level. There are no differences between the DEC and non-DEC flavors which prevent operation. DEC is alleged to require tighter tolerances on drives made for them. There is actually a third RD52 drive type. None of these units were ever shipped to DEC customers, and the vendor ceased pro- duction of disk drives. The manufacturer and model was the Evotek ET5540. If you have one of these drives around, you can use it with your DEC subsystem. Both drives are no longer manufactured in any quantity, so I would advise against buying one. If you must, however, they should be available in the $300-$400 range from various surplus shops. I have used both models without problems. Since this is the largest drive a RQDX1 will support, you might want to investigate this option. ================================================================================ Note 93.2 Low-end disk devices - The Digital difference 2 of 35 EISNER::KENNEDY "Terry Kennedy" 25 lines 16-JAN-1988 01:37 -< RD53 >- -------------------------------------------------------------------------------- The DEC RD53 is a Micropolis 1325 or 1335 drive with one mod- ification. Earlier units were 1325s, current units are 1335s. The modification is as follows [yes, it will probably void your warranty]: Turn the drive over so that the logic board is facing you. Locate the two slotted screws and loosen them. Lift the drive logic board up gently (it is hinged). Toward the hinge end, locate the empty location marked R7. Solder a jumper in. You now have a RD53. You will need to move the drive select jumper to either 3 (for DU0 or expansion cabinet) or 4 (for DU1). One word of caution - if you are re-using your old drive's skid plate (the slide mount bracket), be sure to tape over the metal on the front end of the bracket or it will short out the drive. DEC's RD53 skid plates have foam covering the metal. I have purchased three of these drives from Storage Dimen- sions in California, (408) 395-2688. The cost was under $850 each and they arrived in two days. They offer a one year warranty on the drive, too. They call it a 'Model AT70', but they're probably going to know what you mean if you ask for 'the DEC drive' by now! Of course, you need a RQDX2 or 3 to run this drive... ================================================================================ Note 93.3 Low-end disk devices - The Digital difference 3 of 35 EISNER::KENNEDY "Terry Kennedy" 72 lines 16-JAN-1988 02:05 -< RX33 >- -------------------------------------------------------------------------------- The DEC RX33 is an IBM AT-type drive with some jumper changes. DEC's drive is some variant of the TEAC FD55G. I have exper- imented and discovered that the following drives/jumperings work with the RQDX3: TEAC FD-55GFV-17-U Jumpers in: HG I (Roman numeral 1) U1 U2 DC FG DS0 (for first RX33) DS1 (for second RX33) Jumpers out: (all others) Install the terminator, RA1, on the last (or only) drive TEAC FD-55GFR-540-U Jumpers in: I (Roman numeral 1) U0 U1 DC2 FG D0 (for first RX33) D1 (for second RX33) Jumpers out: (all others) Install the terminator, RA1, on the last (or only) drive Toshiba FDD 6882E1J01 (or ND-08DE-G) Jumpers in: LD DC DE D1 (for first RX33) D2 (for second RX33) Jumpers out: (all others) Install the terminator jumper, TM, on the last (or only) drive I have purchased many of the TEAC 540 drives from JDR Micro- devices in California, (800) 538-5000. They have the follow- ing items of interest: FD-55G High density drive - $129.95 FD-5Y 'Y' cable to power 2 drives from the BAnn box - $2.95 FD55-MHW Mounting hardware for two FD-55G drives - $2.95 FD55-FP Beige faceplate for FD-55G drive (black is std.) - $2.95 FD55-SPEC Specification for FD-55G drive - $5.00 FD55-MAINT Maintenance manual for FD-55G drives - $25.00 Therfore, you can get two RX33's for $265.80 with the Y cable and mounting hardware - a far cry from DEC's $795.00 each... You need an RQDX3 to run these drives, of course. You will also need to save the skid plate from your RX50 drive. If you order these drives elsewhere, please remember that the mounting hole threads are *metric*. Finding the right screws will be a lot harder than finding the drives! ================================================================================ Note 93.4 Low-end disk devices - The Digital difference 4 of 35 EISNER::KOZAM 28 lines 17-JAN-1988 00:17 -< Used DEC = Big Savings, No Compatability Problems >- -------------------------------------------------------------------------------- If you need all-DEC, but are comfortable with used equipment, there are some bargains available. For example, used DEC RD53s sell for $ 700 each, including cables. These are the real thing, DEC labels and all. New DEC RD53s, in the box, can be had for between $ 1000 - $ 1200. The EXACT same item through DECDirect lists for $ 3800. Regardless of whether you got them used or new, field service will take care of them (although at $ 700, it is probably cheaper to replace, rather than repair them). Most used dealers are very reputable. A 30 to 90 day warranty is common and most places let you talk to their technical people if you need help. Some places don't do any testing - they just buy, then ship to you. Others are very thorough and clean up things up so that it looks like new. One caution, however, is to make sure you get the small parts you need. For instance, make sure you get cables. Often, all it takes is a reminder to the dealer that you need them. A great source of used prices is a weekly publication called The Processor, (800-247-4880). I think you can get a free trial subscription by calling. I've been very happy with my used equipment purchases and have had fewer problems with used stuff than with new. Marc Kozam Disclaimer - I'm not a dealer, just a happy player in the used market. ================================================================================ Note 93.5 Low-end disk devices - The Digital difference 5 of 35 EISNER::KENNEDY "Terry Kennedy" 0 lines 4-FEB-1988 19:15 -< RD32 - see Note 92.3 in this conference >- -------------------------------------------------------------------------------- ================================================================================ Note 93.6 Low-end disk devices - The Digital difference 6 of 35 EISNER::KENNEDY "Terry Kennedy" 12 lines 1-APR-1988 01:22 -< RD54 >- -------------------------------------------------------------------------------- The DEC RD54 is a Maxtor XT-2190. I have installed one of these with no modifications in a Micro-11 system (with a RQDX3 as the controller, of course). I purchased the drive from Storage Dimensions in Califor- nia, (408) 395-2688. The cost was under $1700 and it arrived in two days. They offer a one year warranty on the drive, too. They call it a 'Model AT133'. Of course, you need an RQDX3 or the integrated thingie in a MicroVAX 2000 to run this drive... ================================================================================ Note 93.7 Low-end disk devices - The Digital difference 7 of 35 EISNER::ETHINGTON "Superior Klingon Technology" 9 lines 14-MAY-1988 23:41 -< Wanted: Skid plate part # >- -------------------------------------------------------------------------------- Along these lines, folks.....does anyone know the part numbers for your friendly neighborhood field circus to sell you the plastic skid plate and big flat grounding spring/shield or whatever? Lord knows you can get the same disk for enormously less loot than Storage Systems wants, but it would be nice to have the official skid plate hardware to hold it in place instead of rubber bands or string or whatever..... Jerry ================================================================================ Note 93.8 Low-end disk devices - The Digital difference 8 of 35 EISNER::KILLEEN "Jeff Killeen" 0 lines 15-MAY-1988 00:35 -< DEC WILL NOT SELL YOU THE SKID PLATE >- -------------------------------------------------------------------------------- ================================================================================ Note 93.9 Low-end disk devices - The Digital difference 9 of 35 EISNER::RICE "Gary Rice" 12 lines 15-MAY-1988 09:12 -------------------------------------------------------------------------------- >< Note 93.8 by EISNER::KILLEEN "Jeff Killeen" > > -< DEC WILL NOT SELL YOU THE SKID PLATE >- Is this a new policy? I was able to buy 10 of the little puppies about a year and a half ago thru DECdirect. I don't remember the part number. It is packed away in my annual receipts bag for 1986. However, if you need a small quantity, let me know via mail. I could be convinced to part with 4 or 5 of them. Gary ================================================================================ Note 93.10 Low-end disk devices - The Digital difference 10 of 35 EISNER::KOZAM 8 lines 15-MAY-1988 18:21 -< Skid Plate Part Number 74-25652-01 >- -------------------------------------------------------------------------------- The part number for skid plates (which I read off an RD53) is 74-25652-01 and is described as "112684 Skid" and is priced at only $ 5.00 each (that's a real surprise!). I don't know if that includes the metal grounding plate or just the plastic part. Marc Kozam ================================================================================ Note 93.11 Low-end disk devices - The Digital difference 11 of 35 EISNER::KILLEEN "Jeff Killeen" 3 lines 16-MAY-1988 18:26 -< HMMMMMM....... >- -------------------------------------------------------------------------------- When I tried to buy one I was told by the E-store they were not available for sale. If you get them let us know - I will go out and buy a dozen. ================================================================================ Note 93.12 Low-end disk devices - The Digital difference 12 of 35 EISNER::RICE "Gary Rice" 15 lines 18-MAY-1988 06:59 -------------------------------------------------------------------------------- > The part number for skid plates (which I read off an RD53) is > 74-25652-01 and is described as "112684 Skid" and is priced at only $ 5.00 > each (that's a real surprise!). > I don't know if that includes the metal grounding plate or just > the plastic part. 74-25652-01 buys you the plastic "plate". Part number 74-28201-02 gets you the metal "grounding strap". It is a bit more at $11.00 (1986 price). I got BOTH from DECdirect by just calling and requesting these part numbers. They were very helpful and gave no indication that I wasn't allowed to buy these parts. Gary ================================================================================ Note 93.13 Low-end disk devices - The Digital difference 13 of 35 EISNER::ETHINGTON "Superior Klingon Technology" 9 lines 29-MAY-1988 00:32 -< Paydirt! >- -------------------------------------------------------------------------------- Thanks, folks! I had the part number for the plastic skid plate, and had no difficulty at all ordering a batch - I think the price was still around $5. I had never come up with the number for the big spring grounding plate doodad, though - I'll try ordering some of these critters now. DECUS to the rescue again, God help me I do love it so..... Jerry ================================================================================ Note 93.14 Low-end disk devices - The Digital difference 14 of 35 EISNER::KENNEDY "Terry Kennedy" 7 lines 29-MAY-1988 05:06 -< Logic of part numbers! >- -------------------------------------------------------------------------------- > ... still around $5. I had never come up with the number for the > big spring grounding plate doodad, though - I'll try ordering some > of these critters now. The spring p/n should be a -02, where the plastic is -01. Terry ================================================================================ Note 93.15 Low-end disk devices - The Digital difference 15 of 35 EISNER::ETHINGTON "Superior Klingon Technology" 16 lines 21-JUN-1988 00:52 -< Stacking Hardware????? >- -------------------------------------------------------------------------------- Still further along these lines..... I understand that new mounting hardware is available to support stacking two half-height RD3x series Winchesters in a BA23 box. It is also apparently supported to stack a pair of RX33 half-height floppies, allowing you to stuff 2 Winchesters and 2 floppies in the BA23 box. Assuming one has say an official DEC RD31 with skid plate and such, does anyone know what parts and cables need to be ordered to stack these varmints? I've not seen the stacking setup; has anyone got one of these critters to see the cabling? For instance, do you need a "Y" power cable to plug 2 Winchesters into one plug in the BA23 box? How do the 2 floppies cable up? Inquiring minds want to know..... Jerry ================================================================================ Note 93.16 Low-end disk devices - The Digital difference 16 of 35 EISNER::KENNEDY "Terry Kennedy" 31 lines 21-JUN-1988 21:26 -< Possible, proceed carefully >- -------------------------------------------------------------------------------- > I understand that new mounting hardware is available to support > stacking two half-height RD3x series Winchesters in a BA23 box. My sales rep also 'understands' this - however, the only way we've figured out to get the part numbers is to order an 11/53 and dissect it. This is because the '53 is one of the only systems standard with this 2-up stuff, and the 11/53 documentation is a few photocopied addendum sheets which you may or may not get. You *will* need the BA23-UC (@$200.00) upgrade kit to properly put 2 winnies in the BA23. This kit gives you an extra set of ready/write protect buttons, and deletes the two floppy WP indicators. [Yes, you can do it without the -UC, but it isn't very pleasant and if you get the drive selects wrong you will blow the format off *both* drives.] > It is also apparently supported to stack a pair of RX33 half-height > floppies, allowing you to stuff 2 Winchesters and 2 floppies in > the BA23 box. Yes, that is the maximum configuration. If you go for it, be *sure* you have the current revision of the power harness or you'll go up in smoke!. The current harness is individual wires with white plastic con- nectors (between the power supply and the backplane). Anyway, to answer your question, I just went out and got the 'Y' cable and the adapter plates from my local garage-based PC outfit. Most places which will sell you RX33 look-alikes will gladly throw in the Y and the plates with the order. I have ribbon-cable presses here, so I made my own data cables up. Again, your local PC place should be able to make them up, once you convince them it's not a standard PC part and that you know what you're doing... ================================================================================ Note 93.17 Low-end disk devices - The Digital difference 17 of 35 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 13 lines 21-JUN-1988 22:35 -< AFTER MANY HRS OF WONDERING WHY I COULDN'T WRITE TO DU1 >- -------------------------------------------------------------------------------- > [Yes, you > can do it without the -UC, but it isn't very pleasant and if you get > the drive selects wrong you will blow the format off *both* drives.] Yes and No - Without the second set of buttons the second internal drive comes up R/W on an RQDX1 and RO on a RXDX2 or RQDX3. Trimarchi (no I don't own stock in this company) sells the drives with the stacking hardware and a special set of cables. The cables de-activate the two front panel disk buttons and forces the drives ready W/R all the time. I found this one out the hard way...... ================================================================================ Note 93.18 Low-end disk devices - The Digital difference 18 of 35 EISNER::MAYHEW "Bill Mayhew" 15 lines 22-JUN-1988 10:36 -< Drives w/stack hardware, and floating signals >- -------------------------------------------------------------------------------- There is a variant of both the RD31 and RD32 drives that "comes with" the stacking hardware, but that doesn't help a whole bunch if you already have the drive... for those that care, though, you should be able to find it readily (the words STK HDW or such are in the part title) on the Estore or in the USPL. I have reports from other folks who have put multiple Winchesters in a BA23 without the extra switches that what happens is the write-protect and drive-enable signal lines are left "floating". Depending on the phase of the moon, your power supply (AC and DC), etc., you can then have flaky problems where the drive suddenly drops off-line, or write-protects itself, without warning. I don't know which controller was involved, though I think it was the RQDX1. The drives were RD52s, which are risky ANYWAY to double up in a BA23 because of power draw, but these folks did it... ================================================================================ Note 93.19 Low-end disk devices - The Digital difference 19 of 35 EISNER::KENNEDY "Terry Kennedy" 7 lines 22-JUN-1988 20:50 -< Many variables... >- -------------------------------------------------------------------------------- > Yes and No - Without the second set of buttons the second internal drive > comes up R/W on an RQDX1 and RO on a RXDX2 or RQDX3. It depends on a number of things, including the revs of the RQDXn, whether a RQDXE is also in the system, and the rev of the front panel. Also, you need to jumper the first fixed disk as DS3 and the second as DS4, even if you have the new front panel. ================================================================================ Note 93.20 Low-end disk devices - The Digital difference 20 of 35 EISNER::KENNEDY "Terry Kennedy" 26 lines 27-JUN-1988 23:43 -< RX50 drive substitute! >- -------------------------------------------------------------------------------- Finally, a RX50 drive replacement! I have managed to convince a Toshiba IBM AT-type floppy drive that it is an RX50. I've tested extensively with both XXDP+ and RSTS/E, but that doesn't mean that it will work with other systems - you'll have to try it and let me know. Here's what you need to do: 1) Get a Toshiba ND-08DE-G drive kit from your local IBM garage oper- ation. (The actual drive is a Toshiba FDD6882E1J01) 2) Set jumpers as follows: D1, TM, LD and DE on, all others off. 3) Cut the trace leading to pin 2 on the data connector. 4) Solder a wire from the drive side of the wire you cut in (3) to signal ground (the large solid areas on the circuit board) This has been tested on both the RQDX1 (M8639/M8639-YA) controller with firmware 172E5/173E5, as well as with the RQDX2 (M8639-YB) with firmware 188E5/189E5. With an RQDX3, you can use the same drive in RX33 mode (see an earlier reply in this note). For the curious: what the above has done is forced the drive into 300 RPM/low density mode, and disabled the drive ready line. ================================================================================ Note 93.21 Low-end disk devices - The Digital difference 21 of 35 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 10 lines 10-AUG-1989 19:57 -< Does RD53 change for uV2000? >- -------------------------------------------------------------------------------- >< Note 93.2 by EISNER::KENNEDY "Terry Kennedy" > > -< RD53 >- > >The DEC RD53 is a Micropolis 1325 or 1335 drive with one mod- >ification... Is there anything different about the RD53 as used in the uV2000? (We have a 2000 with a failing RD53...) Also, can we drop an RD54 lookalike into instead? ================================================================================ Note 93.22 Low-end disk devices - The Digital difference 22 of 35 EISNER::KOZAM 8 lines 10-AUG-1989 22:29 -< VS2000 RD53 is a regular RD53 >- -------------------------------------------------------------------------------- > Is there anything different about the RD53 as used in the uV2000? > (We have a 2000 with a failing RD53...) Nothing different from a regular RD53. > Also, can we drop an RD54 lookalike into instead? Don't know why not. Never heard otherwise. ================================================================================ Note 93.23 Low-end disk devices - The Digital difference 23 of 35 EISNER::KENNEDY "Terry Kennedy" 0 lines 10-AUG-1989 23:54 -< .-1 is correct on both counts >- -------------------------------------------------------------------------------- ================================================================================ Note 93.24 Low-end disk devices - The Digital difference 24 of 35 EISNER::HARVEY "Jack Harvey" 12 lines 11-AUG-1989 04:44 -< Skid Plate >- -------------------------------------------------------------------------------- Trivial mechanical. The 2000 has mounting brackets that attach to holes in the sides of the raw 5.25" device. You need to screw the black plastic (with floppy metal ground sheet) skid plate to the bottom of the raw device when mounting it in a BA23, 123, etc. I have formatted lookalike RD54's on a 2000, loaded software and then moved the disk to a II. No changes, other than screwing on a skid plate. If you are going the other direction, just remove the skid plate and save it for a friend. ================================================================================ Note 93.25 Low-end disk devices - The Digital difference 25 of 35 EISNER::OSUDAR "John Osudar" 6 lines 11-AUG-1989 13:32 -< Right -- no problems >- -------------------------------------------------------------------------------- Yup, the last several notes are right. We got two VS2000's diskless, bought an RD53-alike and an RD54-alike and put one in each, with no problem (other than the mounting change.) Subsequently got another RD54-alike, ripped out the 53 and put the 54 in with no problems. Beats paying DEC's prices... :-) ================================================================================ Note 93.26 Low-end disk devices - The Digital difference 26 of 35 EISNER::MCCORMICK_B "Virtual Bob" 5 lines 12-AUG-1989 16:35 -< Disk format quirk? >- -------------------------------------------------------------------------------- Don't remember exactly (maybe someone else will remember...?!) but I believe there was a problem with some (maybe early) 2000's where if you formatted the disk on the 2000 it didn't result in the same amount of "formatted capacity" as when you formatted it on the RQDX3 ... ================================================================================ Note 93.27 Low-end disk devices - The Digital difference 27 of 35 EISNER::KOZAM 11 lines 14-AUG-1989 21:54 -< Never seen formatting trouble >- -------------------------------------------------------------------------------- > < Note 93.26 by EISNER::MCCORMICK_B "Virtual Bob" > > -< Disk format quirk? >- > where if you formatted the disk on the 2000 it didn't result in > the same amount of "formatted capacity" as when you formatted it > on the RQDX3 ... While I can't conclusively state that this didn't ever occur, I've never seen it. Moving RD5X drives between 2000s and RQDX3s was pretty common for me for a while and I never had trouble, even though several versions of RQDX3 firmware were involved. ================================================================================ Note 93.28 Low-end disk devices - The Digital difference 28 of 35 EISNER::MCCORMICK_B "Virtual Bob" 14 lines 20-AUG-1989 18:20 -< Maybe not problem - but different! >- -------------------------------------------------------------------------------- | > < Note 93.26 by EISNER::MCCORMICK_B "Virtual Bob" > | > -< Disk format quirk? >- | | > where if you formatted the disk on the 2000 it didn't result in | > the same amount of "formatted capacity" as when you formatted it | > on the RQDX3 ... | | While I can't conclusively state that this didn't ever occur, I've | never seen it. Moving RD5X drives between 2000s and RQDX3s was pretty | common for me for a while and I never had trouble, even though several | versions of RQDX3 firmware were involved. I think you won't see it as a problem (or trouble) but if you did a SHOW DEV/FULL DUAn: you may see different "formatted" sizes! ================================================================================ Note 93.29 Low-end disk devices - The Digital difference 29 of 35 EISNER::KENNEDY "Terry Kennedy" 8 lines 15-SEP-1989 22:41 -< Toshiba-as-RX50 doesn't work on RQDX3 >- -------------------------------------------------------------------------------- > I have managed to convince a Toshiba IBM AT-type floppy drive that it > is an RX50. I've tested extensively with both XXDP+ and RSTS/E, but that > doesn't mean that it will work with other systems - you'll have to try > it and let me know. By the way, this *won't* work on an RQDX3 - it only works on the RQDX1 and RQDX2. Of course, you can use the Toshiba drive on the RQDX3 by setting it up as an RX33 lookalike (see earlier note for jumper settings). ================================================================================ Note 93.30 Low-end disk devices - The Digital difference 30 of 35 EISNER::MCCORMICK_B "Bob McCormick" 2 lines 19-SEP-1989 17:46 -< DEC stuff on PC's? >- -------------------------------------------------------------------------------- How about the (DEC) Teac FD-55GFV-57-U used on an XT or AT style machine? Any ideas how to set up its jumpers, etc? ================================================================================ Note 93.31 Low-end disk devices - The Digital difference 31 of 35 EISNER::KENNEDY "Terry Kennedy" 6 lines 20-SEP-1989 03:23 -< Eh? >- -------------------------------------------------------------------------------- > How about the (DEC) Teac FD-55GFV-57-U used on an XT or AT style > machine? Any ideas how to set up its jumpers, etc? I beg your pardon? You want to take a real DEC RX33 and use it on an XT or AT? I'm not sure what you're trying to accomplish... If you could clarify the "I have" and the "I want" parts, I'll give it a try... ================================================================================ Note 93.32 Low-end disk devices - The Digital difference 32 of 35 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 12 lines 3-OCT-1989 19:38 -< Control panel for drives in BA123? >- -------------------------------------------------------------------------------- So far our adventures in drive replacement have gone well. Now I want to ADD an RD53 lookalike to a uV2 (BA123). I have the drive, and am ordering the skid plate and ground strap. Cables we can make. What about the control panel? Is there a DEC part number for this assembly? (Presumably we need the panel and a cable to plug into the "distribution board") Can I use one of those PC-style Molex Y cables for power, or is there a DEC part I should order? ================================================================================ Note 93.33 Low-end disk devices - The Digital difference 33 of 35 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 1 line 3-OCT-1989 19:38 -< p.s. -- Thanks, folks! >- -------------------------------------------------------------------------------- > So far our adventures in drive replacement have gone well. ================================================================================ Note 93.34 Low-end disk devices - The Digital difference 34 of 35 EISNER::KOZAM 11 lines 3-OCT-1989 20:55 -< No need for front panel >- -------------------------------------------------------------------------------- > What about the control panel? If the drive is going to be on-line and write-enabled at all times, then you have no need for a panel. I've never had any problem with this arrangement. > Can I use one of those PC-style Molex Y cables for power, or is there > a DEC part I should order? You should have a power cable for each device slot in the BA123, so a Y shouldn't be necessary. ================================================================================ Note 93.35 Low-end disk devices - The Digital difference 35 of 35 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 18 lines 5-OCT-1989 05:41 -< BA123 RD CONTROL PANELS >- -------------------------------------------------------------------------------- > What about the control panel? Is there a DEC part number for this > assembly? (Presumably we need the panel and a cable to plug into > the "distribution board") The RD control panel for a BA123 box is part # 54-16244-02 If there is a DEC RD drive in the BA123 you should already have the cable since it is a single cable for all four possible drives. The cable's part # is 17-00862-01 Trimarchi sells a clone for $250 and it includes the cable if you ask for it. > Can I use one of those PC-style Molex Y cables for power, or is there > a DEC part I should order? I don't understand this question. All 5 1/4 inch mounting slots have a power cable available. It is the standard PC type drive power cable.  ================================================================================ Note 94.0 Synch BI-unibus devices No replies EISNER::BOWLER "Bruce Bowler" 5 lines 18-JAN-1988 10:38 -------------------------------------------------------------------------------- I have heard numerous reports that Synch devices like the DUP-11 don't work (and in-fact may hose) on BI-based Unibus's. Can any one out there confirm or deny these rumors? Bruce Bowler  ================================================================================ Note 95.0 Looking for displays 5 replies EISNER::NORTON 14 lines 21-JAN-1988 11:15 -------------------------------------------------------------------------------- I'm looking for a box that acts like a VT220, or at least ANSI, that has an RS232 or 20Ma input and video output, so I can remotely locate a large monitor. While it might be acceptable to consume a VT100 or VT200 just to get the video out signal, I'd rather have a unit without a keyboard and redundant display. For a different project, I need a 'VT220 with a 19" or larger screen'. Could be mono or color. Doesn't need graphics, does need VT220 line-drawing character set. Anyone know of a source for these? \bill ================================================================================ Note 95.1 Looking for displays 1 of 5 EISNER::KILLEEN "Jeff Killeen" 7 lines 21-JAN-1988 11:48 -< HMMMM..... >- -------------------------------------------------------------------------------- > I'm looking for a box that acts like a VT220, or at least ANSI, > that has an RS232 or 20Ma input and video output, so I can remotely > locate a large monitor. I seem to remember someone had a Q-Bus board that could do this. But I don't remember who. ================================================================================ Note 95.2 Looking for displays 2 of 5 EISNER::CETRON 5 lines 21-JAN-1988 14:07 -< TVI 9220 >- -------------------------------------------------------------------------------- I believe the TeleVideo 9220 covers all of the requirements. -ed ================================================================================ Note 95.3 Looking for displays 3 of 5 EISNER::NORTON 4 lines 21-JAN-1988 17:55 -------------------------------------------------------------------------------- >>I seem to remember someone had a Q-Bus board that could do this. But I >>don't remember who. Boards not allowed, since it needs to work with any host system. ================================================================================ Note 95.4 Looking for displays 4 of 5 EISNER::KENNEDY "Terry Kennedy" 15 lines 21-JAN-1988 19:17 -< ALternative solutions >- -------------------------------------------------------------------------------- >>I seem to remember someone had a Q-Bus board that could do this. But I >>don't remember who. > Boards not allowed, since it needs to work with any host system. DEC had a 'VT100-on-a-board' for the Q-Bus. It only got power and ground from the bus, though, so you could use it with an outboard power supply. The most inexpensive thing you could come up with these days is probably a PC (or just the necessary parts) and a terminal pro- gram. If you don't need 220, but 102 with line-drawing will do, consider MS-Kermit 2.30 from Columbia (it's free) Terry ================================================================================ Note 95.5 Looking for displays 5 of 5 EISNER::ERSKINE "Chris Erskine" 4 lines 25-JAN-1988 06:48 -< Very close emulator >- -------------------------------------------------------------------------------- Why not use the system unit from a VT-240/241. This has the jacks for mono or color, supports the VT-220 commands and looks like a DEC terminal (It is). I do not know if you can get it without the monitor but you could try.  ================================================================================ Note 96.0 VT125 doesn't work at 19,200 Baud? 6 replies EISNER::LEDERMAN "Bart Z. Lederman" 26 lines 22-JAN-1988 13:35 -------------------------------------------------------------------------------- I have a VT125 that has started acting "funny". When I am editing a file with EDT and I press the new line key a lot (Keypad zero), I sometimes get kicked back into line mode. It also occasionally happens with the arrow keys or some of the other keypad keys. (With LSE, I get an error message about a bad escape sequence). This only happens on keypad keys in application mode (when they keypad isn't being used a numeric keys). By placing a line monitor on the comm. port, I discovered that the VT125 occasionally drops the low bit on an ESCAPE character, which turns it into a Control-Z (hence causing EDT to go into line mode). This appears to be related to how full the buffer is, and if the VT125 is sending or has just sent an X-off. I think this started happening recently at 9600 Baud, but it really became bad when I switched to a terminal server and 19,200 Baud. The port, line, computer hardware, etc., have been conclusively eliminated. DEC has swapped every board in the VT125 to no effect. They now plan to swap in a new VT125, but they also say that the VT125 is not supposed to work at 19,200 Baud. I don't recall seeing anything which says this, and would like to know if there is anyone out there who is using a VT125 at 19,200 Baud sucessfully (or otherwise). Please note that this is happing with text only, not graphics. Am I being unreasonable to expect this terminal to work properly at 19,200 Baud? ================================================================================ Note 96.1 VT125 doesn't work at 19,200 Baud? 1 of 6 EISNER::OSUDAR "John Osudar" 4 lines 22-JAN-1988 14:32 -------------------------------------------------------------------------------- I've seen the same thing at 9600 baud, and maybe even at 4800 (can't recall, haven't used 4800 in a while); I don't htink it was limited to just a VT125 either -- I seem to recall seeing the same thing from other VT1xx series terminals. ================================================================================ Note 96.2 VT125 doesn't work at 19,200 Baud? 2 of 6 EISNER::HANRAHAN "Jamie Hanrahan" 25 lines 22-JAN-1988 15:17 -< Here's a workaround for EDT >- -------------------------------------------------------------------------------- I've also seen it at 4800. I expect it happens more at 9600, and still more at 19.2K, simply because the VT125 is sending XOFFs more often at higher speeds. A workaround for EDT is to put the following into your EDTINI.EDT file: DEFINE KEY CONTROL Z AS "" and use gold-Z to exit screen mode. I know next to nothing (still!) about TPU, EVE, or LSE, so I don't know if there are equivalents. (I hadn't heard the "drops a bit of the escape character, turning it into a control-Z" explanation -- I just knew that my 125 would sometimes emit a gratuitious control-Z, and came up with this fix. Thanks!) BTW, I wouldn't regard the 125's not being supported at 19.2K as a problem, as 19.2 doesn't make it paint the screen any faster. In fact, I never noticed any difference between 4800 and 9600 -- the 125 is even slower than the VT100. For this reason, among others, I no longer have a 125 on my desk. There are some third-party VT clones that hardly ever send an XOFF, even at 19.2; try editing with one of those and you'll never want to go back... until you get used to a VAXstation... ================================================================================ Note 96.3 VT125 doesn't work at 19,200 Baud? 3 of 6 EISNER::HASSINGER "Bob Hassinger" 9 lines 22-JAN-1988 15:40 -< If it's there *of course* it should work! >- -------------------------------------------------------------------------------- 1) My VAX rarely *outputs* screen paints at a real 19.2 Kb anyway - i.e. the compute time is too long to keep up in a lot of cases - maybe bigger VAXs with light work loads can do it. 2) If the VT125 has a 19.2 Kb setting then *of course* you should expect it to work! Otherwise it is like speedometers on little cars that go up to 130 mph just to impress potential buyers isn't it? Bob H ================================================================================ Note 96.4 VT125 doesn't work at 19,200 Baud? 4 of 6 EISNER::DELARISCH "Arnold S. De Larisch" 22 lines 22-JAN-1988 19:03 -< Just because its says 19.2K ...doesn't mean ... >- -------------------------------------------------------------------------------- >> < Note 96.3 by EISNER::HASSINGER "Bob Hassinger" > >> -< If it's there *of course* it should work! >- >> 2) If the VT125 has a 19.2 Kb setting then *of course* you should >> expect it to work! Otherwise it is like speedometers on little cars >> that go up to 130 mph just to impress potential buyers isn't it? Right you are! The VT100 can JUST barely keep up at 4800 Baud ... the VT125 is even S-L-O-W-E-R. If you want to see how fast your VT1xx series terminal really is ... turn off X-on/X-off recognition. Then, edit a file and have the screen repainted a few times. Watch for dropped characters. In general ... VT1xx 4800 Baud VT125 2400 Baud With out Flow control active. AND BTW ... if you need flow control active, you should probably slow down the link. All you are doing (if a lot of flow control characters are being used) is putting UNNECESSARY load on your Terminal/Computer/MUX/Communications link! -Arnold ================================================================================ Note 96.5 VT125 doesn't work at 19,200 Baud? 5 of 6 EISNER::LEDERMAN "Bart Z. Lederman" 14 lines 22-JAN-1988 19:51 -< ex >- -------------------------------------------------------------------------------- It's interesting to see that others have seen problems: I have never seen any problems at 9600 Baud. I agree that most VT100 family terminals can't keep up at even 9600 baud, especially if there are screen control or formatting instructions: however, in repainting a screen "top down" 9600 is much faster than 2400 and some faster than 4800. I don't expect that 19,200 is significantly different than 9600: but it should work. Yes, I can see what happens even at 9600 baud if there is no x-off. And I don't think 19,200 is there just to impress the yokels. Thanks for the response. ================================================================================ Note 96.6 VT125 doesn't work at 19,200 Baud? 6 of 6 EISNER::ERSKINE "Chris Erskine" 0 lines 25-JAN-1988 06:52 -< And smooth scroll = 1200 baud >- --------------------------------------------------------------------------------  ================================================================================ Note 97.0 RX33 on a PRO? No replies EISNER::RICE "Gary Rice" 12 lines 31-JAN-1988 09:25 -------------------------------------------------------------------------------- This note appears in the PRO_300_SERIES conference (topic 22) as well. For some time, I have been curious as to WHY the PRO is GEN'd for 4 floppy drives. Is is possible to add another RX50 controller and another RX50 box? Or better yet, Add an RX50 controller and plug in an RX33 box? Also, Virtual Microsystems is no longer selling the MS-DOS bridge that came with new ROMs for the RX50. (You know the ones that let the RX50 read IBM-PC format). Do you know if these ROMs are available from ANY other source? Gary  ================================================================================ Note 98.0 WHAT IS THE BEST WAY TO SELL OFF SOME HARDWARE? 5 replies EISNER::KILLEEN "Jeff Killeen" 78 lines 1-FEB-1988 00:33 -------------------------------------------------------------------------------- The following note is being posted by the DECUServe Executive Committee on behalf of Alan Frisbie. ================================================================================ I have a task/problem that I hope some of you can help me with. As most of you know, Dan Eisner (a founder of this system) died a few months ago. I have been asked by his family to take care of selling off Dan's computer equipment. I would like your advice on the "best" way to handle this task. For many years I have been on the buying end, trying to get the lowest price possible. Now I find myself on the opposite side of the fence, and have no experience. The methods I have thought of so far include: o Used equipment vendors (Digital Basics, Midwest Systems, Newman, etc.). Easy to do, but how much will they offer? o Ad in The Processor, Computer Shopper, etc. Lead time for ads is ???. How are results? o Auction. Fast, but are prices any better than from used equipment vendors? o Direct sales to people I know. Possibly a better price, but might be accused of conflict of interest. o Auction or direct sale via DECUServe SWAP_MEET. No track record, plus unknown delay in waiting for conference to be created. Are there any other methods that you people can think of? To give you an idea of the kinds of items I have to deal with, this list covers the primary pieces: THIS IS NOT AN OFFER TO SELL, NOR IS IT A SOLICITATION OF AN OFFER TO BUY. IT IS A REQUEST FOR HELP. o VAX-11/730 system. 5 MB, RA80, TU80, DMF32, LA120. VMS v4.4, Fortran v4.3(?), VMS Doc Set, Hardware manuals. o 1 VT-240 and 2 VT-220 terminals. o 2 BA123 boxes. One empty, one containing a small (home-grown) LSI-11/23 system with ~32MB disk. o MicroVAX to VAXstation conversion kit. Single-plane monochrome. Mouse currently missing, but we are looking for it. o RRD50 CDROM player with Q-Bus controller and DECUS Disk #1. o Unibus options: DMF32 and DEUNA, both with cabinet kits. o Q-Bus options: RX02/RXV21, Dataram 0.5 MB memory. o Octek 2200 Image Analysis processor o Assorted other hardware and Doc. Sets. None of the equipment is under a maintenance agreement, and the condition is unknown. Given this collection of equipment, how would YOU go about selling it to benefit Dan's estate the most? Remember, I am doing this as his friend and will receive no commission or other consideration. It would be desirable to act as quickly as possible to spare the family the pain of long delays. Thank you, Alan ================================================================================ Note 98.1 WHAT IS THE BEST WAY TO SELL OFF SOME HARDWARE? 1 of 5 EISNER::LEDERMAN "Bart Z. Lederman" 21 lines 1-FEB-1988 08:29 -< Some thourhgts >- -------------------------------------------------------------------------------- {Used equipment dealers such as Midwest:} Unless they have a customer on hand screeming for what you have to sell, they will offer rock-bottom prices for what you have. And if they already have too much of what you are selling, they won't buy it from you at all unless you are a big potential customer they want to oblige. {Direct sale}: From observation, not experience, direct sale through user bulletin boards, such as Compu-Serve, seem to go quite well. Response if fast, and since there is no middle man you are likely to get more. On the other hand, most of the pepole looking are also hobbyists with low budgets: but occasionally people are looking on behalf of the companies they work for. {Unrelated response:} Y doubt if you will get much for an 11/730 unless sombody already has several and want's another for spares or something. ================================================================================ Note 98.2 WHAT IS THE BEST WAY TO SELL OFF SOME HARDWARE? 2 of 5 EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" 8 lines 1-FEB-1988 13:14 -< A couple of suggestions... >- -------------------------------------------------------------------------------- If you have access like I do, perhaps a note in some newsgroup on CSNET would do the trick. I sell things through this method all the time (as do hundreds of others, too). Contact me if you need more information. Another thought: how about a donation to a local school in the area ? They could write it off as a charitable contribution. ================================================================================ Note 98.3 WHAT IS THE BEST WAY TO SELL OFF SOME HARDWARE? 3 of 5 EISNER::KENNEDY "Terry Kennedy" 11 lines 1-FEB-1988 20:54 -< Sealed bids via MAIL >- -------------------------------------------------------------------------------- If you post a list (possibly with minimum acceptable offers) on a variety of systems, with bids to be MAILed to you by a particular cut-off date, you can do a 'sealed-bid' sort of thing. This will probably be the way to maximize income from the sale. Regarding selling to a used-equipment dealer, they will generally offer you less than half of their 'sell' price, minus transit costs. Also, saying 'no' to the 'is it on DEC maintenance' will cause the bid to fall through the floor in most cases. An exception to the above is equipment that dealers currently have the hots for, like RA82, DELQA, etc. ================================================================================ Note 98.4 WHAT IS THE BEST WAY TO SELL OFF SOME HARDWARE? 4 of 5 EISNER::CETRON 14 lines 2-FEB-1988 13:10 -< my .02 >- -------------------------------------------------------------------------------- 1. Personal sales are the best. 2. I will be happy to post a message/sale prospectus for you on all the available network connections I have (Bitnet, Usenet, and Internet) 3. Used resellers will not give you fair market value. 4. Note to XCOM: this sounds like the perfect 'trial' for the new swap_meet conference, eh? -ed ================================================================================ Note 98.5 WHAT IS THE BEST WAY TO SELL OFF SOME HARDWARE? 5 of 5 EISNER::FRISBIE "Alan E. Frisbie" 21 lines 13-FEB-1988 22:46 -< Thanks for your suggestions >- -------------------------------------------------------------------------------- After reading your suggestions and checking with a used equipment dealer, I have decided to do try the following: 1. Direct sales offers to DECUS members 2. Direct sales to others if anything is left 3. Offer to used equipment vendors if anything is left 4. Drop asking prices and go to #1 This seems to be the best compromise between a fast sale and getting the most money for Dan's estate. The original asking prices will be between what the used equipment vendors offer and what they charge. Some of the prices were difficult to arrive at because the dealers don't know what to offer ("What is an RRD50?"). If the Exec Committee approves, I can make more information available. I am sure that Dan would have liked DECUServe users to get first crack at this. Alan  ================================================================================ Note 99.0 Neet KA620/630 SBC VAX help... 1 reply EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" 9 lines 5-FEB-1988 12:35 -------------------------------------------------------------------------------- We're thinking about getting a KA620/630 SBC VAX for our vision system (the 11 address space just won't make it in this case). Can anyone tell me what size the on-board EROM is ? Does the user have access to it ? Could we blast our own boot program ? Do you HAVE to use ELN for this system (gag, barf..). Bob ================================================================================ Note 99.1 Neet KA620/630 SBC VAX help... 1 of 1 EISNER::PROVOST 27 lines 25-FEB-1988 14:37 -< ELN vs NOTHING YET >- -------------------------------------------------------------------------------- > < Note 99.0 by EISNER::PERRY "Bob Perry - Skydiver/Sky-Scum" > > -< Neet KA620/630 SBC VAX help... >- > Can > anyone tell me what size the on-board EROM is ? Does the user have > access to it ? Could we blast our own boot program ? Do you HAVE > to use ELN for this system (gag, barf..). You don't have to run ELN, but I know of no other operating system which will run on it. Looks like a good hole for a 3rd party product. "The KA620-BA boot and diagnostic ROM provides 64Kbytes of power-up diagnostics, boot programs for standard devices, and a subset of the VAX console program." That doesn't tell us much. In the description for VAXELN in the REALTIME SYSTEMS AND OPTIONS CATALOG it says: Completed VAXELN SYSTEM applications can be loaded from FILES-11 format disk media, tape, or downloaded through Ethernet from the VMS host to the target VAX. Additionally, any VAXELN application can be blasted into read-only memory (ROM) for installation and execution on the target system. Therefor something is available and reachable, but I don't know what or where. Tom  ================================================================================ Note 100.0 Looking for a good key... 3 replies EISNER::ROBITAILLE "Mike Robitaille" 12 lines 10-FEB-1988 16:05 -------------------------------------------------------------------------------- Over in the security conference we've noticed that the keys on just about all DEC CPU's can be interchanged. Tends to make the concept of a lock on "my VAX" kind of ridiculous. Does anyone know of a source out there for key/lock mechanisms that can be easily used to replace the apparently useless "stock" key/lock provided by DEC? Also, we've also observed that uVAXen don't have console keys at all! Is there a way to put one in? advTHANKSance, Mike R. ================================================================================ Note 100.1 Looking for a good key... 1 of 3 EISNER::NORTON 9 lines 10-FEB-1988 17:50 -< partial answer >- -------------------------------------------------------------------------------- >> Also, we've also observed that uVAXen don't have console keys at >> all! Is there a way to put one in? An outfit called Secure Technologies sells locking front and rear covers for BA23's, BA123's, H9642's and uV2000's. They're at 800-832-LOCK, or 617-326-7979 in Mass. I haven't used any of their products, but the pictures in the brochure look good. \bill ================================================================================ Note 100.2 Looking for a good key... 2 of 3 EISNER::HASSINGER "Bob Hassinger" 5 lines 10-FEB-1988 22:31 -< It just *looks* like a lock... >- -------------------------------------------------------------------------------- When I ask about the one-key-fits-all feature the usual answer is that it just *looks* like a lock. It is really a switch with a removable "knob". Bob H ================================================================================ Note 100.3 Looking for a good key... 3 of 3 EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 6 lines 11-FEB-1988 12:33 -< Thats what locked doors are for... >- -------------------------------------------------------------------------------- Thats why some of have locks on our computer rooms... We even keep our MicroVAX in the computer room. (I admit, I would have trouble keeping a VAXstation in the computer room and using it my office, however.) Seton  ================================================================================ Note 101.0 LN03 weirdness 2 replies EISNER::OSUDAR "John Osudar" 34 lines 12-FEB-1988 21:50 -------------------------------------------------------------------------------- I just spent a couple of weeks working on some LN03 fonts, modifying some TeX fonts to allow a word processor (MASS-11) to use them for making viewgraphs. Once the font work was done, I printed some test viewgraphs. On an LN03-Plus, they came out fine. On a plain LN03, some of the lines started at different places than others. Same file, same fonts, same rev level (DEC004.4) of the base firmware. The text was laid out like this: but the LN03 printed it like this: * This is a sample viewgraph, * This is a sample viewgraph, printed on the LN03 printed on the LN03 * And so on, and so on, and * And so on, and so on, and so on, all the way to the so on, all the way to the end of the page end of the page Positioning for the bullets and the first word on each line was done by absolute cursor position commands (HPA, "[nnn`"). In the lines that contain the bullets, there are some additional commands to select the DEC Tech font for the bullet symbol and go back to the text font for the rest; in those lines, the first word (i.e. "This" or "And" above) is positioned at the requested absolute horizontal position on the page, but the bullet is at the wrong place. On all the other lines, which contain no other commands, the first word starts at the wrong place (and always at the same wrong place!) Also, all of the mis-positioned text is off by the same distance. Each line is terminated by a carriage return, with no line feed. (The output file has embedded carriage control.) Does anyone have any ideas what this is all about? I have a vague recollection of seeing this before, and misinterpreting it to mean that a font change command requires a carriage return in order to avoid mis-positioned text. I now suspect that my earlier problem's "solution" was actually making all of the text be mis-positioned by the same amount, so that the final product "looked right". ================================================================================ Note 101.1 LN03 weirdness 1 of 2 EISNER::HAHN 3 lines 13-FEB-1988 09:48 -< Wrap/noWrap >- -------------------------------------------------------------------------------- It looks like you may have wrap on. Try to set Switch pack 2-4 set to off = no wrap. Pierre ================================================================================ Note 101.2 LN03 weirdness 2 of 2 EISNER::OSUDAR "John Osudar" 2 lines 15-FEB-1988 10:43 -< why wrap/nowrap? >- -------------------------------------------------------------------------------- How does wrap/nowrap affect absolute positioning commands? Am I missing something???  ================================================================================ Note 102.0 Western Datacomm - Vadic 4224 4 replies EISNER::TILLMAN "Brian Tillman, Smiths Industries." 3 lines 19-FEB-1988 13:40 -------------------------------------------------------------------------------- Anyone out there have any experience with a Western Datacomm Line Guard coupled with Vadic 4224 modems for dialback? If so, have they been reliable? ================================================================================ Note 102.1 Western Datacomm - Vadic 4224 1 of 4 EISNER::ERSKINE "Chris Erskine" 5 lines 22-FEB-1988 06:23 -< We like it >- -------------------------------------------------------------------------------- I installed one here and on the most part have not had any problems. There are some times where it may hang but it is very rare. If I had to do it again, I would. To save people money, we put in an 800 number to dial in. This means my users do not have to spend any money to initiate the call even. ================================================================================ Note 102.2 Western Datacomm - Vadic 4224 2 of 4 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 15 lines 22-FEB-1988 08:32 -< The trouble we've seen. >- -------------------------------------------------------------------------------- We had a lot of problems installing this combination, Chris. After a week of trying, we finally got it so we MIGHT get a call-back. We sent it back to the dealer and told them to get it working, then bring it back and install it. It turns out the documentation concerning some of the jumpers and switches for the modem boards was in error. It's reinstalled and the modems work flawlessly, but we're still having problems with the line guard. We can put two call-back names in it, but as soon as we put in the third, the first acquires the same phone number as the third. The fourth name will cause the second name to get its phone number and the fifth will cause both the first and third to get its number. We called the dealer again and they said to send it back to Western Datacomm. We said, "Nuts! You sold us this trash and it's still under warranty. You send us a new board and we'll send back the old one." So far, no response. I guess we just picked a few lemons. ================================================================================ Note 102.3 Western Datacomm - Vadic 4224 3 of 4 EISNER::ERSKINE "Chris Erskine" 14 lines 23-FEB-1988 06:38 -< Do you know how to make lemonade? >- -------------------------------------------------------------------------------- Of the sites that I know of, we have not had any problem. I even have used the dialup to create an async DECnet link through a router box. We use the 407 Tone Modems for input on 2 of the 3 lines. I of the numbers is an incoming 800 number which says the user does not even have to pay for the first call. The modems have changed a couple of times since I first got them. I have some of the first MNP modems from Racal and had to convert them from normal to MNP support. There was one bug that I found in the manual for that be I thought is was due to changing the size of the ROM's from 8K to 16K. We have had ours for 2-3 years. The rep for the Western Datacom and Racal was very helpful and even had the President of Western Datacom come out and discuss changes that were coming. ================================================================================ Note 102.4 Western Datacomm - Vadic 4224 4 of 4 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 2 lines 23-FEB-1988 09:09 -< Update >- -------------------------------------------------------------------------------- We received new ROMs for the Lineguard yesterday and installed them. Everything seems to work now. Perhaps we'll wind up happy with it.  ================================================================================ Note 103.0 DF124 switch setting info needed 22 replies EISNER::KENNEDY "Terry Kennedy" 6 lines 23-FEB-1988 23:42 -------------------------------------------------------------------------------- Could someone upload the switch settings for a DEC DF124 modem? We got a bunch as a donation (without manuals, of course) and DEC [in]direct is quoting 60-90 days. Someone went after these units as the switches are *all* off, so I need the complete list... Thanks... ================================================================================ Note 103.1 DF124 switch setting info needed 1 of 22 EISNER::ROCHELEAU "Rod Rocheleau" 99 lines 24-FEB-1988 19:28 -< Switch packs and jumpers ... whew!! >- -------------------------------------------------------------------------------- < Note 103.0 by EISNER::KENNEDY "Terry Kennedy" > -< DF124 switch setting info needed >- >>> Could someone upload the switch settings for a DEC DF124 modem? From the trusty dusty mini-reference manual Volume 3 for DF124-AM and DF124-BM Switchpack 1 s1 long space diconnect enable=off s2 loss of carrier disconnect enable=off s3 answer on originate ans=off orig=on s4 slave clock receive=on external=off s5 master clock internal=off external=on s6 remote loopback v.54=off v.22=on s7 parity select parity=on noparity=off s8 local echo enabled=on disabled=off s9 parity type odd=on even=off s10 fallback select ccitt=on (-BM) bell=off (-AM) S1 (E48) factory default 1 2 3 4 5 6 7 8 9 10 o o o o o o o o o o f f f f f n f n f f f f f f f f f f switchpack 2 s1 character length 8 bits/10 bits=on 9 bits/11 bits=off s2 async rate basic=on extended=off s3 character length 8 bits/9 bits=on 10 bits/11 bits=off s4 mode select sync=off async=on s5 response message terminal=off cpu=on s6 MI enabled=off disabled=on s7 auto answer manual=on auto=off s8 PSTN or P/LTN pstn=off p/ltn=on s9 Signal detector threshold -49 dbm=off -37 dbm=on s10 retrain enabled=on disabled=off s2 (E33) factoru default 1 2 3 4 5 6 7 8 9 10 o o o o o o o o o o n f f n f n f f f n f f f f f f switchpack 3 s1 abort timer enabled=on disabled=off s2 DTR external=on forced ON=off s3 RTS external=off forced ON=off s4 interface rate select enabled=on (-BM) disabled=off (-AM) s5 guard tone select enabled=off (-BM) disabled=on (-AM) s6 pulse dialing duty cycle 60%=off (-AM) 67%=on (-BM) s7 interface local loop control enabled=on disabled=off s8 interface remote digital enabled=on disabled=off loop control s3 (E23) factory default 1 2 3 4 5 6 7 8 o o o o o o o o n n f n f f n n f f f switchpack 4 PSTN xmit level s1 s2 s3 s4 -2dbm off off off off This option is disabled for -3 on off off off DF124-AM, it is hard wired for -4 off on off off -10 dbm -5 on on off off ........ for the DF124-BM it is set to -16 off on on on -9 dbm -17 on on on on P/LTN xmit level s5 s6 s7 s8 -2dbm off off off off This is the same counting sequence -17dbm on on on on as S1 through s4 DF124-AM is set to -2 dbm DF124-BM is set to -14 dbm s4 (E18) factory default 1 2 3 4 5 6 7 8 o o o o o o o o f f f f f f f f f f f f f f f f s5 (beside s4) is for field service use only. The manual lists the X-Y scope tests this switchpack is used for. factory defaults 1 2 3 4 5 6 7 8 o o o o o o o o f n f f n n f f f f f f f The jumpers listed for U.S./Canada are as follows ... Thats a lot of typing W1=in w2=out w3=in w4=out w5=in w6=n/a w7=out w8=in w9=in w10=in w11=in w12=in w13=out w14=in w15=in w16=out w17=n/a w18=n/a w19=in r198=in If you need any more info give me a call at 214-506-6036 Rod Rocheleau ================================================================================ Note 103.2 DF124 switch setting info needed 2 of 22 EISNER::KOZAM 20 lines 25-FEB-1988 02:25 -< Making the DF124s pay attention to DTR >- -------------------------------------------------------------------------------- -< Switch packs and jumpers ... whew!! >- Gee, I also happened to get a few DF124's in about a week ago. I got the DF124 Modem Family User Guide from DECDirect (Quoted 30 days, but I got it in under 2 weeks). I've got them set up according to factory settings, but have one major problem: they always answer, regardless of the state of DTR. In fact, I even completely removed the RS232 connector, but they still answer. And no, I don't have DTR forced ON. The only way I could get them NOT to answer was to set the switch for MANUAL ANSWER to the ON position, but then they NEVER answer (regardless of the position of the DATA/TALK switch on the front panel) unless you physically pick up a telephone connected to them, then fiddle with the DATA/TALK switch. Anyone else encounter this problem? Any solutions? Marc Kozam ================================================================================ Note 103.3 DF124 switch setting info needed 3 of 22 EISNER::KENNEDY "Terry Kennedy" 16 lines 8-MAR-1988 03:55 -< Switches, Rev. 2 >- -------------------------------------------------------------------------------- Ok, I decided my 124's were broken and I DECMailered one of 'em to see what would happen. It actually came back with a sheet of paper listing the switch settings. Now, I'm even more confused: o I can get it to answer either at 1200 or 2400, depending on switch settings, but can't get it to guess automatically. This sort of renders it useless. Symptom: when running with mismatched baud rate, it answers, gives tone, calling modem gives tone (both whistling at each other), but *no* carrier detect on either end. o The darned thing also answers *and* gives carrier at 300 baud, but won't pass any data to the host system. I yanked another DF124 out of my remote service port and it behaves the exact same way. This has got to be a jumper problem - even DEC couldn't do this bad when building a modem. (close, but not this bad) ================================================================================ Note 103.4 DF124 switch setting info needed 4 of 22 EISNER::ROCHELEAU "Rod Rocheleau" 33 lines 13-MAR-1988 16:04 -< Switched from a live working DF124 >- -------------------------------------------------------------------------------- >>> < Note 103.3 by EISNER::KENNEDY "Terry Kennedy" > >>> -< Switches, Rev. 2 >- Ok I just pulled one from a dialin rack that was functioning. Here are the switches. o = on or closed...f = off or open s1 1 2 3 4 5 6 7 8 9 10 f f f f f o f o f f s2 1 2 3 4 5 6 7 8 9 10 o f f o f o f f f o s3 1 2 3 4 5 6 7 8 o o f f o f o o s4 1 2 3 4 5 6 7 8 f f f f f f f f s5 1 2 3 4 5 6 7 8 f o f f o o f f Jumpers ... those not listed are in w2, w4, w13, w16 all fromt panel switched are out. Hope this helps Rod ================================================================================ Note 103.5 DF124 switch setting info needed 5 of 22 EISNER::KENNEDY "Terry Kennedy" 3 lines 14-MAR-1988 03:26 -< 1200 *and* 2400? >- -------------------------------------------------------------------------------- Does the one in your rack work for both 1200 and 2400 incoming calls without switching anything? That is the major problem I am now having - will try your new settings tomorrow - thanks... ================================================================================ Note 103.6 DF124 switch setting info needed 6 of 22 EISNER::KILLEEN "Jeff Killeen" 8 lines 14-MAR-1988 08:09 -< PIN WHAT PIN ON THE DHV11 >- -------------------------------------------------------------------------------- > Does the one in your rack work for both 1200 and 2400 incoming > calls without switching anything? That is the major problem I am > now having - will try your new settings tomorrow - thanks... We sent back a bunch of DEC DF1xx modems because the only way they would change speed was if a pin was raised - and non of the DEC muxs had the capability to raise that pin. I believe they were DF124's but I am not 100 percent sure since it was 12 months ago. ================================================================================ Note 103.7 DF124 switch setting info needed 7 of 22 EISNER::HASSINGER "Bob Hassinger" 1 line 14-MAR-1988 13:05 -< DECserver 200s do control speed on pin 12... >- -------------------------------------------------------------------------------- ================================================================================ Note 103.8 DF124 switch setting info needed 8 of 22 EISNER::KENNEDY "Terry Kennedy" 7 lines 14-MAR-1988 19:22 -< DEC modems - still useless after all these years >- -------------------------------------------------------------------------------- > We sent back a bunch of DEC DF1xx modems because the only way they would > change speed was if a pin was raised - and non of the DEC muxs had the... Brilliant! You can't get incoming calls because the modem will not give carrier unless it knows what spped you're calling in at, which can only be determined by the host CPU after a successful autobaud, which can never happen! ================================================================================ Note 103.9 DF124 switch setting info needed 9 of 22 EISNER::KENNEDY "Terry Kennedy" 9 lines 14-MAR-1988 19:27 -< If you cant get it (at least) half-right, don't do it at all... >- -------------------------------------------------------------------------------- By the way, I read a recent Digital News article about DEC's new security modems that quoted DEC's modem product line manager as saying "We make modems so we can sell a complete system, and that is the only reason". When will DEC ever learn that they should not waste time on things like this - if they feel a crucial need for such a product, sign a contract for a few zillion of the things with someone who knows how to make working product - just like RD and RX33 series drives... ================================================================================ Note 103.10 DF124 switch setting info needed 10 of 22 EISNER::KILLEEN "Jeff Killeen" 11 lines 14-MAR-1988 20:07 -< I THINK THEY DID IT RIGHT THIS TIME >- -------------------------------------------------------------------------------- > When will DEC ever learn that they should not waste time on things > like this - if they feel a crucial need for such a product, sign a Generally I agree with this - however I have just got done playing with the DF242 modems we installed on DECUServe. I am seriously considering putting these on future systems we sell. They look like they will work better with PDP/VAX systems than any other modem I have seen. They have the ^B activation feature which prevents the modem and system from looping back on each other. Also they have a full "AT" command set. ================================================================================ Note 103.11 DF124 switch setting info needed 11 of 22 EISNER::GARDNER "Tim Gardner" 14 lines 15-MAR-1988 06:59 -< ^B and AT! Wow!! >- -------------------------------------------------------------------------------- Re: < Note 103.10 by EISNER::KILLEEN "Jeff Killeen" > > They have the ^B activation feature which prevents the modem and system > from looping back on each other. Also they have a full "AT" command set. Holy cow!! This is what I've been looking for! Can you tell us more? Like how well they work as both INcoming and OUTgoing modems? Like what happens when caller one uses 1200 baud and caller two uses 2400 baud? Does the RS232 port use a constant speed output to the host, or does it vary with modem speed? tg ================================================================================ Note 103.12 DF124 switch setting info needed 12 of 22 EISNER::KILLEEN "Jeff Killeen" 16 lines 15-MAR-1988 08:43 -< DOES THIS ANSWER THE QUESTION? >- -------------------------------------------------------------------------------- > Like how well they work as both INcoming and OUTgoing modems? > Like what happens when caller one uses 1200 baud and caller two > uses 2400 baud? Does the RS232 port use a constant speed output to the > host, or does it vary with modem speed? On incomming the modem sychs to whatever speed/protocol the incomming user is using. On outgoing it will autobaud to whatever speed the line is set to. When the modem is activated by a ^B it autobauds on that character - or you can set a fixed speed between the modem and the user (1200,2400,4800,9600). The thing I like about ^B activation is it is not a normal character the O/S will send a terminal - therefore you don't get your system turning on your modem by mistake. ================================================================================ Note 103.13 DF124 switch setting info needed 13 of 22 EISNER::STAMERJOHN "RW Stamerjohn" 9 lines 16-MAR-1988 14:19 -< DF242 and Racal-Vadic >- -------------------------------------------------------------------------------- The new modems (DF242?) do seem too take longer to connect up with my 1200 baud Radic-Vadics's, long enough for the R-V assume 'FAILED CALL' and hang-up. This keepts from keyboard dialing DCS. I can get my modem to work if I manual dial the phone and switch the toggle to data. This is a small hassle as the modem is three rooms away. (there was no problem with the old modems, on here or DCS). ================================================================================ Note 103.14 DF124 switch setting info needed 14 of 22 EISNER::KILGALLEN "Larry Kilgallen" 5 lines 16-MAR-1988 14:40 -< May Be Ignoring My VA3451-PA's Also >- -------------------------------------------------------------------------------- Has this machine been down a lot since last Friday's upgrade? If not, I have been having the same problem (also with Vadic 3451-PA's at all sites I dial in from). It is not nearly so bad as the DEC Electronic Store, however, where I connect 1 time in 40. ================================================================================ Note 103.15 DF124 switch setting info needed 15 of 22 EISNER::KILLEEN "Jeff Killeen" 0 lines 16-MAR-1988 19:47 -< WE HAVE NOT BEEN DOWN SINCE FRIDAY >- -------------------------------------------------------------------------------- ================================================================================ Note 103.16 DF124 switch setting info needed 16 of 22 EISNER::WALLIS "Keep the 'us' in DECUServe" 4 lines 18-MAR-1988 01:39 -< FYI >- -------------------------------------------------------------------------------- I dial in using a Hayes compatible modem @ 1200 baud and have no problems. I have noticed neither a slowdown when initially connecting nor any other changes (except I haven't gotten any line noise recently). ================================================================================ Note 103.17 DF124 switch setting info needed 17 of 22 EISNER::KILGALLEN "Larry Kilgallen" 4 lines 19-MAR-1988 07:32 -< Hayes compatible? >- -------------------------------------------------------------------------------- My understanding of the phrase "Hayes compatible" is that it deals strictly with the command set used to control the modem from the DTE, and has nothing to do with the modulation scheme used over the telephone network. ================================================================================ Note 103.18 DF124 switch setting info needed 18 of 22 EISNER::KILLEEN "Jeff Killeen" 5 lines 19-MAR-1988 11:13 -< CORRECT >- -------------------------------------------------------------------------------- > My understanding of the phrase "Hayes compatible" is that it deals > strictly with the command set used to control the modem from the > DTE, and has nothing to do with the modulation scheme used over > the telephone network. ================================================================================ Note 103.19 DF124 switch setting info needed 19 of 22 EISNER::WALLIS "Keep the 'us' in DECUServe" 12 lines 19-MAR-1988 17:40 -< I guess my ignorance is showing >- -------------------------------------------------------------------------------- >> My understanding of the phrase "Hayes compatible" is that it deals >> strictly with the command set used to control the modem from the >> DTE, and has nothing to do with the modulation scheme used over >> the telephone network. Well I guess a software guy shouldn't venture into the realms of hardware help ;-). On looking at the documentation its Bell 103 and 212A compatible (I say this even though my mother told me not to use terms I don't know the meaning of). Barry ================================================================================ Note 103.20 DF124 switch setting info needed 20 of 22 EISNER::CETRON 9 lines 12-APR-1988 22:53 -< hmm, I've had problems too >- -------------------------------------------------------------------------------- I've found the new modems will answer fine at 2400, give carrier etc, and my modems will show "connect 2400" and then I will type a and voila! a 1200 baud connect to EISNER....woops... It has kept me off for quite a while (the modem I am forced to use is on a unix box and for POLITICAL reasons must call at 2400 always..) -ed ================================================================================ Note 103.21 DF124 switch setting info needed 21 of 22 EISNER::KENNEDY "Terry Kennedy" 33 lines 13-APR-1988 02:25 -------------------------------------------------------------------------------- > I've found the new modems will answer fine at 2400, give carrier > etc, and my modems will show "connect 2400" and then I will type > a and voila! a 1200 baud connect to EISNER....woops... I think this is a system problem versus a modem problem - I know of other VAX systems with these (DF224) modems which do not exhibit the same problem (I've had it, too). Most DEC interfaces can't 'really' autobaud - they just look for a few specific characters and error conditions and 'guess'. 'Real' autobaud requires you to time a char- acter from start bit to stop bit (with external hardware) and then compute the speed. The requirements of such systems for you to type a specific character is usually so it can guess parity or code set (ASCII / EBCDIC / etc) at the same time. (You can prove this with Tymnet, which claims to want an 'A'. CR's get you a 'please type your terminal identifier' message. Back to the issue - since recent DEC terminal controllers are intelligent (on-board Intel 8048-series, for example, in the DHV11), they seem to want some set-up time after carrier before they gettheir act together. You may want to wait 5 seconds or so and see if the problem goes away. The problem might also be related to a VMS bug I've SPR'd about XOFF'd lines not correctly autobauding for the next caller. The problem Larry mentions earlier is because these (and some other) modems give a different answer tone before giving the 'right' answer tone. Therefore, you'll have the CONNECT message delayed at least the 3 seconds or so from the first tone, plus a bit if such tone confused your modem. By the way, it *does* con- fuse anything capable of using Vadic 1200 modulation. I think the first tone is a call for protocol for one of the error-correcting protocols. ================================================================================ Note 103.22 DF124 switch setting info needed 22 of 22 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 13 lines 13-APR-1988 08:56 -< Good suggestion. >- -------------------------------------------------------------------------------- RE: < Note 103.21 by EISNER::KENNEDY "Terry Kennedy" > | Back to the issue - since recent DEC terminal controllers are |intelligent (on-board Intel 8048-series, for example, in the |DHV11), they seem to want some set-up time after carrier before |they gettheir act together. You may want to wait 5 seconds or so |and see if the problem goes away. Indeed, this helps. I find consistently that if I start typeing my CRs immediately after I see my connect message, even though the connection says 2400, I get garbage spewed at me, but if I just let it sit for a few seconds, either EISNER will put up the username prompt for me by itself, or my CRs will elicit the prompt.  ================================================================================ Note 104.0 Ethernet Hardware Interface Chips 6 replies EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 20 lines 25-FEB-1988 22:15 -------------------------------------------------------------------------------- First of all, excuse this note. there was an accident on my keyboard today and the SHIFT key is not working. (among other things]. i have to use the SHIFT LOCK to type in UPPERCASE etc. i am looking at comparing performace of of dec's ethernet boards versus 3rd party offerings, particulary a popular line of programmable ethernet boards. the 3rd party boards all interface with ether using an off the shelf intel chip, something like the 82578 (?. i am interested in finding out what chip or chips or non-vlsi circuitry dec is using in their boards for the ether interface to handle data-link functions. are they all proprietary elecTroincs, do they use a variety of other manufacturers chips, or what? if i can't get an answer can someone supply me with a pointer to where to look? thanks in advance, apologies for the messy presentation level. ================================================================================ Note 104.1 Ethernet Hardware Interface Chips 1 of 6 EISNER::SHANNON "Terry C. Shannon (R.I.P. Charlie M." 3 lines 26-FEB-1988 08:56 -< Ethernet's "Cutting Edge"??? >- -------------------------------------------------------------------------------- I believe DEC uses the so-called LANCE Ethernet chip. ================================================================================ Note 104.2 Ethernet Hardware Interface Chips 2 of 6 EISNER::KENNEDY "Terry Kennedy" 13 lines 26-FEB-1988 18:06 -------------------------------------------------------------------------------- > I believe DEC uses the so-called LANCE Ethernet chip. Right - I think this chip is from National Semi. Note that the DEUNA did it with loose parts. The DEQNA (and also the DELQA, I think) are 'powered' by an Intel iAPX 86 (80186, etc family) chip driving the Lance. It appears the 80x86 familiy is now the favorite chip to use (over the T-11) in DEC's intelligent controllers (it also appears on the RQDX3 and the TK50 controller). Also, remember that most non-DEC Ethernet cards do not emulate DEC products, and so come with their own drivers and cannot [readily] be used with DEC software like DECnet. The only exception to this that I know of is the MEQNA, a DEQNA emulation. ================================================================================ Note 104.3 Ethernet Hardware Interface Chips 3 of 6 EISNER::FRISBIE "Alan E. Frisbie" 18 lines 26-FEB-1988 21:11 -< Different flavored chips >- -------------------------------------------------------------------------------- >> The DEQNA (and also the DELQA, I think) are >> 'powered' by an Intel iAPX 86 (80186, etc family) chip The two DEQNA boards that I have here use the Intel 8051 (masked ROM version) and the 8751 (EPROM version) CPU/ROM chips (yes, I have an unusual FCO). True, they do have PDP-11 code in their ROM, but it is just the downline-boot code for PDP-11s. >> It appears the 80x86 familiy is now the favorite chip to use >> (over the T-11) in DEC's intelligent controllers (it also appears on >> the RQDX3 and the TK50 controller). Well, I can't find a T-11 on the TK50, but there sure is one on my RQDX3. Could there be a later revision that uses the 80x86? Alan ================================================================================ Note 104.4 Ethernet Hardware Interface Chips 4 of 6 EISNER::KENNEDY "Terry Kennedy" 19 lines 26-FEB-1988 22:55 -< Operating instructions: 1) open mouth 2) insert foot >- -------------------------------------------------------------------------------- > The two DEQNA boards that I have here use the Intel 8051 > (masked ROM version) and the 8751 (EPROM version) CPU/ROM chips... Oops - I meant to reverse DELQA and DEQNA there. You are correct on the DEQNA. > Well, I can't find a T-11 on the TK50, but there sure is > one on my RQDX3. Could there be a later revision that > uses the 80x86? You got me again - the RQDX3 is T-11 powered (if you can call it that...). It's claim to fame is the use of the SMC HD controller chipset instead of the DEC 'glue engine' on older RQDXn. It is a 3rd-party RQDX emulator that is 80186 powered. And now, for something completely similar...  The RRD50 (CD-ROM) Q-bus interface is 80186-powered. Yes, I checked this time! ================================================================================ Note 104.5 Ethernet Hardware Interface Chips 5 of 6 EISNER::ROBITAILLE "From the Mac of Mike Robitaille" 3 lines 1-FEB-1989 01:18 -< What about the DEBNA? >- -------------------------------------------------------------------------------- Does DEC use the Lance chip on the BI Ethernet Card (DEBNA?) as well? What's so different about this beastie that they will only rate it at 5 megabits/s? ================================================================================ Note 104.6 Ethernet Hardware Interface Chips 6 of 6 EISNER::KENNEDY "Terry Kennedy" 15 lines 1-FEB-1989 09:22 -< I believe *all* dec boards are 5 Mb/s or less >- -------------------------------------------------------------------------------- > Does DEC use the Lance chip on the BI Ethernet Card (DEBNA?) as well? I don't know. But it *seems* likely. Everything I've seen from DEC for EThernet has been LANCEd in recent years. > What's so different about this beastie that they will only rate it > at 5 megabits/s? Different? I haven't seen DEC rate any of their Ethernet products for > 5 Mb/s... The software (DECnet) certainly can't run flat out at 5 anyway... Note that this is an issue of how much of the 10 Mb/s Ethernet bandwidth any one controller can hog, as opposed to the internal Xerox test im- plementation of Ethernet at 3 Mb/s, which isn't interoperable with cur- rent 10 Mb/s Ethernet.h  ================================================================================ Note 105.0 M9301 Bootstrap info wanted 3 replies EISNER::FRISBIE "Alan E. Frisbie" 27 lines 4-MAR-1988 15:45 -------------------------------------------------------------------------------- I need some information on the M9301 Unibus Bootstrap board: 1) What is the function of the five jumpers W1 - W5? 2) Why must they be removed when the board is used with an 11/34? The 11/34 User's Guide says they must, but doesn't explain why. 3) Is there a ROM set for the M9301 that will boot both RK05 and RL02 disks? If not, what would you suggest as a replacement? Three months ago, DEC serviced my client's 11/34 and replaced the M9301. Ever since then, the bootstrap has been very flakey. DEC came out several times to try and fix it and never succeeded. They did have some novel explainations, such as the RK05-F disk platter being corroded. It is in perfect condition. Yesterday, I pulled out the M9301 and checked it against the installation notes in the 11/34 User's Guide. Yes, all five jumpers were installed on the M9301. Removing the jumpers fixed the entire problem. Yay, Field Service! Needless to say, the client does not intend to pay for the Field Service calls. Alan ================================================================================ Note 105.1 M9301 Bootstrap info wanted 1 of 3 EISNER::FULLER_S "Stuart Fuller" 6 lines 28-APR-1988 14:02 -< Here's why! >- -------------------------------------------------------------------------------- The jumpers W1 thru W5 are used to include BUS GRANT pullup resi{tors. Since the 11/04 and 11/34 processor modules have these resistors on board, you must pull the jumpers on the M9301. Otherwise, the GRANT lines will be pulled too high. Stu ================================================================================ Note 105.2 M9301 Bootstrap info wanted 2 of 3 EISNER::BOSTWICK 9 lines 24-JUN-1988 18:53 -< depends upon where you put the card >- -------------------------------------------------------------------------------- > The jumpers W1 thru W5 are used to include BUS GRANT pullup resi{tors. > Since the 11/04 and 11/34 processor modules have these resistors > on board, you must pull the jumpers on the M9301. Otherwise, the > GRANT lines will be pulled too high. This applies only if the M9301 is in the (usual for 11/34) position at the CPU end of the UNIBUS. If it is installed in place of the terminator, at the 'back' end of the bus, leave the jumpers in. ================================================================================ Note 105.3 M9301 Bootstrap info wanted 3 of 3 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 24 lines 5-JUL-1988 13:50 -< M9312 upgrade for M9301-XX >- -------------------------------------------------------------------------------- < Note 105.0 by EISNER::FRISBIE "Alan E. Frisbie" > -< M9301 Bootstrap info wanted >- >>>Is there a ROM set for the M9301 that will boot both >>>RK05 and RL02 disks? If not, what would you suggest >>>as a replacement? I don't know if there is a M9301 chip, but those were a pain because you had to take DEC's selection of bootable devices. The newer M9312s will boot most anything, as individual chips are used for each device (or sometimes for several related devices), so you can fill the boot board with whatever assortment you need. The card takes 4 chips. A few (e.g. network) boot proms are on more than one chip, but the devices you ask for should be a chip each. The M9312 boot chips also fit the 11/24's KT24 and can be used on the PMI/Unibus adapter in the 11/84 if your device isn't one the 11/84 otherwise can boot. M9312s used to cost a lot more than M9301-XXs, but now are VERY cheap as are most Unibus cards. It may well cost more to get the right chips, that the card costs, so it is worth while shopping with helpful dealers that may have the chips needed and throw them in with the board, rather than dealing with iron mongers that just give you the next card in the pile.  ================================================================================ Note 106.0 Lap_top Terminals 12 replies EISNER::RICE "Gary Rice" 5 lines 7-MAR-1988 15:15 -------------------------------------------------------------------------------- I am in the market for some Lap-Top terminals. Do you use one that you particularly like? Gary ================================================================================ Note 106.1 Lap_top Terminals 1 of 12 EISNER::KILLEEN "Jeff Killeen" 8 lines 7-MAR-1988 17:59 -< RANDOM PORTABLE >- -------------------------------------------------------------------------------- I am using the RANDOM COLLEAGUE plus. I am very happy with it. Two options I would get are the back-lit screen and the keypad. Skip the word processor. I went the Lap-Top terminal route rather than the Lap-Top computer route because I did not want to put up the keyboard mapping issues. The RANDOM terminal is pretty close to a VT100 keyboard. ================================================================================ Note 106.2 Lap_top Terminals 2 of 12 EISNER::RICE "Gary Rice" 5 lines 8-MAR-1988 08:06 -< How do I get one? >- -------------------------------------------------------------------------------- > I am using the RANDOM COLLEAGUE plus. Do you have an address for the company that makes the uit? Gary ================================================================================ Note 106.3 Lap_top Terminals 3 of 12 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 3 lines 8-MAR-1988 08:37 -------------------------------------------------------------------------------- We have some of the Radio Shack TRS-1000s. I personally don't use them, but those who do think it does a pretty good job. I don't think it emulates a VT100, though. ================================================================================ Note 106.4 Lap_top Terminals 4 of 12 EISNER::KILLEEN "Jeff Killeen" 7 lines 8-MAR-1988 10:14 -< ADDRESS OF RANDOM >- -------------------------------------------------------------------------------- > Do you have an address for the company that makes the unit? Random Corp 581 Northland Blvd Cincinnati Ohio 45240 513 825-0880 ================================================================================ Note 106.5 Lap_top Terminals 5 of 12 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 1 line 8-MAR-1988 17:30 -< Another suggestion >- -------------------------------------------------------------------------------- See also Notes 85.* in this conference. ================================================================================ Note 106.6 Lap_top Terminals 6 of 12 EISNER::GARDNER "Tim Gardner" 11 lines 9-MAR-1988 12:32 -< Terminal? Who needs a dumb old terminal? >- -------------------------------------------------------------------------------- Re: < Note 106.0 by EISNER::RICE "Gary Rice" > -< Lap_top Terminals >- >I am in the market for some Lap-Top terminals. Do you use one that ^^^^^^^^^ >you particularly like? Why not just buy a lap top PC and a communications package. Not much additional cost, and infinitely more useful! tg ================================================================================ Note 106.7 Lap_top Terminals 7 of 12 EISNER::RICE "Gary Rice" 6 lines 9-MAR-1988 13:57 -------------------------------------------------------------------------------- > Why not just buy a lap top PC and a communications package. Not much > additional cost, and infinitely more useful! I HATE MS-DOS! A terminal is infinitely simpler. Gary ================================================================================ Note 106.8 Lap_top Terminals 8 of 12 EISNER::KILLEEN "Jeff Killeen" 1 line 9-MAR-1988 18:10 -< *** DITTO *** >- -------------------------------------------------------------------------------- > A terminal is infinitely simpler. ================================================================================ Note 106.9 Lap_top Terminals 9 of 12 EISNER::GARDNER "Tim Gardner" 13 lines 10-MAR-1988 09:19 -< ñ >- -------------------------------------------------------------------------------- Re: < Note 106.7 by EISNER::RICE "Gary Rice" > > I HATE MS-DOS! A terminal is infinitely simpler. Simpler, sure. My point is that the PC is more useful than a terminal. Besides, who said anything about MS-DOS? You are certainly not limited to DOS on lap top PCs (Note: the term PC does not mean IBM-PC). You can buy a Macintosh laptop, or as someone else has suggested, a Radio Shack Model 100 or 200 (not especially recommeded if you expect to do a lot of typing). tg ================================================================================ Note 106.10 Lap_top Terminals 10 of 12 EISNER::HASSINGER "Bob Hassinger" 17 lines 10-MAR-1988 10:22 -< LAP top *terminals* SHOULD find a valid, distinct market... >- -------------------------------------------------------------------------------- Lap top terminals certainly look like they would be interesting for some of our needs. Our people (non-computer types) travel a good bit and would like to be able to call back here to read and write MAIL and maybe run a few applications from time to time. They DON'T WANT TO KNOW about another computer and they don't need one! They WANT a pocket size VT220 that acts just like the one on their desk at the lowest possible cost/size/weight/training. The terminal approach SHOULD be better on all these scores. It seems however that the lap top products are not as well developed as the PCs at this point so the differences are not as distinct as one would like and expect. I am not satisfied so far with the available products with respect to easy to use, low training emulation for DEC VT100/200/300 oriented applications, particularly things that use the full screen and the keypad like EDT, TPU, FMS applications, etc. Bob H ================================================================================ Note 106.11 Lap_top Terminals 11 of 12 EISNER::HANRAHAN "Jamie Hanrahan" 46 lines 22-APR-1988 16:16 -< A vote for portable PCs >- -------------------------------------------------------------------------------- I too would rather not learn MS-DOS, but... According to this week's Digital News, you can buy a Random Corp. "Colleague" with a built-in 2400 bps modem for $1195. The unit weighs 7.5 pounds. For $300 more you get internal word-processing software and 65 Kbytes of nonvolatile RAM. No mention of battery operability. I can also buy an Amstrad laptop PC with: 640 K RAM 2 720K 3-1/2 floppies ( = nonvolatile RAM) 8 MHz 8086 processor full-size enhanced-style keyboard backlit, supertwist LCD screen built-in 2400 bps modem power from AC, batteries, or car adapter socket for PC bus expansion box socket for 8087 math coprocessor MS-DOS, communications, word processing, spreadsheet, and database software weight under 12 pounds ... for $1299. Now, I'm no MS-DOS fan, but I use it. And my experience is that I don't spend enough time talking to MS-DOS to be all that bothered by it -- as I spend most of my time on PCs inside applications. I used a system like this on a recent trip and got along with it just fine! I found out that a DECUS deadline had almost gotten by me, and, worse, I hadn't brought along the files I needed to work on the project. No problem, I just dialed the 8200, started a Kermit download, and went to dinner. By the time I got back from dinner the transfer (about an hours' worth) was done and my Procomm [a shareware communications package] script had hung up the phone for me, avoiding excess l-d charges. I spent about four hours editing, dialed DCS, started a Kermit upload, and went to sleep. This time my Procomm script woke me with bells when the transfer was done, so I could move the stuff from DCS's Kermit upload area (which gets cleaned out upon logout) into All-in-None mail. If I trusted DCS more I could have had Procomm do that auto- matically, but what the heck. Try doing that with a portable terminal! If you really want the laptop PC to feel more like a VAX, you can always get PC/EDT and PC/DCL for it, right? ================================================================================ Note 106.12 Lap_top Terminals 12 of 12 EISNER::EARLE "David Earle" 7 lines 18-JUN-1988 15:35 -< Random Colleague problems >- -------------------------------------------------------------------------------- I am using the RANDOM COLLEAGUE plus. I am very happy with it. Two options I would get are the back-lit screen and the keypad. Skip the word processor. See my note 36.4 on the problems I am having with a Colleage. Would appreciate any help.  ================================================================================ Note 107.0 LA-50 problems! 12 replies EISNER::DELARISCH "Arnold S. De Larisch" 28 lines 7-MAR-1988 20:24 -------------------------------------------------------------------------------- I don't know if this is the right conference ... but I'll start here. I have a friend at a RSTS site who is having trouble with an LA-50 printer. The printer is connected to a Rainbow and a DZ-11 port on a PDP 11/44 thru a switch box. The length of the cable between the PDP 11/44 and the printer is about 200 feet or so. The LA-50 printer works just fine at 4800 baud when the data is comming from the Rainbow. The printer works fine on the PDP-11 for the first page or so ... it then prints out a bunch of backward question marks and stops. Lowering the baud rate only causes the problem to occur latter (at 1200 Baud it gets about 2 pages before going out to lunch). If you lower the thing to 300 baud, it works fine (but is VERY slow). The printer has been swapped out a number of times, with no difference. I suspect the problem to be related to the XON/Xoff handling. The PDP-11 port is set for HOSTSYNC and TTYSYNC (I know only one of these is needed, but both are currently enabled.) The LA-50 dip switches are set for XON/XOFF recognition (if the manual doesn't have a misprint!). The operating system is RSTS/E V9.1 ! Any ideas? -Arnold ================================================================================ Note 107.1 LA-50 problems! 1 of 12 EISNER::KILLEEN "Jeff Killeen" 5 lines 7-MAR-1988 21:00 -< QUESTIONS >- -------------------------------------------------------------------------------- Is the LA50 set seven or eight bit? If it is set seven bit is it set seven/space parity? - and of course is the RSTS system set the same? P.S. on RSTS 7-bit no parity = 7-bit/space parity (just like VMS) ================================================================================ Note 107.2 LA-50 problems! 2 of 12 EISNER::DELARISCH "Arnold S. De Larisch" 10 lines 7-MAR-1988 21:12 -< 8 Bit /No Parity! >- -------------------------------------------------------------------------------- >> < Note 107.1 by EISNER::KILLEEN "Jeff Killeen" > >> -< QUESTIONS >- >> Is the LA50 set seven or eight bit? If it is set seven bit is it set >> seven/space parity? - and of course is the RSTS system set the same? Sorry about forgetting that ... Eight Bit, no parity! -Arnold ================================================================================ Note 107.3 LA-50 problems! 3 of 12 EISNER::KILLEEN "Jeff Killeen" 1 line 7-MAR-1988 22:35 -< SET TERM/PERM/EIGHT KBxx: >- -------------------------------------------------------------------------------- Are you sure the RSTS is set EIGHT_BIT? The norm in RSTS land is 7-bit. ================================================================================ Note 107.4 LA-50 problems! 4 of 12 EISNER::DELARISCH "Arnold S. De Larisch" 9 lines 8-MAR-1988 02:38 -< Yup ... 8-bit! >- -------------------------------------------------------------------------------- >> < Note 107.3 by EISNER::KILLEEN "Jeff Killeen" > >> -< SET TERM/PERM/EIGHT KBxx: >- >> >> Are you sure the RSTS is set EIGHT_BIT? The norm in RSTS land is 7-bit. >> Yup! 8-bit no parity ... any trouble with that??? -Arnold ================================================================================ Note 107.5 LA-50 problems! 5 of 12 EISNER::KILLEEN "Jeff Killeen" 15 lines 8-MAR-1988 06:08 -< NOW THE FUN STARTS >- -------------------------------------------------------------------------------- > Yup! 8-bit no parity ... any trouble with that??? This has happened to us with LA210's. Since we can rule out the LA50 because it works with th PC it must be on the RSTS side. With one exception all the problems the LA210's have had was due to 7/8-bit, parity problems. The one exception was a DZ11 port that was throwing framing errors and that is what caused the printer errors. If I was trouble shooting this the next thing I would do is set the DZ to 7-bit and the LA50 to 7/space and try it. P.S. On the 210's there is something call restraint mode which sets printer busy on pin 11 - I don't know if the LA50 has this but if it does it most likely should not be set. ================================================================================ Note 107.6 LA-50 problems! 6 of 12 EISNER::ERSKINE "Chris Erskine" 14 lines 8-MAR-1988 06:42 -< Flow control and testing >- -------------------------------------------------------------------------------- As Jeff said there is a ready/busy mode. Be sure that it is set to Xon/Xoff. Swithc 1-6 should be open. I do not know if the Rainbow supports the hardware restaint and that may be the reason you are having the problem. Other tricks that can be done are: 1) Put a data scope on the line 2) Put a tube without a buffer on the line and by hand, send the Xon/Xoff's. 3) If you do not have a data scope, you can fake a simple data scope with 2 VT-220 terminals (1 on the send line and 1 on the rcv line). 4) Move the printer to the computer room and test it there. The LA-50 may have weak drivers on the transmit line. ================================================================================ Note 107.7 LA-50 problems! 7 of 12 EISNER::DELARISCH "Arnold S. De Larisch" 20 lines 7-APR-1988 14:52 -< What Next? >- -------------------------------------------------------------------------------- >> < Note 107.5 by EISNER::KILLEEN "Jeff Killeen" > >> -< NOW THE FUN STARTS >- >> If I was trouble shooting this the next thing I would do is set the DZ >> to 7-bit and the LA50 to 7/space and try it. Didn't work ... got a page full of backward questionmarks that ran 'screeming' down the page! >> P.S. On the 210's there is something call restraint mode which sets >> printer busy on pin 11 - I don't know if the LA50 has this but if it >> does it most likely should not be set. It is indeed NOT set and XON/XOFF Recognition is enabled on the printer! What next??? -Arnold ================================================================================ Note 107.8 LA-50 problems! 8 of 12 EISNER::HAHN 20 lines 7-APR-1988 16:17 -< change baud rate to 300 >- -------------------------------------------------------------------------------- 1) per previous notes the LA50 works with a PC.. I would set it for 7S (Sw2-, Sw2-5o, Sw2-6c) the correct baud rate (Sw2-1 t to Sw2-3) and first test it with a PC. 2) Maek sure Sw1-1 to Sw1-4 set to Open (US ASCII). 3) in the same manner the RSTS system be set at the same baud rate and TESTED with the PC (eg use KERMIT) or a terminal. 4) decide on whether you want to use the Xon/off protocol (Sw1-6o) or the ready/busy (Sw1-6c and pin 11) I presume that turning the printer on to the test mode (FF/power on) works. Connect only 2,3 and 7 for all the above tests unless you need 11 for ready/busy. Last question does the reverse ? occur immediately or after a while, and at what baud rate.. There may be a problem with the input buffer. Try to set it up at a slow baud rate e.g. 300 or 600. If this solves the problem there is a bad ship in your LA50 as there was in ours. Pierre ================================================================================ Note 107.9 LA-50 problems! 9 of 12 EISNER::DELARISCH "Arnold S. De Larisch" 21 lines 7-APR-1988 19:43 -< 300 Baud works ... but not above! >- -------------------------------------------------------------------------------- >> < Note 107.8 by EISNER::HAHN > >> -< change baud rate to 300 >- >> I presume that turning the printer on to the test mode (FF/power >> on) works. YES! >> Last question does the reverse ? occur immediately or after a while, >> and at what baud rate.. There may be a problem with the input buffer. >> Try to set it up at a slow baud rate e.g. 300 or 600. >> If this solves the problem there is a bad ship in your LA50 as there >> was in ours. It works fine at 300 Baud ... but anything above causes the data to get garbled. As the baud rate is increased less and less of the actual information is correctly printed. At 2400 Baud about a page an a half is printed befor the problem rears its ugly head! The printer is on the President Secretary's Desk ... and 300 Baud is beginning to frustrate her! -Arnold ================================================================================ Note 107.10 LA-50 problems! 10 of 12 EISNER::HAHN 8 lines 8-APR-1988 08:25 -< Tough it out at 300 baud! >- -------------------------------------------------------------------------------- You may not like the answer, but you have identified the problem. The x-on/off no longer works when the buffer is full. This is not an uncommon failure of the LA-50s that we have. Over the past 3 years 4 out of 11 have had that failure. When we are in dire financial status (being a State University this occurs towards the end of the fiscal year) we tough it out at 300 baud. Pierre ================================================================================ Note 107.11 LA-50 problems! 11 of 12 EISNER::PROVOST 7 lines 8-APR-1988 09:33 -< Known LA50 X-OFF problem >- -------------------------------------------------------------------------------- There is a known problem with LA50 DC-DC connector causing loss of X-OFF. Check for ripple on DC input. IF (RIPPLE) REPLACE CONNECTOR Replace capacitor in any case. Oil bearings with sewing machine oil Let me know if this helps... Tom Provost ================================================================================ Note 107.12 LA-50 problems! 12 of 12 EISNER::ERSKINE "Chris Erskine" 8 lines 13-APR-1988 06:37 -< New thoughts >- -------------------------------------------------------------------------------- Something just came to mind on this after rereading .0. Have you tried running the printer adjacent to the computer. The problem may be in either the cable to the PDP or the DZ port. Also for the 200 foot distance, the LA-50 may be having a problem driving it. You could also try placing a tube in place of the printer and seeing if ^S and ^Q works there. Chris  ================================================================================ Note 108.0 Q-BUS Board for TU-80? 13 replies EISNER::DELARISCH "Arnold S. De Larisch" 10 lines 9-MAR-1988 15:46 -------------------------------------------------------------------------------- Does anyone know if anybody offers a Q-bus board for DEC's TU-80 tape drive? I can't see trashing a perfectly good 1600 BPI drive off an 11/44 system which will be replaced with a uVAX 3600 just because DEC chooses to market only the TSV05! I don't like the idea of a Q-bus to Unibus converter either! (I'm trying to keep this beast under DEC Maintenance contract!) -Arnold ================================================================================ Note 108.1 Q-BUS Board for TU-80? 1 of 13 EISNER::KILLEEN "Jeff Killeen" 10 lines 9-MAR-1988 18:21 -< YOU ARE IN LUCK >- -------------------------------------------------------------------------------- > Does anyone know if anybody offers a Q-bus board for DEC's TU-80 tape drive? What I am about to say does *NOT* apply to the TU-81. I believe the interface on the TU80 is an un-modified CDC 9218x tape interface. Remember the TU80 is based on the CDC 92180 tape drive. That interface is an industry standard Pertec interface. An Emulex TC02 will do it - and so will a DEC TSV05 interface. Any Pertec tape interface should work. ================================================================================ Note 108.2 Q-BUS Board for TU-80? 2 of 13 EISNER::KENNEDY "Terry Kennedy" 25 lines 9-MAR-1988 19:11 -< Dilog is preferred solution here, other observations >- -------------------------------------------------------------------------------- > That interface is an industry standard Pertec interface. An Emulex > TC02 will do it - and so will a DEC TSV05 interface. I'd be careful with the TC02 - early revisions didn't work well - they thought they were TK25 emulation rather than anything useful. They may have fixed this to everybody's astisfaction, but read on... The Unibus TU80 interface card (M7454) was designed by Dilog. They also build the first run of 10,000 or so. They offer a Q-bus version of the board *which has different ROMS from their regular Q-bus card, the DQ132*. Therefore, they say that there is *something* different. As they designed both boards, I'd believe them. [Come to think of it, I *know* what the difference is - the regular boards (TSV05 included) think the drive is dual-speed. The TU80 has internal speed control. That's what confuses the vanilla cards.] Now, for the silly parts: 1) If you are running VMS, I know of no way to extract the TSV05 driver from DEC without buying a TSV05 to go with it. 2) Doesn't the 3600 uses the ruggedized cabinets? If so, you'll have to get the DEC handle conversion/perversion kit. 3) You will need to hack a cabinet kit together. You could use the old TU80 cabkit, shorten the internal cables, and punch two DB-50-sized cutouts in a blank filler panel. I have actually done this, so if you want yours chopped up, drop me MAIL. ================================================================================ Note 108.3 Q-BUS Board for TU-80? 3 of 13 EISNER::FRISBIE "Alan E. Frisbie" 9 lines 9-MAR-1988 21:34 -< Another Dilog fan >- -------------------------------------------------------------------------------- I, too, prefer the Dilog solution. I have two DQ-132 controllers with Cipher 910 drives that have given me no trouble at all. As for the driver, it is identical to the TSDRIVER.EXE on the full VMS kit. Since it will be distributed with all v5.0 kits, the problem of getting it goes away. Alan ================================================================================ Note 108.4 Q-BUS Board for TU-80? 4 of 13 EISNER::DELARISCH "Arnold S. De Larisch" 21 lines 9-MAR-1988 23:10 -< More Input! >- -------------------------------------------------------------------------------- >> < Note 108.3 by EISNER::FRISBIE "Alan E. Frisbie" > >> -< Another Dilog fan >- >> I, too, prefer the Dilog solution. I have two DQ-132 >> controllers with Cipher 910 drives that have given me no >> trouble at all. >> As for the driver, it is identical to the TSDRIVER.EXE on >> the full VMS kit. Since it will be distributed with all >> v5.0 kits, the problem of getting it goes away. As I understand it (correct me if I'm wrong) the uVAX 3600 is being shipped with FULL VMS not Micro VMS. Therefore, the problem with getting the driver is mute. As for the cabinet kit ... the TU80 is in a H9642 type cabinet with an RA-81 underneath. I'm gonna get the uVAX 3600 in the "computer room" cabinet (i.e. an H9642 style cabinet). Does anyone make a clean Cabinet Kit for doing this??? -Arnold ================================================================================ Note 108.5 Q-BUS Board for TU-80? 5 of 13 EISNER::KILLEEN "Jeff Killeen" 19 lines 9-MAR-1988 23:33 -< U-DO-IT CAB-KITS >- -------------------------------------------------------------------------------- > As for the cabinet kit ... the TU80 is in a H9642 type cabinet with > an RA-81 underneath. I'm gonna get the uVAX 3600 in the "computer room" > cabinet (i.e. an H9642 style cabinet). Does anyone make a clean Cabinet > Kit for doing this??? The Cab-Kit problem is on the uVAX 3x00 side - not the H9642. I don't think any of the third party controller manufacturers have built one yet. I would be even more surprised if they built one for the DEC DB50 cables. When they do build them I will bet they have ribbon cable headers. Also DEC is very careful to *ALWAYS* describe the module handle system as * P A T E N T E D * DEC also states if you don't use the module handle system you may have cooling problems. DEC will sell you a $81 kit with a blank strip. All you will need to do is punch two DB50 cutouts in it - then use your old kit. ================================================================================ Note 108.6 Q-BUS Board for TU-80? 6 of 13 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 8 lines 10-MAR-1988 09:50 -------------------------------------------------------------------------------- RE: < Note 108.4 by EISNER::DELARISCH "Arnold S. De Larisch" > -< More Input! >- | Therefore, the problem with | getting the driver is mute. If only more problems were mute, most of our discussions would be moot! ;-) ================================================================================ Note 108.7 Q-BUS Board for TU-80? 7 of 13 EISNER::HASSINGER "Bob Hassinger" 1 line 10-MAR-1988 10:30 -< Arnold, I am convinced 70% of the world does not understand .4 >- -------------------------------------------------------------------------------- ================================================================================ Note 108.8 Q-BUS Board for TU-80? 8 of 13 EISNER::DELARISCH "Arnold S. De Larisch" 16 lines 10-MAR-1988 17:05 -< Clear as Mud? >- -------------------------------------------------------------------------------- >> < Note 108.7 by EISNER::HASSINGER "Bob Hassinger" > >> -< Arnold, I am convinced 70% of the world does not understand .4 >- A small clearification may be in order. What I meant was that DEC had been purposely stripping certain code from the uVMS kit ... forceing you to buy only Digital! Case in point is the TSDRIVER.exe for the TSV05 tape drive. D$G$TAL would NOT sell you the license to get the driver UNLESS you provided them with the serial number for their Poor Excuse of a 1600 BPI tape drive ... the S-L-O-W TSV05! Full VMS comes with this driver stock ... hence no problem if the uVAX 3600 gets shipped with "full VMS". -Arnold P.S. Sorry for the confussion ... let me know if more clarification is needed! ================================================================================ Note 108.9 Q-BUS Board for TU-80? 9 of 13 EISNER::KENNEDY "Terry Kennedy" 8 lines 10-MAR-1988 21:18 -< Don't blame the hardware >- -------------------------------------------------------------------------------- > provided them with the serial number for their Poor Excuse of a 1600 BPI > tape drive ... the S-L-O-W TSV05! Full VMS comes with this driver stock ... Don't blame the poor Cipher transport (The TSV05 drive). The problem is that the VMS driver group doesn't want to properly support it! It requires that a bit be set in the command packet. If this is done, the drive runs (quite well, actually) at 100 IPS. By the way, RSTS/E on an 11/23 runs the drive at 100 IPS with no trouble... ================================================================================ Note 108.10 Q-BUS Board for TU-80? 10 of 13 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 31 lines 1-JUL-1988 00:06 -< Yes! Dilog, but not DQ132s >- -------------------------------------------------------------------------------- I am yet another Dilog fan, but rather than the DQ132, may I suggest one of their newer cards. The DQ142 is a better choice, and if you ever got a CDC 92185 you will be thankful. The DQ142 has higher thruput for GCR drives, and shortly should have new PROMS to make using the 92185 (what DEC turns into a TU81) more reasonable with BACKUP. This upgrade W O N ' T be possible on the DQ132 cards. I think the problem is the way read-backwards was implemented actually was backspace a record, read forward, backspace again. This is VERY slow on a streamer. Carl Freidberg spent a lot of time on this with Dilog. For the lower powered CDC drive in the TU80 having the DQ142 is probably just nice to have. DON'T read into this that TU81s will run on Pertec interfaces! The TU80 is vanilla and should work on various controllers, The TU81s are NOT just 92185s. DEC has done something to implement their LESI interface (Low End Storage Interface - look at a TU81's price and see if it seems LOW END!!) on the TU81. The BIG secret is that you probably could have used any old TU81 on a Q-bus system before the cached TU81-e came out. The only other product I know of on the LESI bus is the RC25 and if you plugged the TU81 into its Q-bus KLESI controller years ago, it should have worked (slowly). This does raise the question of whether a TU81 could be restored to a normal CDC 92185 state to use with anyones Pertec controller. One more NEW Dilog card that has been long long in coming and may actually be here is the DQ153. It has TMSCP (MU:) emulation. I would LOVE to be able to boot a 9 track tape on a microvax, but have heard ugly rumors that the boot code that works on TK50s may somehow fail on reel (real?) tape drives. ================================================================================ Note 108.11 Q-BUS Board for TU-80? 11 of 13 EISNER::WICHMAN "Jon Wichman" 6 lines 16-NOV-1988 15:13 -< DEC TU8x controller >- -------------------------------------------------------------------------------- >> adapter for TU80 on Q-BUS Might check with CSS re: the KLESI-xx. I ran across this before dealing with moving a TU81E. Not sure they use the same controller - anybody know off-hand? The TU81 version was definately availible from CSS as a special order item. ================================================================================ Note 108.12 Q-BUS Board for TU-80? 12 of 13 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 19 lines 16-NOV-1988 19:17 -< KLESI .not. for TU80 >- -------------------------------------------------------------------------------- >> adapter for TU80 on Q-BUS > Might check with CSS re: the KLESI-xx. I ran across this before > dealing with moving a TU81E. Not sure they use the same controller > - anybody know off-hand? KLESI is only for RC25 disks and TU81, NOT tu80, tapes. TU80 has the standard Pertec interface popular on MANY brands of tape drives, including the CDC 92185 (TU81) if bought from CDC. It is only when DEC buys a 92185 that it winds up being sold with a weird interface... IF someone NEEDS a Q Klesi controller, check a recent SWAP-MEET offering only listed as for RC25, but also ok for TU81s. Another trick: All the OTHER brand Pertec interface tape controllers seem to do up to 4 drives. Most 8mm drives are SCSI, but at least one is available in Pertec interface, so it could be daisy-chained off the same controller used for a 9 track. Saves $ and a Q slot. ================================================================================ Note 108.13 Q-BUS Board for TU-80? 13 of 13 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 21 lines 16-NOV-1988 20:48 -< Dilog Tape controllers for MV3s >- -------------------------------------------------------------------------------- >>I can't see trashing a perfectly good 1600 BPI drive off an 11/44 >>system which will be replaced with a uVAX 3600 just because DEC >>chooses to market only the TSV05! More data on DILOG Q tape controllers has surfaced. The DQ132 does have some sort of Q bus problem on MV3 systems that can't be ECO'd. Older DQ142 cards need new proms and actually some hard hardware eco and so must briefly visit the factory. Newer DQ142s and all DQ153s will have no problem. The newer DQ142s are fixed so CDC-92185 drives under BACKUP will use BACKSPACE rather than READBACKWARDS (which is emulated on this drive that can't do it by: BACKSPACE/READ-FORWARD/BACKSPACE!!) which was taking too long. The DQ153, which has the preferable MU: emulation, is bootable on MV2s if declaring itself to be a TK50, but not as a TU81. You have your choice. AND for BA2xx box users, the DQ153 has a SKUNKED DQ3153 version available.  ================================================================================ Note 109.0 uV3000 Machine Checks 2 replies EISNER::FRISBIE "Alan E. Frisbie" 16 lines 9-MAR-1988 21:42 -------------------------------------------------------------------------------- Does anyone have a copy of the KA650 User's Guide (EK-KA650-UG)? One is on order, but not expected for a few weeks. I need some information on machine checks. In particular, what kind of error is implied by the following register contents: DSER 00000004 QBEAR 0000000F DEAR 00000000 CBTCR C0000004 I think it is some kind of Q22 bus timeout, but I need more details. Alan ================================================================================ Note 109.1 uV3000 Machine Checks 1 of 2 EISNER::KILGALLEN "Larry Kilgallen" 21 lines 12-MAR-1988 00:14 -< I have EK-KA650-UG-001 >- -------------------------------------------------------------------------------- DSER <2> No grant timeout. Read/Write to clear. Sets if the Q22-bus does not return a bus grant within 10 ms of the bus request from a CPU demand read cycle or write cycle. Not set during interrupt acknowledge or request read cycles. Cleared by writing a 1, on power-up, by the negation of DCOK when the processor halts, and by writes to IPR 55 (IORESET). QBEAR <12:0> Contains Q22-bus address bits <21:9>. DEAR Reading this register when DSER <4> and DSER <0> are clear returns undefined data. CBTCR I could not find this one. ================================================================================ Note 109.2 uV3000 Machine Checks 2 of 2 EISNER::FRISBIE "Alan E. Frisbie" 7 lines 12-MAR-1988 00:35 -< Thanks! >- -------------------------------------------------------------------------------- >> -< I have EK-KA650-UG-001 >- Thanks, Larry. You just saved me many hours. Now it's time to get out the logic analyzer and figure out WHY it happens. Alan  ================================================================================ Note 110.0 Third Party Maintenance 9 replies EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sa" 11 lines 15-MAR-1988 05:54 -------------------------------------------------------------------------------- We are being forced to investigate third party maintenance for our VAX Cluster. So far we have been pretty pleased with Field Service (except for a couple of cases) from Digital. Now that we are going to place a request for proposal on the streets, I want to make sure whoever we get, be it Digital or some third party, is at least as good as Digital. Any suggestions for what to ask for ? Any things to watch out for in third party maintenance programs ? What does Digital Field Service have (other that a kit with at least one DOA board in it) that the third party won't have ? ================================================================================ Note 110.1 Third Party Maintenance 1 of 9 EISNER::KILLEEN "Jeff Killeen" 3 lines 15-MAR-1988 08:25 -< ECO/FCO'S >- -------------------------------------------------------------------------------- Make sure you require the vendor to install any mandatory ECO/FCO's *AND* any ECO/FCO's that are required to make the system software function. ================================================================================ Note 110.2 Third Party Maintenance 2 of 9 EISNER::HASSINGER "Bob Hassinger" 26 lines 15-MAR-1988 11:50 -< Very good point - Be careful about FUTURE ECO/FCOs >- -------------------------------------------------------------------------------- >Make sure you require the vendor to install any mandatory ECO/FCO's *AND* >any ECO/FCO's that are required to make the system software function. A very good point. When I looked into it I could not get the third parties to address the ECO/FCO issue. The case in point was that there was a MAJOR ECO/FCO for 11/750s some time ago. It involved replacing "half the CPU boards". We were on DEC maintenance so they just did it as part of the contract - no extra charge or hassle. The third parties were unable to give me an assurance that they could and would do the same. I had to take the risk of something like this coming along. Also, from time to time there have been questions raised about the ability of the third parties to actually get the parts and what-not for one of these. It has been suggested that if they can get them at all it might not be until after DEC had taken care of all it's own customers and it has also been suggested that we, the hardware owners, might have to be the ones to actually get the materials from DEC for the third party to install. In any case, I never got answers from the third parties that I was satisfied with. They kept seeming to answer different questions from the ones I was asking (i.e. either unable to understand the issue or unwilling to face it up front). You should at least keep this point in mind when talking with them. Bob H ================================================================================ Note 110.3 Third Party Maintenance 3 of 9 EISNER::KENNEDY "Terry Kennedy" 19 lines 15-MAR-1988 20:42 -------------------------------------------------------------------------------- ECO/FCO issue: My Sorbus service contract says: "Sorbus will control and instal all original equipment manufacturer hardware and firmware engineering changes, including microcode changes, categorized as Mandatory, Required, Specification, or Improvement on Equipment covered by this Agreement..." In fact, they installed several to get systems previously on DEC maintenance up to their minimum acceptable rev levels. Other issues: If you want to continue using DEC diagnostics on your VAX after leaving DEC field service, you will have to buy a license and media kit. Note that most third-party service vendors have their own, 'Brand-X' diagnostics, but the newest DEC hardware may not be supported yet. Also, you give up DEC remote diagnostics. ================================================================================ Note 110.4 Third Party Maintenance 4 of 9 EISNER::PROVOST 77 lines 22-MAR-1988 09:13 -< Hints & Kinks of 3rd party maintenance >- -------------------------------------------------------------------------------- We have been using third party maintenance for a number of years. Our current contractor is CDC. We had chosen McDonnell Douglas originally, but they withdrew their bid. Very disconcerting. Most vendors, including Digital, have a 90 day cancellation clause. This means you need to make you decision 3 months before the contract expires. Most vendors, including Digital, will not install all FCO's unless your system is having grief. A new contractor can charge you to bring your equipment up to rev at contract turnover, or can charge back updates after the contract is in force. It pays to have a pre-contract inspection by the incoming vendor. You should then get your current contractor to install all pertinent FCOs before your contract runs out. Third party vendors are more likely to maintain foreign equipment at reasonable prices. Unfortunately this work is sometimes subcontracted. It can be difficult to get the subcontractor to abide by the terms and conditions of the primary contractor. With smaller vendors, we find that the first man in the door tends to be more competent and experienced than the Digital equivalent. This is due to the lower number of competence levels. You pay the price at the other end, where Digital has a higher top-level. With some contractors there is an escape clause stating that if they can't fix your machine, they will give you your maintenance fees back. This is not comforting when faced with an HDA crash. The clause applies item by item. Installing FCO's is not the same as supplying FCO's. In some cases you are required to supply the FCO, and they will install it. This can be both costly and incovenient. Take the case of the DEQNA board-swap. Assume you are hardware-illiterate. You must have vendor de-install the board. You must bring or send the board to Digital for swap. You must have vendor re-install. Hopefully you remembered to have someone preserve your hardware Ethernet address. Meanwhile your machine is down. The cost of VAX diagnostics is enormous! One vendor I know of will absorb the cost of the diagnostics if their diagnostics don't resolve the problem. You end up the owner of the diagnostic license even if you later change vendors. Installation of new equipment is always a problem. If you buy new equipment from Digital, and have Digital install it, then you get an on-site warrantee. If your third party installs it, you get a return-to-factory warrantee. Couple this with lack of diagnostics for new devices and you have a problem. Even if you have Digital do the installation, you may encounter the finger-pointing problem. Some of these problems can be avoided by carefully phrasing your RFP. Savings on the contract costs can offset the others. You should know hat you're paying for. One more point... Once you've begun serious negotiations with third parties, Digital often has a change of heart about pricing. If you want to sole-source Digital, there are several unique features which can help you. As far as I know Digital is the only vendor to offer true remote diagnostic capability. Only Digital can supply the full range of diagnostics including newly announced products. If the vendor is required to include the cost of a diagnostic license in his bid, Digital begins to look good. Of course you should write off the cost of that license over several years of the contract. Doing business with third partes is much easier than with Digital. Third parties usually give you a single point of contact, whereas Digital likes to issue several contracts with different cost-centers depending upon the class and location of the equipment. On the other hand Digital always oils the squeaky wheel. If you complain loud and frequently, you will get a higher level of service. There seems to be no limit to the extent to which you can escalate a problem. Third parties are more likely to give a consistent level of service even if you don't complain. ================================================================================ Note 110.5 Third Party Maintenance 5 of 9 EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 29 lines 24-MAR-1988 13:04 -< We moved... >- -------------------------------------------------------------------------------- We have moved from DECservice to third party (CDC) service. We were the first commercial sight with 8650s in the Wash. DC area and this has caused some startup problems, like CDC not knowing what was in the local kits until it wasn't there. My impression from DEC is that if the new machine is moved into the area they get a kit. They have even delayed installation of some equipment (our LPS40, in particular) until they have the kit available. I would be careful to find out if you are the first or the upteenth customer with your equipment. I have found that the CDC central suport organization to be just as helpful as RDC, they have been more than happy to dial in and look over equipment and crash dumps on more than one occasion. The only problem is the lack of RDC or the like if the machine is down. I have pointed this out to them and they are looking into what they can do. They (and I am sure other vendors as well) counter by claiming to send a higher level rep to the sight with a more complete kit. This has seemed to be the case a couple of times, so far. The best thing that we did was to ensure that all our requirements were in the contract, ours reads something like "CDC will supply and instal ALL FCOs...". When you get to the bottom line be sure to get you negotiate a contract as favoriable to you as possible. This, I suppose, may not be an option for the small company. We negotiated a fairley specific contract with CDC, but DEC would do no such thing. Seton ================================================================================ Note 110.6 Third Party Maintenance 6 of 9 EISNER::KILLEEN "Jeff Killeen" 5 lines 24-MAR-1988 20:42 -< MANDATORY <> ALL >- -------------------------------------------------------------------------------- > The best thing that we did was to ensure that all our requirements > were in the contract, ours reads something like "CDC will supply > and instal ALL FCOs...". When you get to the bottom line be sure I hope it is not the standard CDC statement which is all *MANDATORY* FCO's. ================================================================================ Note 110.7 Third Party Maintenance 7 of 9 EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 12 lines 25-MAR-1988 21:36 -< We took exception to "MANDATORY" >- -------------------------------------------------------------------------------- We took exception to the "manditory" statement, and had them rewrite it "All". There is also a clause in the contract that states something along the lines of "if DEC does it, or provides it, with DECservice, CDC will do the same"... Also, back to how one pays for diagnostics... Our agreement with CDC states that we will purchase the licenses and initial media for DEC diagnostics, but that CDC will give us credit for the purchase. CDC will also have us purchase the updates, as needed, and give us credit, etc. Seton ================================================================================ Note 110.8 Third Party Maintenance 8 of 9 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 5 lines 7-JUN-1988 23:26 -< Include a escalation clause >- -------------------------------------------------------------------------------- We use DEC for our VAX systems, but have a third party for our IBM 4381. We put a clause in that contract that said if the third party cant fix the problem within some time period (24 hrs I think) they call in IBM to fix it and they foot the bill. I dont know if thats typical or not but it seemed like a good idea. ================================================================================ Note 110.9 Third Party Maintenance 9 of 9 EISNER::MERRIMAN "George Merriman - CCA/NY" 5 lines 2-OCT-1988 17:39 -< See SITE_MANAGEMENT for more information >- -------------------------------------------------------------------------------- See topic 24 in SITE_MANAGEMENT for a discussion of third party maintenance. It seems to me that SITE_MANAGEMENT would be a better place to discuss service contract issues in the future. George Merriman -- HARDWARE_HELP co-moderator  ================================================================================ Note 111.0 DEC RS232 VS DEC RS423 8 replies EISNER::KILLEEN "Jeff Killeen" 8 lines 17-MAR-1988 23:20 -------------------------------------------------------------------------------- Can someone explain what the difference is between DEC RS232 and DEC RS243? I thought the main difference was the distance and the connectors - however DEC also sells RS232 to RS243 *ACTIVE* converters to "....allow full implementation of DEC RS243...." Also does anyone but DEC sell the Modified Jacks with the DEC offset? ================================================================================ Note 111.1 DEC RS232 VS DEC RS423 1 of 8 EISNER::KENNEDY "Terry Kennedy" 38 lines 18-MAR-1988 02:36 -< A standard (sort of) >- -------------------------------------------------------------------------------- > Can someone explain what the difference is between DEC RS232 and DEC RS243? First, some tidying up on the nomenclature: DEC's previous interface standard was 'RS-232'. The new DEC standard is 'DEC423'. I have not seen any DEC literature claiming that DEC423=RS423. While they may be the same electrically, the funny connectors on DEC423 are a uniquely DEC perversion (although, as has happened in the past, other vendors may decide to follow along). The proper connector for RS-423 is specified as a 9-pin D subminiature. > I thought the main difference was the distance and the connectors - > however DEC also sells RS232 to RS243 *ACTIVE* converters to "....allow > full implementation of DEC RS243...." RS-232 is specified as 50 feet at 9600 baud when using a cable with [some] specified capacitance. RS-422 (note - not the same as 423) was developed to address the distance limitations of 232. It did this by using differential rather than polarity-based signalling. For example, on 232 a '1' is one polarity and a '0' the other. On 422, a '1' is one line being at a higher voltage than the other, and vice versa for a '0'. This allowed for substantially longer runs (on the order of thousands of feet). However, there was the usual hue and cry about compatibility - so RS-423 was developed. 423 is a single-ended version of 422, with the reference point being signal ground. Now, if you treat ground as one of the differential signals, you still have a differential input with 423. However, you *also* have something that is close enough to RS-232 that simple passive adapters can be constructed. > Also does anyone but DEC sell the Modified Jacks with the DEC offset? Yes - but getting the part numbers can be a bear. AMP Special In- dustries makes the DEC ones, and will probably sell them to you as well. Unfortunately, their latest full line catalog (2005-8) was issued in 1985 (it dwarfs the New York phone book). You could call your nearest distributor - CRONIN Electronics at (617) 449-0500. [For others reading this note, AMP's national number is (215) 647- 1000]. ================================================================================ Note 111.2 DEC RS232 VS DEC RS423 2 of 8 EISNER::ERSKINE "Chris Erskine" 3 lines 18-MAR-1988 06:41 -< Connector >- -------------------------------------------------------------------------------- I thought that 423 was an electrical spec only and it used RS-449 as the spec for the connector. This connector is a 39 pin sub-D connector. ================================================================================ Note 111.3 DEC RS232 VS DEC RS423 3 of 8 EISNER::KILLEEN "Jeff Killeen" 8 lines 18-MAR-1988 07:53 -< MORE 423 QUESTIONS >- -------------------------------------------------------------------------------- > However, you *also* have something that is close enough to > RS-232 that simple passive adapters can be constructed. Are the passive adapters just wires? - or do they have something else in them? Do you know what the assignment of the 6 wires are? ================================================================================ Note 111.4 DEC RS232 VS DEC RS423 4 of 8 EISNER::PRIGOT "Jonathan M. Prigot" 16 lines 18-MAR-1988 09:00 -< MMJ <==> RS232 >- -------------------------------------------------------------------------------- According to the VT320 manual, the signal assignments are: 1 - DTR 2 - TXD+ 3 - TXD- 4 - RXD- 5 - RXD+ 6 - DSR <-- this is the pin closest to the clip The - pins (TXD- and RXD-) correspond to signal grounds (RS232 pin 7). DTR is referenced to TXD-, while DSR is referenced to RXD-. The passive adapter is just wires, although they jumper RS232 pins 6 and 8 to MMJ pin 6. An excellent source of MMJ stuff is MOD-TAP Systems (617-456-3500 or 800-252-1100 [outside MA]). ================================================================================ Note 111.5 DEC RS232 VS DEC RS423 5 of 8 EISNER::ERSKINE "Chris Erskine" 5 lines 18-MAR-1988 11:55 -< It is not all as it seems. >- -------------------------------------------------------------------------------- Also you have to be careful of what you use to make the link between the port and terminal. The MMJ block which DEC uses for the DECConnet system of wiring connects only 2, 3, 4, 5. The block has 8 pins, the jack has 6 pins and the circuit board connecting them only connects 4 four of the pins. ================================================================================ Note 111.6 DEC RS232 VS DEC RS423 6 of 8 EISNER::HANRAHAN "Jamie Hanrahan" 55 lines 19-MAR-1988 15:02 -< Differential details >- -------------------------------------------------------------------------------- >It [RS422] did this [allowed use with longer lines at higher speeds] by >using differential rather than polarity-based signalling. For example, >on 232 a '1' is one polarity and a '0' the other. On 422, a '1' is one >line being at a higher voltage than the other, and vice versa for a '0'. A slight clarification. The voltages for RS422 driver outputs are spec'd at 0 and 5 volts, plus or minus somethingorother. The point of differential signalling is that you have two wires for each signal, plus a common ground, rather than a bunch of individiual signal wires and a common ground, as is used in RS232 and 423. The two wires are driven with opposite signals -- that is, a "1" is represented by +5 on one wire and 0 on the other, and a "0" is represented by the opposite. The "inverted" signal is generated by a simple operational amplifier (not a digital inverter, at least not directly, as we still have to be concerned with slew rate limiting, etc.). At the receiving end the two wires go to the inverting and noninverting inputs of an opamp, which inverts one input and then adds them together. Thus, the output of the receiving opamp is the *difference* of the signals on the two wires. Here's the neat part: Any noise picked up by the wires tends to be picked up more or less equally by *both* signal wires. When the receiving opamp inverts one side of the signal and adds it to the other, the noise is cancelled! This is called "common mode noise rejection", and its efficacy is expressed as a ratio, expressed in dB. The common mode rejection ratio (CMMR) of a simple differential hookup as described can easily be 60 dB or better. Of course you're sending the inverted signal through an extra opamp stage at the driver end... but if you use similar-behaving opamps for the line drivers and receivers, many of the nonlinearities in the opamps get cancelled too. Cute, eh? If your cable is going through very-high-noise environments, you can use a shielded, twisted pair for your two signal wires. With proper grounding techniques and high-quality cable this reduces the noise pickup of the cable by around 71 dB, according to the textbook. Your mileage may vary. (The same trick is used in professional audio gear; this is one of the many things that separate pro gear from home stuff -- they use differential inputs and shielded, twisted-pair cable, though the outputs (except from mics) are single-ended; note that CM noise rejection works whether the extra wire carries an inverse of the signal wire or a 0-voltage reference. This is why you can run a mic cable for hundreds of feet across a stage and get negligible AC hum, while a six-foot run from your preamp to your power amp in your living room may do much worse.) I believe that RS422's 0 and +5 voltages (as opposed to RS232's +/- 3- to 25 volts, or whatever it is; in practice, many mfrs use +/- 12, and some use +5, -12) were chosen simply to eliminate the need for extra power supplies, for which most digital devices had no other use. The availability of RS232 driver chips from Maxim (and others) which have on-board DC-to-DC converters, operating from a single +5 supply and costing little more than old-style RS232 drivers, has reduced the benefit of this somewhat. ================================================================================ Note 111.7 DEC RS232 VS DEC RS423 7 of 8 EISNER::WICHMAN "Jon Wichman" 14 lines 16-NOV-1988 15:08 -< DEC423 pin-outs & specs. >- -------------------------------------------------------------------------------- >>>> DEC423 information See DEC publications EK-DECSY-CG-001 (DECconnect System Planning and Configuration Guide) and EK-DECSY-GD-001 (DECconnect System General Description) for DEC423 pin outs and adapter info (esp. the Configuration guide). These two are usually shrink-wrapped together, but I don't know the part number for the combination (my DEC rep gave it to me). The latest version (as of about March 1988) was dated April 1986. It has the cable and H8571-A and -B adapters in it. For the newer adapters, check out the DEC store ((800) 332-3366) under "CABLES" and also Insight magazine (option 10 on the main menu) around March 1988 edition. They are up to "F" on the H8571's last I knew... ================================================================================ Note 111.8 DEC RS232 VS DEC RS423 8 of 8 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 7 lines 16-NOV-1988 19:04 -< newer DecConnect manuals exist >- -------------------------------------------------------------------------------- >> See DEC publications EK-DECSY-CG-001 (DECconnect System Planning >> and Configuration Guide) and EK-DECSY-GD-001 (DECconnect System >> General Description) ... The latest version (as of about March 1988 >> was dated April 1986. These are now both ...002, and both are dated AUG 1988 and have lots of new stuff. See 171.14 here in HARDWARE.  ================================================================================ Note 112.0 LN03 Toner Options 1 reply EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sap" 9 lines 25-MAR-1988 12:04 -------------------------------------------------------------------------------- I went ahead and bought the little kit from Digital to replace my toner cartridge in my LN03. It was fairly expensive and I wondered can you buy the actual toner and refill these things yourself ? What commercial brand of toner do you use ? Also has anyone used those companies that say they refill your cartridge for you and what has your experiences been ? ================================================================================ Note 112.1 LN03 Toner Options 1 of 1 EISNER::LEDERMAN "Bart Z. Lederman" 11 lines 26-MAR-1988 18:32 -< Could this be false economy? >- -------------------------------------------------------------------------------- The toner kits I've received have some sort of environmental notice in them saying what the toner contains (basically plastic and a pigment made of iron oxide or "rust"). The LN03 uses the Ricoh "guts" common to several other laser printers, so there may be other sources of toner. But considering how expensive a laser printer is, I'd be very cautious about trying to save a few dollars on toner that might wreak the printer. On the other hand, it'd be nice to know if any of the parts that come out of the printer when doing maintenence (like the image band and such) had any trade in value for rebuilds.  ================================================================================ Note 113.0 2nd DW750 installtion woes 3 replies EISNER::RENES "Stuart Renes, DFWLug Chair" 11 lines 7-APR-1988 11:38 -------------------------------------------------------------------------------- Has anyone experienced problems (and found a solution) with adding a second DW750 (Unibus Adapter) to an 11/750? One user here has tried unsuccessfully to do that, even after following the installation manual, and verifying that all the hardware passes DEC diagnostics. The symptom is that the heartbeat light goes out as soon as the UBA ribbon cables are attached from the DW750 to a CMI slot. Thanks in advance. ================================================================================ Note 113.1 2nd DW750 installtion woes 1 of 3 EISNER::DOHERTY "Bob Doherty" 16 lines 7-APR-1988 22:57 -< We have one, and it works. >- -------------------------------------------------------------------------------- We had a second DW750 put on our 750, and had a great deal of problems until we realized that the clown who installed our Massbus controller had installed it at the same CMI arbitration level as the second Unibus. Hence, the network traffic, which we have on the second unibus, was getting mixed up with disk I/O on the Massbus. Once we got that straightened out, we've had not trouble at all. The second Unibus has a v.35 DMR11 and a DELUA on it, while all the other peripherals are on the first Unibus. The tech who installed the second DW750, as I remember it had a great deal of trouble getting the cables from the DW750 to the first DD11 inserted correctly. However, until he did, as I remember it, the second Unibus wasn't recognized at all. Bob. ================================================================================ Note 113.2 2nd DW750 installtion woes 2 of 3 EISNER::TINKELMAN "Bob Tinkelman - CCA" 7 lines 8-APR-1988 06:22 -< Second UBA working here too >- -------------------------------------------------------------------------------- We had a second UBA installed on a client's 750 when it needed a second UDA50. It's working fine. One or two of the other peripherals were moved over to it also. Although it was installed fairly recently (January), I don't remember the details of the problems that DEC Field Service had before getting it working smoothly. I believe that the problems were simple things - like setting CSRs incorrectly for the options moved from one UBA to another. ================================================================================ Note 113.3 2nd DW750 installtion woes 3 of 3 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 5 lines 1-JUL-1988 00:29 -< did you need the 2nd Unibus? >- -------------------------------------------------------------------------------- No first hand experience in this but I have heard that some situations where a 2nd Unibus adapter would normally be needed can easily be handled by using Setasi's Unibus high speed segmenter/isolator and keeping just one Unibus. This unit is apparently also used by DEC field service to trouble shoot Unibus problems!  ================================================================================ Note 114.0 TK25 No replies EISNER::KILLEEN "Jeff Killeen" 3 lines 8-APR-1988 00:14 -------------------------------------------------------------------------------- From what I understand Dilog produced the TK25 controller for the TQK25 tape drive. Does anyone know what the number of the "real" Dilog controller is and does anyone know if there is a UNIBUS version also?  ================================================================================ Note 115.0 CSU/DSU DEVICES ON DDS TERMINATIONS 5 replies EISNER::KILLEEN "Jeff Killeen" 6 lines 9-APR-1988 00:03 -------------------------------------------------------------------------------- Has anyone used this type of phone service? 1. What are the advantages/disadvantages over modems 2. What type of lines do you order from your local TELCO ================================================================================ Note 115.1 CSU/DSU DEVICES ON DDS TERMINATIONS 1 of 5 EISNER::TINKELMAN "Bob Tinkelman - CCA" 41 lines 9-APR-1988 23:32 -< A little about digital vs analog >- -------------------------------------------------------------------------------- Digital Data Sevice provides a dedicated digital circuit at one of the standard speeds - 2400, 4800, 9600, 19200 (in some few locations), 56000 and (in Europe) 64000 bps. You use a digital circuit in much the same way you use a leased voice circuit (what used to be called a 3002 circuit). Instead of leased line modems, you use DSU/CSUs. A DSU/CSU is typically a box with one connector for the telephone company's two pair and one RS232 or V.35 connector to go to your equipment. (Historically, there were once two separate units - CSUs which the phone company provided and DSUs which you could provide yourself.) Advantages of Digital service: o Quality of service. It is my experience that phone companies fix problems with the digital circuits more quickly than with analog circuits. Also problems occur much less frequently. o Cost of equipment. DSU/CSUs are generally much cheaper (approx $500 each) than leased line modems ($1000 to $4000). Disadvantages of Digital service: o Cost of service. In every case I've investigated, 9600bps digital service was more expensive than an analog circuit between the same two points. Sometimes the difference was slight. Sometimes it was substantial. o Installation delay. This may be changing. But the facilities needed to support digital circuits seem to be in short supply, resulting in longer lead times for installations. Six months ago, I was quoted 30 working days for an analog circuit and 60 working days for a digital circuit between the same two points. o Limited choice of speeds. With an analog cicuit, you can buy modems for speeds like 14.4K and 19.2K. The first is simply not available as a digital service and the second is available in only certain locations. o Difficulty of "dial backup". A leased analog circuit consists of two pairs - a transmit pair and a receive pair. If the leased circuit is bad, you can use the same modems, substituting two dial-up circuits. There are various "dial backup" units sold. Some modems have this capability built in. With digital circuits, there is no corresponding fallback. ================================================================================ Note 115.2 CSU/DSU DEVICES ON DDS TERMINATIONS 2 of 5 EISNER::KENNEDY "Terry Kennedy" 14 lines 10-APR-1988 04:26 -------------------------------------------------------------------------------- Also a few more things to consider: o Be especially careful if you are pricing multi-point drops (such as DMC/DMR tributaries). You may get 2 or 3 different rates when quoted for digital service. o With digital service, you are limited to a set of synchronous proto- cols (I'm unsure if DDCMP qualifies directly). Some inexpensive stat muxes have a 9600 *async* composite feed, which won't play. o Since it's digital, service is all-or-none, with no fallback to a slower speed with line degradation. Of course, analog circuits go completely dead more often than fading, but in fading cases your analog 9600 modem can fall back to 7200 and then to 4800. ================================================================================ Note 115.3 CSU/DSU DEVICES ON DDS TERMINATIONS 3 of 5 EISNER::TINKELMAN "Bob Tinkelman - CCA" 6 lines 10-APR-1988 07:33 -< DDCMP runs on DSU/CSUs >- -------------------------------------------------------------------------------- Re: < Note 115.2 by EISNER::KENNEDY "Terry Kennedy" > > o With digital service, you are limited to a set of synchronous proto- > cols (I'm unsure if DDCMP qualifies directly). The service provided moves bits and should support *all* synchronous protocols. ================================================================================ Note 115.4 CSU/DSU DEVICES ON DDS TERMINATIONS 4 of 5 EISNER::KENNEDY "Terry Kennedy" 5 lines 10-APR-1988 21:06 -< Inquiring minds at the Telco want to know... >- -------------------------------------------------------------------------------- > The service provided moves bits and should support *all* synchronous protocols. When we investigated this service, we had to specify the protocol we would be using. Maybe this was just a 'survey' - but it was not pre- sented that way... ================================================================================ Note 115.5 CSU/DSU DEVICES ON DDS TERMINATIONS 5 of 5 EISNER::ERSKINE "Chris Erskine" 3 lines 13-APR-1988 06:42 -< No problems with digital lines >- -------------------------------------------------------------------------------- We push DDCMP over digital lines all of the time. The service is also more reliable and for an idea of where things are going, TELCO is changing their system to digital including the telephones.  ================================================================================ Note 116.0 9600 BAUD DIALUP MODEMS 32 replies EISNER::KILLEEN "Jeff Killeen" 11 lines 9-APR-1988 00:33 -------------------------------------------------------------------------------- There are two types of dialup 9600 baud modems available now V.29 EDI V.32 What are the advantages/disadvantages of the two standards. The V.29 modems are about 2/3 the price of the V.32 modems. Is this price difference because the V.32 or better or because they are newer? ================================================================================ Note 116.1 9600 BAUD DIALUP MODEMS 1 of 32 EISNER::KENNEDY "Terry Kennedy" 39 lines 9-APR-1988 05:41 -< Lots of incompatibilities >- -------------------------------------------------------------------------------- > What are the advantages/disadvantages of the two standards. The V.29 > modems are about 2/3 the price of the V.32 modems. Is this price > difference because the V.32 or better or because they are newer? First, almost none of these modems will talk to any other 9600 baud modem except another unit of the same brand. The V.29 prototcol is a half-duplex protocol, initially used on digital fax machines. Therefore, you have the line turn-around time for things like character echo. Turn-around time varies drastically between the various brands, and is one of the major incompatibilities. Some of the modems also use a slow (~300 baud) reverse channel to avoid the turn- around delay. Also, some units implement higher levels of the MNP proto- col, so that things like Kermit acknowledgements are generated internally by the modem, rather than having to turn the channel around. V.32 is a full-duples 9600 baud protocol, but unfortunately it's been bastardized right out of the starting gate. It requires very complicated (hence expensive) echo-cancellation circuitry in order to avoid hearing itself on the line. Some manufacturers have adopted half- duplex implementations to avoid the cost of the circuitry. Of course, that means that the units aren't strictly V.32 anymore, but that won't stop the misleading advertisements. A third protocol is the proprietary DCA/Telebit/Ven-Tel protocol, which uses a digital signal processor (DSP) to band-split the phone channel into about 500 carriers. These units adjust to phone line quality in steps of about 75 baud. The above brings up an important point - most of these units have internal buffers, and deliver data to the attached computer or ter- minal at a fixed rate. This may well render them useless for appli- cations like Async DECnet, which gets scared if there is any intel- ligence in the link it doesn't know about. You may want to refer to PC magazine, Vol. 7, No. 8, April 26, 1988. It has a review of a dozen or so 9600 baud modems. ================================================================================ Note 116.2 9600 BAUD DIALUP MODEMS 2 of 32 EISNER::KILLEEN "Jeff Killeen" 12 lines 9-APR-1988 11:53 -< SOUNDS LIKE THE EARLY DAYS OF 1200 BAUD MODEMS >- -------------------------------------------------------------------------------- > V.32 is a full-duples 9600 baud protocol, but unfortunately it's > been bastardized right out of the starting gate. It requires very > complicated (hence expensive) echo-cancellation circuitry in order to > avoid hearing itself on the line. Some manufacturers have adopted half- > duplex implementations to avoid the cost of the circuitry. Of course, > that means that the units aren't strictly V.32 anymore, but that won't >stop the misleading advertisements. Sounds like these are the ones to go with provided they have fully complied with the standard. ================================================================================ Note 116.3 9600 BAUD DIALUP MODEMS 3 of 32 EISNER::KILGALLEN "Larry Kilgallen" 2 lines 9-APR-1988 20:46 -< What about Codex and Concord V.32 >- -------------------------------------------------------------------------------- I was under the impression that they were both fully compliant and therefore should talk to each other. ================================================================================ Note 116.4 9600 BAUD DIALUP MODEMS 4 of 32 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 13 lines 16-AUG-1989 12:11 -< Two (Well, 2.5) questions... >- -------------------------------------------------------------------------------- I have two questions about V.32: 1. Is V.32 here to stay? I mean, can I expect Tymnet, Telenet, DECUServe, and anybody else I dial up to end up getting V.32 eventually? 2. I'm looking at Telebit's modems, in particular the Trailblazer Plus, and the T2500 (Supposedly a T-blazer+ with V.32). Assuming the answer to #1 is positive, is the T2500 a "good" implementation of V.32, and is it in the ballpark price-wise? Ed ================================================================================ Note 116.5 9600 BAUD DIALUP MODEMS 5 of 32 EISNER::LEDERMAN "Bart Z. Lederman" 23 lines 16-AUG-1989 13:30 -< Probably fairly stable, but hard to predict >- -------------------------------------------------------------------------------- V.32 is a reasonably stable CCITT recommendation. Since these recommendations have no expiration date, you can expect V.32 to last quite a while. While other recommendations get modified occasionally, past history shows relatively little change to modem recommendations, and they would almost certainly be compatible with previous versions. It is also almost certain that there will be faster dial-up modems some day, and though I'd hesitate to make any really hard predictions, 9600 Baud is hitting a reasonable speed limit for dial-up operation. Also, a lot of computer equipment doesn't work well over 9600 Baud, so there is less incentive to make faster modems than this (for this type of application, anyway. There are 14,400 Baud FAX modems.) V.32, in my limited experience, is far more popular than any of the proprietary 9600 Baud modems, and since it has international acceptance, I'd say you are more likely to encounter a V.32 9600 Baud modem than other 9600 Baud modems. But I can't really say if any of the commercial services will offer this. If enough customers have them and ask for that service, the various companies (Tymnet, Telenet, etc.) will comply: at least, if they have any business sense they will, and it's a very competitive industry. ================================================================================ Note 116.6 9600 BAUD DIALUP MODEMS 6 of 32 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 11 lines 16-AUG-1989 18:39 -< Telenet has v.32 >- -------------------------------------------------------------------------------- I have been told that Telenet HAS v.32 modems deployed, and will have more as time goes on. I was also told that PC Pursuit may not be available on them. Rockwell makes the chips that many v.32 modems use, and there is supposed to be some newer version that will have a raw speed of 12000, rather than the current 9.6k limit. To that add the compression and other games, and you get some fancy speeds. I haven't seen any modem companies talking about 12k yet, nor have I seen the V.32 spec, so don't know how it actually fits in. ================================================================================ Note 116.7 9600 BAUD DIALUP MODEMS 7 of 32 EISNER::BURNET "John Burnet" 16 lines 16-AUG-1989 21:35 -< Yes -- V.32 is a winner. All systems go. >- -------------------------------------------------------------------------------- > V.32, in my limited experience, is far more popular than any of > the proprietary 9600 Baud modems, and since it has international > acceptance, I'd say you are more likely to encounter a V.32 9600 > Baud modem than other 9600 Baud modems. Agreed. And in addition to the CCITT seal of approval, one big advantage of V.32 is that it's a standard for *symmetrical, full-duplex* I/O -- none of this "300 baud one way, 9600 baud the other way, nn milliseconds to turn the line around" nonsense that most other 9600-baud transmission schemes rely on. In an article in the current issue of *Data Communications*, a Multitech higher-up is quoted as saying that the market for non-V.32 9600-baud modems is drying up quickly, now that V.32 is here. They expect that their model 696E (non-V.32, asymmetrical 9600 bps) won't last too much longer, while sales of their V.32 product are increasing rapidly. ================================================================================ Note 116.8 9600 BAUD DIALUP MODEMS 8 of 32 EISNER::BURNET "John Burnet" 2 lines 16-AUG-1989 21:37 -< ... and prices are dropping, as expected >- -------------------------------------------------------------------------------- Quantity-1, end-user (list) prices for V.32 modems are down to $800-1000 per unit now. ================================================================================ Note 116.9 9600 BAUD DIALUP MODEMS 9 of 32 EISNER::DOHERTY "Bob Doherty" 10 lines 16-AUG-1989 23:04 -< When V.32 <> V.32 ... >- -------------------------------------------------------------------------------- One thing to watch for is vendors playing fast and loose with the language. For example, Hayes was advertising a 9600 baud modem as "half-duplex V.32". There ain't no such thing, and this puppy was actually doing a "ping-pong" protocol so that the only modems these things would talk to as "V.32" were each other. We picked up a number of Courier HST Dual-Standard modems for just under $1000 each (GSA price) about three months ago, and I expect the price to go lower than that, maybe to the $5-600 range in the next year or so. ================================================================================ Note 116.10 9600 BAUD DIALUP MODEMS 10 of 32 EISNER::HAHN "DECUServe Vice Chair" 3 lines 17-AUG-1989 06:43 -< Vendor and Model numbers requested >- -------------------------------------------------------------------------------- We are interested in acquiring V.32 modems in the near future, could you give vendor name and model number, these would be modems you would be willing to use in your company. ================================================================================ Note 116.11 9600 BAUD DIALUP MODEMS 11 of 32 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 33 lines 17-AUG-1989 11:13 -< Take a look at Telebit... >- -------------------------------------------------------------------------------- Wow! Thanks for all the good info. Sounds like V.32 is the way to go, though that 12K bps extension doesn't suck pond water, either... ;-) >< Note 116.10 by EISNER::HAHN "DECUServe Vice Chair" > > -< Vendor and Model numbers requested >- > > We are interested in acquiring V.32 modems in the near future, could > you give vendor name and model number, these would be modems you > would be willing to use in your company. Well, I'm waiting for some spec sheets on their products, but give some thought to Telebit. Their Trailblazer Plus is something of a standard in U**X-land, where they routinely push 14Kbps through them with uucp. The high-speed mode on this unit is not based on any national or international standard, but if DECUS uucp is going to be running on this system, (as it appears to be running now) it might behoove the powers that be to consider at least some of these. Telebit also has a V.32 modem, which is supposedly a T-blazer+ with a V.32 mode added in. This could handle uucp traffic for those folks that "grep their cat" (or use DECUS uucp), as well as provide a universally recognized standard means of 9.6Kbps dialup access. One further point. Telebit has a pretty attractive discount agreement in place for subscribers to UUNET, a uucp dialup service. Perhaps something similar could be set up for DECUServe... Ed (I should mention that I have zero experience with Telebit products, but all the talk I've heard about them is positive. If somebody that has laid hands on their products can provide additional info (good or bad), I'd appreciate it; I'm interested in a good fast modem...) ================================================================================ Note 116.12 9600 BAUD DIALUP MODEMS 12 of 32 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 11 lines 17-AUG-1989 11:48 -< Looking at the front-runner >- -------------------------------------------------------------------------------- > One further point. Telebit has a pretty attractive discount > agreement in place for subscribers to UUNET, a uucp dialup > service. Perhaps something similar could be set up for > DECUServe... Could you give us some more information on this, or point towards a source? I can only afford to buy one high speed modem, and I want to make sure I buy the "correct" one. As usual, I will be watching this topic for everyones opinions. ================================================================================ Note 116.13 9600 BAUD DIALUP MODEMS 13 of 32 EISNER::KENNEDY "Terry Kennedy" 9 lines 17-AUG-1989 12:01 -< US Robotics Dual Standard here >- -------------------------------------------------------------------------------- Well, we just bought a couple rackfuls (15 modems/rack) of US Robotics Dual Standard modems. These do both V.32 and USR's proprietary 9600 mod- ulation scheme. We would probably get just the V.32 version for future acquisitions, but we have a large installed base of HST 9600 modems to support, so we went with the Dual Standard units. US Robotics is at 800-DIAL USR. ================================================================================ Note 116.14 9600 BAUD DIALUP MODEMS 14 of 32 EISNER::HARVEY "Jack Harvey" 8 lines 17-AUG-1989 12:17 -< TrailBlazer Distributor >- -------------------------------------------------------------------------------- One source is Glasgal Communications, Inc. They are a stocking distributor of a wide variety of modems and related data com products. Home office 201 768-8082 (NJ). Many branches: 714 380-0161 (S.CA), 312 705-3830 (IL), etc. Ask for their big catalog (1/2 inch thick). They offer engineering help too. I have no connection with Glasgal. ================================================================================ Note 116.15 9600 BAUD DIALUP MODEMS 15 of 32 EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 13 lines 17-AUG-1989 16:36 -< Some more V.32 >- -------------------------------------------------------------------------------- Microcom is also introducing a V.32 modem in their AX line of products. It also supports MNP class 9, I believe. Don't ask what class 9 does; I don't know. Anderson Jacobson also makes V.32 compliant modems, model AJ 9631-S for synchronous; AJ 9631-SA for synch/asynch. Concord Data Systems makes the DialAccess 296 and the V.32 Trellis. Penril makes the Alliance ALX V.32 These all came from the ARC catalog. ARC is at 1-800-321-2510 or, in Ohio, 1-800-686-0066. I have no association... ================================================================================ Note 116.16 9600 BAUD DIALUP MODEMS 16 of 32 EISNER::BURNET "John Burnet" 14 lines 17-AUG-1989 21:37 -< Multitech >- -------------------------------------------------------------------------------- We don't have any V.32 modems yet, but we've used Multitech 2400E modems for a while and we like them, so we'll probably consider the Multimodem V.32 when we go shopping for V.32 modems. The list price of that unit as of a couple of months ago was $1200 (quantity 1). If that hasn't already come down by one or two hundred dollars, it probably will before too much longer, since many other vendors' prices are under $1k. Multitechs are fairly rugged, they're Hayes-compatible to the nth degree, and they also have lots of extra features in the command set. The 2400E model implements MNP class 5; I'm not sure about the V.32 model, but I assume that they'd keep MNP (you can always disable it). The V.32 can also be used at 2400, 1200, and 300 bps. I am not affiliated with Multitech. ================================================================================ Note 116.17 9600 BAUD DIALUP MODEMS 17 of 32 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 9 lines 18-AUG-1989 10:16 -< TYMNET has V.32 >- -------------------------------------------------------------------------------- TYMNET also supports V.32 9600 bps in a few large cities (like Chicago 312-427-7579). We have recently purchased a few Codex 2264 V.32 modems and are very happy with them. They have MNP error correction (level ?) and also have a V.22 2400 bps modem built in for "backward compatibility". It also speaks the "AT" command set. It's great to read the E-Store at 9.6 k !! ================================================================================ Note 116.18 9600 BAUD DIALUP MODEMS 18 of 32 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 39 lines 18-AUG-1989 10:28 -< More info on the UUNET deal... >- -------------------------------------------------------------------------------- I should preface this by saying that this deal is only for subscribers to UUNET. The reason I mentioned it here was to (hopefully) get some of the DECUServe powers-that-be thinking about trying to set up something similar for DECUServe subscribers. After all, if Telebit is discounting equipment for one group of users, maybe they'll do it for another, right? Here's the deal for UUNET subscribers: Trailblazer Plus modems for half-price (Can purchase no more than two at this price). This is a non-V.32 modem, but with a high-speed proprietary protocol that uucp-types have adopted as a sort of standard. UUNET subscribers can also purchase (with no quantity limitations) the following modems at a 33% discount: The T2500, which is a Trailblazer Plus that also supports V.32. The Trailblazer Plus. The T1000, Which is a slowed-down version of the T-blazer+ (only 9600bps). The wording on this information sheet is interesting. It starts off by saying that only UUNET sites that have not purchased modems under one of Telebit's previous promotional offers can get two T-blazer+'s at 50% off. It then goes on to say that the 33% discounts are offered only to UUNET subscribers. Seems to imply that (1.) Telebit has been (or is) doing this sort of thing with other groups of users, and (2.) UUNET, perhaps due to the number of users involved, was able to negotiate the 1/3 off deal. All this sounds like Telebit might be willing to cut a similar deal for DECUServe (or maybe DECUS as a whole). To you folks running DECUServe: The sheet also lists the address and phone number of the Telebit people that are coordinating this offer for UUNET subscribers. If you don't have this information already, and feel a discount for DECUServe subscribers might be worth looking into, I'd be happy to pass it along. Send me mail if you need it... Ed ================================================================================ Note 116.19 9600 BAUD DIALUP MODEMS 19 of 32 EISNER::DELARISCH "Arnold S. De Larisch" 33 lines 19-AUG-1989 00:59 -< Trailblazers are GREAT! >- -------------------------------------------------------------------------------- >> < Note 116.12 by EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" > >> -< Looking at the front-runner >- >> > One further point. Telebit has a pretty attractive discount >> > agreement in place for subscribers to UUNET, a uucp dialup >> > service. Perhaps something similar could be set up for >> > DECUServe... >> Could you give us some more information on this, or point >> towards a source? If you run a public BBS or you participate in UUNET you can purchase a Trailblazer modem at half price. >> I can only afford to buy one high speed modem, and I want >> to make sure I buy the "correct" one. As usual, I will >> be watching this topic for everyones opinions. The trailblazer+ (I think its called) give you both V.32 and the Trailblazer high speed 14KBps. The techology used in the Modem is very interesting. Instead of using a single carrier, it determines how many different carriers can be used (with appropriate signal processing) and uses as many good channels as possible. Neet concept. It also has advantages in VERY bad connections. On most 'standard' modems, if the particular carrier isn't getting thru the telephone line, you are hosed. With the trailblazer adaptive technology, it simply discards that channel and uses others. I herd a TECHNICAL talk by an Engineer from Trailblazer, I was impressed! Check out the TrailBlazer and TrailBlazer+, I think they are the way to go! -Arnold ================================================================================ Note 116.20 9600 BAUD DIALUP MODEMS 20 of 32 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 24 lines 19-AUG-1989 16:02 -< Clarification on TB+ vs. T2500 >- -------------------------------------------------------------------------------- >The trailblazer+ (I think its called) give you both V.32 and the >Trailblazer high speed 14KBps. No, the TB+ "only" runs 300/1200/2400 and Telebit's proprietary PEP code (and, as I've said elsewhere, we really do get 1400 char/sec through it in uucp-mode). PEP mode is the adaptive multicarrier mode described by Arnold. >. With the trailblazer adaptive technology, it >simply discards that channel and uses others. And it is adaptive DURING a connection -- as the channel characteristics change, the modems adapt! The Telebit T2500 is their 300/1200/2400/PEP/V.32 version. Incidently, I believe that Telebit is selling a ROM upgrade for Trailblazer Plus's to turn them into T2500's for $500 or so. (Wow! A sanely-priced upgrade option! Telebit is obviously interested in maximizing short-term customer satisfaction, which should lead to long-term income advantages, rather than grabbing for the bucks in the short term... are you listening, DEC?) And with the uunet discount on the TrailBlazer Plus, this could get you a T2500 for a very reasonable price. ================================================================================ Note 116.21 9600 BAUD DIALUP MODEMS 21 of 32 EISNER::BURNET "John Burnet" 3 lines 19-AUG-1989 16:38 -< More on Multitech >- -------------------------------------------------------------------------------- I just got a flyer from American Digital, a reseller in Chicago (312/342-9600) that lists Multitech V.32EA modems (V.32 + MNP class 5) for $710 each. I think the price wars have started... ================================================================================ Note 116.22 9600 BAUD DIALUP MODEMS 22 of 32 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 24 lines 20-AUG-1989 10:09 -< Is the price right? >- -------------------------------------------------------------------------------- >< Note 116.20 by EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" > > -< Clarification on TB+ vs. T2500 >- > >The Telebit T2500 is their 300/1200/2400/PEP/V.32 version. Incidently, I >believe that Telebit is selling a ROM upgrade for Trailblazer Plus's to >turn them into T2500's for $500 or so. (Wow! A sanely-priced upgrade >option! ... >And with the uunet discount on the TrailBlazer Plus, this could get you >a T2500 for a very reasonable price. Uh, am I missing something here? With the discount from uunet, a TB+ plus $500 to make it into a T2500, puts it right in the ballpark (possibly even a bit higher) than a uunet-discounted T2500... I (personally) wouldn't label a ROM upgrade costing $500 "sanely-priced"; though I think that Telebit's strategy to allow upgrades of their proprietary hardware to include industry-standard protocols is laudable. BTW, do the Telebit modems utilize MNP on the slower speeds and/or V.32 (in the T2500/upgraded TB+)? Ed ================================================================================ Note 116.23 9600 BAUD DIALUP MODEMS 23 of 32 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 15 lines 21-AUG-1989 12:08 -< T2500 looks like the ultimate >- -------------------------------------------------------------------------------- > Check out the TrailBlazer and TrailBlazer+, I think they are > the way to go! I spoke with another satisfied user yesterday who is very happy with his two TrailBlazer+ modems. (What would be the plural of TrailBlazer+ if you misused it as a noun?) He says that he consistently gets 1100 bytes/second through them, even after the uucp overhead. He said that one way they get such good throughput is by "spoofing" the various protocols by generating Kermit (and similar protocol) ACKs locally, rather than waiting for the remote system to do it. This makes me just a bit nervous (data links shouldn't know about higher-level protocols), but if it works, why complain. Does the T2500 do this? ================================================================================ Note 116.24 9600 BAUD DIALUP MODEMS 24 of 32 EISNER::PRIGOT "Jonathan M. Prigot" 12 lines 21-AUG-1989 13:47 -< It's in there >- -------------------------------------------------------------------------------- > but if it works, why complain. Does the T2500 do this? Yes it does. You can view the T2500 as a TB+ with V.32. In answer to an earlier question, it does have MNP Class 5. HOWEVER! the MNP Class 5 and the synchronous features of the modem were not implemented in the unit I evaluated. There was a postcard stating this and by filling out the postcard, the will send you the updated ROMS when they are available. (This will be a no-charge upgrade.) I sent a message to them (they maintain a connection to the Internet), and they stated that this upgrade would be out by the end of September. ================================================================================ Note 116.25 9600 BAUD DIALUP MODEMS 25 of 32 EISNER::MCFARLAND "Sysmanageri Non Carborundum" 13 lines 14-OCT-1989 20:45 -< DataComm 296B 9600 baud modem >- -------------------------------------------------------------------------------- We are testing a DataComm 296B 9600 baud modem from General DataComm Inc, Middlebury, Connecticut. It is advertised as being "a full featured CCITT V.32 and V.54 compliant modem." I have tried calling DECUServe with the 296B. They answer the phone, but the handshake never completes. I don't have any trouble calling DECUServe with a VEN-tel 1200 baud modem, although the 10-20 second handshake wait is annoying. Reading mail at 1200 baud is also annoying. Does anyone have experience with this modem? Is it compatible with the DF242 modems that DECUServe uses? Bob ================================================================================ Note 116.26 9600 BAUD DIALUP MODEMS 26 of 32 EISNER::MAYHEW "Bill Mayhew" 1 line 15-OCT-1989 23:30 -< Unless it's V.22bis compatible, it won't talk 2400 baud >- -------------------------------------------------------------------------------- ================================================================================ Note 116.27 9600 BAUD DIALUP MODEMS 27 of 32 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 3 lines 16-OCT-1989 11:35 -< Just curious >- -------------------------------------------------------------------------------- > -< Unless it's V.22bis compatible, it won't talk 2400 baud >- What does "bis" stand for (if anything)? ================================================================================ Note 116.28 9600 BAUD DIALUP MODEMS 28 of 32 EISNER::LEDERMAN "Bart Z. Lederman" 8 lines 16-OCT-1989 14:14 -< C.C.I.T.T. Lingo >- -------------------------------------------------------------------------------- "bis" and "ter" are French abbreviations meaning something like 'above' and 'below' (Pierre, you out there?). They are sometimes used for C.C.I.T.T. reccomendations which have been changed enough or have had varients which are enough different from the original that they need a spearate designation, but not enough different that they want to give it a new number. In this case, I think V.22 was modified, and as far as I can recall nobody bothers with the original V.22 anymore. ================================================================================ Note 116.29 9600 BAUD DIALUP MODEMS 29 of 32 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 8 lines 16-OCT-1989 15:48 -< BIS + TER = 2nd + 3rd >- -------------------------------------------------------------------------------- > "bis" and "ter" are French abbreviations meaning something like > 'above' and 'below' (Pierre, you out there?). BIScuit is literally 'twice cooked', (TRIscuit 'thrice cooked'). BIS is the less familiar one. I think TER is probably readily recognised as 3rd, as in TERtiary. BIS would here read as 2nd or maybe 'again'. ================================================================================ Note 116.30 9600 BAUD DIALUP MODEMS 30 of 32 EISNER::HAHN "DECUServe Vice Chair" 2 lines 16-OCT-1989 20:48 -< BIS + TER etymology >- -------------------------------------------------------------------------------- BIS and TER are both of Latin origin not French. ================================================================================ Note 116.31 9600 BAUD DIALUP MODEMS 31 of 32 EISNER::MCFARLAND "Sysmanageri Non Carborundum" 7 lines 16-OCT-1989 20:52 -< DF242 isn't V.32 >- -------------------------------------------------------------------------------- I had thought that the DF242 was V.32 compatible, but it isn't. That's why I couldn't connect. Thanks for the help...and the French lesson. Bob ================================================================================ Note 116.32 9600 BAUD DIALUP MODEMS 32 of 32 EISNER::KENNEDY "Terry Kennedy" 9 lines 17-OCT-1989 05:33 -< 9600 *only*? Ick! >- -------------------------------------------------------------------------------- > We are testing a DataComm 296B 9600 baud modem from General DataComm Inc, > Middlebury, Connecticut. It is advertised as being "a full featured CCITT > V.32 and V.54 compliant modem." From subsequent notes indicate, this is *only* a 9600 baud modem. I find this rather unusual (all the V.32 / funny-9600 modems I've seen also do 1200 and 2400, and most do 300). Unless you're planning on converting all your gear to 9600 and giving up talking to (and being talked to by) older modems, this one looks like a loser...  ================================================================================ Note 117.0 Emulex LAT server 14 replies EISNER::OSUDAR "John Osudar" 31 lines 12-APR-1988 11:55 -------------------------------------------------------------------------------- We just spent a couple of days testing a pre-production version of the new Emulex LAT terminal server. Since other people may be interested in such a beast, I thought I'd post some comments and observations: First of all, it works. As you would expect with a pre-production version, there are still some bugs in the software (that runs in the server box), but Emulex seems interested in getting them resolved as quickly as possible. The box emulates a DECserver 200 in terms of commands, management, etc., but can handle up to 32 lines with one box (instead of eight). The user interface (in local mode) is nicely done, and appears to provide all of the 200's commands plus some additional useful ones. We ran into some problems with the wiring between the modular jacks and the four "full modem" lines (they worked in local mode but not through a data switch), and with the parallel printer port that comes standard with the box (although all of those problems are resolvable, and probably will be fixed by the time production servers are shipped.) The "partial modem" lines worked well with our data switch -- these provide one control signal in each direction, which is all we needed to make and break connections. Another problem we ran into was that the software had a hard limit of 48 services in its table; we're on a large Ethernet that has more than 48 LAT services. That problem is being looked into -- since they're trying to emulate a 200, it's likely this will be fixed. Otherwise, it was a fairly impressive product. Their software loads over the Ethernet similar to DEC servers, and of course they require no special software on the host systems (they talk LAT, and can connect to VMS and M-Plus systems that are running DEC's LAT software). The box is quite small, even with the 16-line expander that makes it a full 32-line server. (We didn't have an expander, just the 16-line base unit.) Our overall impression was quite positive. ================================================================================ Note 117.1 Emulex LAT server 1 of 14 EISNER::DELARISCH "Arnold S. De Larisch" 23 lines 12-APR-1988 12:19 -< Inquiring minds want to know...more! >- -------------------------------------------------------------------------------- >> < Note 117.0 by EISNER::OSUDAR "John Osudar" > >> -< Emulex LAT server >- >> >> We just spent a couple of days testing a pre-production version >> of the new Emulex LAT terminal server. Since other people may be >> interested in such a beast, I thought I'd post some comments and >> observations: A couple of questions: 1) Is the configurateion stored in Non-Volitle RAM or is loaded with the server image from the load host? 2) If not loaded from the load host, are their tools in place from the host side to configure the beastie with out being physically connected to it? 3) Does it do REVERSE-LAT well ... and how many lines (MAX) have partial modem control? Inquring minds want to know! -Arnold ================================================================================ Note 117.2 Emulex LAT server 2 of 14 EISNER::KILLEEN "Jeff Killeen" 2 lines 12-APR-1988 13:15 -< ALSO $$$$ COST >- -------------------------------------------------------------------------------- > -< Inquiring minds want to know...more! >- ================================================================================ Note 117.3 Emulex LAT server 3 of 14 EISNER::ERSKINE "Chris Erskine" 2 lines 13-APR-1988 06:46 -< Inquiring minds want to know...more and more! >- -------------------------------------------------------------------------------- Also do they support TSM or the NCP CONNECT command? ================================================================================ Note 117.4 Emulex LAT server 4 of 14 EISNER::OSUDAR "John Osudar" 94 lines 13-APR-1988 12:39 -< answers... >- -------------------------------------------------------------------------------- > Is the configurateion stored in Non-Volitle RAM or is loaded with the > server image from the load host? The configuration is stored in nonvolatile RAM. The Performance 4000 (that's the product name) is supposed to emulate a DECserver 200 so it works a lot like a 200 in most aspects. > If not loaded from the load host, are their tools in place from > the host side to configure the beastie with out being physically > connected to it? It has the same problems that 200's have. Unlike the old DECSA, there's no way to change configuration information in the server image on the loading host, so you have to do it either via a port on the server, or via NCP to the remote port. > Does it do REVERSE-LAT well ... and how many lines (MAX) have > partial modem control? It does reverse LAT; in the day and a half that we had our test 4000, we didn't get much of a chance to see how WELL it does reverse LAT -- but Emulex realizes that reverse LAT is an important aspect of LAT terminal servers. The available configurations include either 12 or 16 partial modem control lines in the base unit (with 12, you can either get 4 full modem control lines for a total of 16 lines, or just have 12 lines in the unit). Then there's an expander box that sits under the base unit, and comes with either 16 partial modem control lines, or 12 partial and 4 full modem control lines. List prices: (RJ12 connectors go with partial modem lines, DB25's with full) P4012 -- 12 line base unit (12 RJ12 connectors) $3595 P4016 -- 16 line base unit (16 RJ12 connectors) $4595 P4016M -- 16 line base unit (12 RJ12 + 4 DB25) $4995 Each base unit comes with a parallel printer port (Centronics or Dataproducts compatible) that looks like a LAT port to the host software. PE4016 -- 16 line expansion unit (16 RJ12 connectors) $2000 PE4016M -- 16 line expansion unit (12 RJ12 + 4 DB25) $2400 (One expansion unit can be attached to each base unit, giving a maximum of 32 lines per server) PS4001 -- Load software (RX50, TK50, or 1600 bpi tape) $200 Eventually, you'll be able to get a "PerformancePak" (personality module) that has the software already on it, and you won't have to load the server over a cable at all; that feature isn't available yet. > Also do they support TSM or the NCP CONNECT command? I don't have a definite answer about TSM -- however, their configuration file can be stored in DECserver 200 format, and they've tried to make it easy for a site with both 200's and P4000's to keep everything together in one directory, for easier management. The NCP CONNECT command is definitely supported; it gets you into port 34. (The parallel printer port is port 33.) A little more information: Emulex claims a peak throughput rate of 60000 characters/second (i.e. 32 lines at 19200 baud). Firmware is on a "PerformancePak" module that plugs in through the front of the server, right next to an LCD status display. This makes it easier to install firmware upgrades (e.g. the eventual full server software module). The server has DECnet RCF master capabilities for remotely configuring other (DEC or Emulex) terminal servers without using a host. (I am not sure exactly what this means -- but that's what the literature says!) The user interface (local mode) features VMS-like command recall and line editing, and a typeahead buffer. Also, all messages are time-stamped (the current time is loaded from the host VAX at server load.) The display of sessions (in local mode) includes a flag that indicates "data waiting", i.e. the host for that session has sent output to the terminal for that session. Each unit (base and expansion) is about 2.25" high, 14.5" wide, and 11" deep. Any more questions? ================================================================================ Note 117.5 Emulex LAT server 5 of 14 EISNER::OSUDAR "John Osudar" 13 lines 14-APR-1988 16:31 -< more info >- -------------------------------------------------------------------------------- Just found out one potentially nasty problem with the Emulex Performance 4000: the current version has a hard limit on services (and also nodes? -- that one's not clear) of 48 (compared to default 100, maximum 1000 on DECserver 200's) This may not be a problem for some, but our Ethernet currently has about 68 services (and even more nodes) and is growing... Our Emulex rep doesn't seem to consider this a serious/widespread problem -- so if you are interested in this product and the limit WOULD be a problem, be sure to inquire explicitly about it. (It has been stated that "the limit will be increased in the future", but increased TO WHAT, and WHEN, are unanswered questions -- and our rep thinks this is a low priority item. Of course, a salesman's opinion may not match the product manager's!) ================================================================================ Note 117.6 Emulex LAT server 6 of 14 EISNER::CLAPP 21 lines 16-JUN-1988 11:24 -< More P4000 experience >- -------------------------------------------------------------------------------- >> -< Emulex LAT server >- We've also been looking at the Emulex P4000 box. So far, I've also been favorably impressed. We don't have anywhere near 48 services, so that limit hasn't been a problem for us. We saw two problems using release 0.5 of the load image - one minor and one major (after logging off a TOPS-20 host, the server would crash...) Telephone support was prompt, knowledgeable and helpful. They have just shipped us version 0.6 and it has fixed both problems. I was reluctant to put users on a "release 0" product, but it seems to be (knock wood) robust so far. Version "1.0" is due out within a month. The release notes do say that the current version will not work with Ultrix. I guess what I like best about it is that the size is right. DEC's 8 line boxes are too small and the 128 line box is too big (or too expensive if only partially populated. The user (and management) interface is nicely done and the front panel LCD beats a single green LED. ================================================================================ Note 117.7 Emulex LAT server 7 of 14 EISNER::HASSINGER "Bob Hassinger" 16 lines 16-JUN-1988 12:28 -< Let the buyer beware... >- -------------------------------------------------------------------------------- I believe this product depends on reverse engineering of DEC's proprietary LAT protocol. As such it could be vulnerable to changes DEC makes. Note for example that both ends of a VAX to DECserver 200 have the LAT protocol implemented in software so DEC could make changes quite easily. Even if there is no intentional change it is very possible that changes and incompatibilities will creep into something this complex by the normal process of product engineering and support. So... you need to be confident your vendor will stand behind the product and guaranty to keep it working with new DEC releases and other changes for the full life time you expect to get out of it. I for one have been particularly disappointed with Emulex's track record in this regard on other products so I have very little interest in this one as a result, slick features and low price not-withstanding... Bob H ================================================================================ Note 117.8 Emulex LAT server 8 of 14 EISNER::DELARISCH "Arnold S. De Larisch" 13 lines 16-JUN-1988 13:39 -< I've been told LAT is Licensed to EMULEX >- -------------------------------------------------------------------------------- >> < Note 117.7 by EISNER::HASSINGER "Bob Hassinger" > >> -< Let the buyer beware... >- >> >> I believe this product depends on reverse engineering of DEC's >> proprietary LAT protocol. As such it could be vulnerable to changes >> DEC makes. I understand (although I've not confirmed it) that EMULEX is the first company to LICENSE the LAT protocol from DEC. This is 3rd hand info ... so I'm interested if anyone can confirm or deny this! -Arnold ================================================================================ Note 117.9 Emulex LAT server 9 of 14 EISNER::HASSINGER "Bob Hassinger" 12 lines 16-JUN-1988 15:52 -< Could you be thinking of Able rather than Emulex? >- -------------------------------------------------------------------------------- > -< I've been told LAT is Licensed to EMULEX >- At a recent LUG meeting here I believe they said no, it was strictly reverse engineered with no cooperation from DEC. I do think it was Able that just announced a product that *does* involve a DEC license however. I think I saw that it uses the DECserver 500 software directly - i.e. they just license the use of DEC software that loads into their box as opposed to licensing rights to write their own LAT compliant software. Details so far have been very scarce so this is more guess than fact. Bob H ================================================================================ Note 117.10 Emulex LAT server 10 of 14 EISNER::DELARISCH "Arnold S. De Larisch" 6 lines 16-JUN-1988 18:40 -< Able ... not Emulex Licenses LAT software! >- -------------------------------------------------------------------------------- >> < Note 117.9 by EISNER::HASSINGER "Bob Hassinger" > >> -< Could you be thinking of Able rather than Emulex? >- Me thinks you are right ... I think it was Able not Emulex! -Arnold ================================================================================ Note 117.11 Emulex LAT server 11 of 14 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 16 lines 1-JUL-1988 00:58 -< our brief test >- -------------------------------------------------------------------------------- We had an all to brief test of it, and so I will only make the following comments. 1) Won't download from RSX11M+ yet. Emulex said they know how to fix it. 2) I hate MMJs. At least these folks used normal modular jacks AND they USE the 1 + 6 wires for modem control! Dec's MMJ version of the 200 does not. 3) Emulex DID NOT SEEM to know much about the dual session support for VT340 class terminals that magically worked from DS200s running the latest software! They have more reverse engineering to do... 4) I wish Emulex and DEC would simply give me 25 pair telco connectors for each 8 ports wired as though I had Modtap DTE-SYS modulars collected into a 'harmonica' or octopus cable ending in a 25-pair connector. ================================================================================ Note 117.12 Emulex LAT server 12 of 14 EISNER::ZIMMERMAN "John Zimmerman" 1 line 8-JUL-1988 14:51 -< Parallel? >- -------------------------------------------------------------------------------- Does anyone have experience with the parallel port on the server? ================================================================================ Note 117.13 Emulex LAT server 13 of 14 EISNER::OSUDAR "John Osudar" 5 lines 27-JUN-1989 20:06 -< P4000+LN03R+Scriptprinter V2 won't work >- -------------------------------------------------------------------------------- The Emulex LAT terminal servers (Performance 4000) apparently are unable to drive an LN03R when using DEC's ScriptPrinter V2.0 software. The symptom is a "CONTERMINATED" error when attempting to start the queue. Emulex support people say they've seen the problem before and are working on it... I'll post more details when/if I get them. ================================================================================ Note 117.14 Emulex LAT server 14 of 14 EISNER::ALDRICH "Jeff Aldrich" 5 lines 25-OCT-1989 17:32 -< More thoughts anyone? >- -------------------------------------------------------------------------------- We are now in the process of deciding what to buy next. Any more recent experiences would be helpful. Now that DEC is licensing LAT, can someone confirm this feeling that I have that Emulex has purchased a license?  ================================================================================ Note 118.0 TK25 BOOT ON A PDP-11/73A 3 replies EISNER::KILLEEN "Jeff Killeen" 24 lines 13-APR-1988 00:23 -------------------------------------------------------------------------------- If you have not found out the hard way - there is no way to boot a TK25 on a PDP-11/73A (dual height) CPU with the DEC MXV11-B2 boot ROMs. The boot code is as follows..... 7776/ 46522 12701 172522 10102 5000 105711 100376 10704 112737 200 172523 5242 105711 100376 5711 240 5007 start at 10000G *NOT* 7776 ================================================================================ Note 118.1 TK25 BOOT ON A PDP-11/73A 1 of 3 EISNER::KENNEDY "Terry Kennedy" 20 lines 13-APR-1988 02:35 -< Possible work-arounds >- -------------------------------------------------------------------------------- > If you have not found out the hard way - there is no way to boot a TK25 > on a PDP-11/73A (dual height) CPU with the DEC MXV11-B2 boot ROMs. I thought TK25's spoke MS:? If so, some set of roms for a MXV11 should do it. Or, many 3rd-party disk controllers contain 'universal boot' code which can boot MS:. Of course, the question is why you want to boot a TK25 from ground zero in the first place - you should only have to do it once. If the reason is to launch an OS such as RSTS, and the system has an RX50 on it, you could do the following. [I presume similar stunts could be pulled for RT-11, and maybe RSX]: DSKINT a RX50 or RX33 diskette (or other suitable media found on both machines), cluster size 1, SATT.SYS base 16 <-- Important!. Copy INIT.SYS to it and HOOK it. This disk will then boot your '73 and you can say BOOT MS0: (or whatever) at the Option: prompt. This is non-obvious because INIT.SYS must be contiguous, and the defaults for DSKINT place the sat in the middle of the disk, effectivly halving your storage space. ================================================================================ Note 118.2 TK25 BOOT ON A PDP-11/73A 2 of 3 EISNER::KILLEEN "Jeff Killeen" 5 lines 13-APR-1988 10:02 -< I DON'T KNOW WHY >- -------------------------------------------------------------------------------- > I thought TK25's spoke MS:? If so, some set of roms for a MXV11 > should do it. Or, many 3rd-party disk controllers contain 'universal The MXV11 boot ROM's will boot a TSV05 with no problem - *BUT NOT A TK25*. DEC has confirmed this problem. ================================================================================ Note 118.3 TK25 BOOT ON A PDP-11/73A 3 of 3 EISNER::KENNEDY "Terry Kennedy" 0 lines 13-APR-1988 16:35 -< FPR - Firmware Performance Report - time? >- --------------------------------------------------------------------------------  ================================================================================ Note 119.0 LQP03 a viable item? 6 replies EISNER::LEDERMAN "Bart Z. Lederman" 20 lines 13-APR-1988 09:36 -------------------------------------------------------------------------------- The electronic store is now listing some items which DEC is going to sell at the next DECUS Symposium (for those of you who don't know it, DEC sells a number of items there at what are often very large discounts). One item listed is the LQP03 letter printer (daisy wheel) at what looks to me to be a very attractive price. Before I buy one, I'd like to know what people who have used it think of it (does it jam, is the print quality good, does it break down often, etc.). More important: since DEC isn't making them anymore, am I going to be able to get ribbons and print wheels for it? The office supply companies I know of which sell ribbons and print wheels, and which normally stock ribbons for DEC printers (LA120, LQP02, LP25, etc.) don't list LQP03 ribbons or print wheels. DEC sometimes uses industry components, but sometimes uses it's own design which can't be replaced with any other parts. Would I be getting a printer which, in a few years, would be out of service for a lack of ribbons and print wheels? ================================================================================ Note 119.1 LQP03 a viable item? 1 of 6 EISNER::KILLEEN "Jeff Killeen" 4 lines 13-APR-1988 10:08 -< THE PRICE EQUALS THE VALUE >- -------------------------------------------------------------------------------- The LQP03 is neither user freindly or reliable. I had to give a customer a credit on the one we sold. My advice is only buy it if you are going to use it for *LOW VOLUME* printing and a techie is going to running it. ================================================================================ Note 119.2 LQP03 a viable item? 2 of 6 EISNER::MAYHEW "Bill Mayhew" 8 lines 19-APR-1988 16:37 -< The Value Exceeds The (DECUS) Price >- -------------------------------------------------------------------------------- Jeff's experience does not correspond with ours. We have roughly a half-dozen installed at customer sites, with EXTREMELY non-techie users. Zero reliability problems for standard work (letters, etc., though they may have some problems with carrier-mounted envelopes due to the paper path -- anybody know of bottom-feed LQPs these days??) The only major user-hostile feature I find is the lack of a platen knob. (You have to use pushbuttons instead.) ================================================================================ Note 119.3 LQP03 a viable item? 3 of 6 EISNER::ROCHELEAU "Rod Rocheleau" 9 lines 23-APR-1988 02:30 -< Nobody wanted the one I had >- -------------------------------------------------------------------------------- I bought a used one for a secretary who wanted LETTER QUALITY print and didn't have the bread for a LQP02. She used it for 2 days and pitched it into the hall, note attached about where I could put it. I then presented it to another department that had been demanding letter quality, a week later they asked for their LA120 back. Need I say more. ================================================================================ Note 119.4 LQP03 a viable item? 4 of 6 EISNER::LEDERMAN "Bart Z. Lederman" 20 lines 16-AUG-1988 14:04 -< Well, I'm happy so far. >- -------------------------------------------------------------------------------- Well, I've had mine for a while now. It's not the fastest letter quality printer around: but it's a lot faster and neater than my Underwood "Noisless" (and a lot less effort to use, too!). I especially like the proportional spacing mode, and the fact that it's done in hardware so I don't have to send a lot of individual letter positioning commands. Overall, I'm reasonably pleased with the output. I have had no problems positioning paper with the "up" and "down" buttons. It's no more difficult than using a platten roller. I like the tractor feed. I have had NO jams at all. That is a lot more than I can say for other DEC printers: I have never yet used an LA50 or LA12 which did not jam constantly. At least one supplier identifies the LQP-03 ribbon as a Qume V ribbon (and the LQP03 documentation gives Qume part numbers). This is possibly the least common Qume ribbon, but at least it's not a totally foreign part. ================================================================================ Note 119.5 LQP03 a viable item? 5 of 6 EISNER::MAYHEW "Bill Mayhew" 6 lines 16-AUG-1988 14:54 -< Software to drive proportional hardware? >- -------------------------------------------------------------------------------- Just out of curiosity, Bart, what software are you using to drive the LQP03 in proportional mode? (The hardware can't do the right-justification, can it?) BTW, I got one too, and wish I had bought two... the second to act as a "live spare" since it's now getting fairly regular use in a new adjunct to our main business. ================================================================================ Note 119.6 LQP03 a viable item? 6 of 6 EISNER::LEDERMAN "Bart Z. Lederman" 3 lines 16-AUG-1988 17:34 -< None yet, but Bonner RNO in the future? >- -------------------------------------------------------------------------------- I'm not using any sofware to drive proportional mode now, that's what I like. According to the manual, I believe it could also do right justification, but I haven't tried it yet.  ================================================================================ Note 120.0 VAX 62xx question(s) 25 replies EISNER::OSUDAR "John Osudar" 2 lines 19-APR-1988 15:40 -------------------------------------------------------------------------------- Details, please, about the new VAX 62xx series ("Calypso") announced today (19-Apr-1988)? ================================================================================ Note 120.1 VAX 62xx question(s) 1 of 25 EISNER::TRAYSER "C J "Buck" Trayser - Notes Addict!" 17 lines 19-APR-1988 16:05 -< From sales literature... >- -------------------------------------------------------------------------------- I'm not a hardware type (If it can't be fixed from this side of my keyboard, then I usually gotta call someone), but here is some of the stuff I got from a piece of Sales literature... 6210 6220 6230 6240 CPUs 1 2 3 4 perf x 780 2.7 up to 5.5 up to 8.3 up to 11 memory 32-256 64-256 64-256 128-256 VAXBI buses 2-6 2-6 2-6 2-6 in-cab-expans 6220,6230,6240 6230,6240 6240 - Cabinet looks to be about the size of an 8550 I've heard its based on the MicroVAX 3000 chip, but couldn't find that in print. $ ================================================================================ Note 120.2 VAX 62xx question(s) 2 of 25 EISNER::OSUDAR "John Osudar" 1 line 19-APR-1988 16:23 -------------------------------------------------------------------------------- How about prices? Or delivery? ================================================================================ Note 120.3 VAX 62xx question(s) 3 of 25 EISNER::TRAYSER "C J "Buck" Trayser - Notes Addict!" 12 lines 19-APR-1988 16:35 -< Digital has it now! :-) >- -------------------------------------------------------------------------------- According to the Sales people "...they are available now!" I spoke to one of the engineers, his comment was "its based on the 3000 chip, so no production delays are expected". Price:  Hmmm, what is the DECUS policy on price quotes?  I'll play it safe...call your sales rep. $ ================================================================================ Note 120.4 VAX 62xx question(s) 4 of 25 EISNER::SCHWARTZ "Dana J. Schwartz" 1 line 19-APR-1988 19:43 -< Stoppit >- -------------------------------------------------------------------------------- Sheesh! Three keystrokes to get through one note! Gimme a break! ================================================================================ Note 120.5 VAX 62xx question(s) 5 of 25 EISNER::LYMAN "Rich Lyman" 23 lines 19-APR-1988 20:46 -< Hardware configuration of 6200... >- -------------------------------------------------------------------------------- The 6200 is a MicroVax 3xxx cmos chip with the clock speeded up to 80 ns. (35/3600 clock is 90 ns) System configurations: 1) 32-256 Mbytes in multiples of 64 meg. This is due memory interleaving between boards. One board (32 meg) is also a valid configuration. 2) 2-6 BI's 3) TK50 is load device. (too bad, DEC must have had lots of extra TK50's) 4) 100 MB/sec backplane (XMI) 5) 14 available slots. Can plug in CPU's, BI controllers and memory cards. (14 cards total) Can expand from 1 to four processors. Available immediately and will be shipped with VMS V5. ================================================================================ Note 120.6 VAX 62xx question(s) 6 of 25 EISNER::TRAYSER "C J "Buck" Trayser - Notes Addict!" 0 lines 20-APR-1988 10:20 -< I check with the proper people - price starts near 131k >- -------------------------------------------------------------------------------- ================================================================================ Note 120.7 VAX 62xx question(s) 7 of 25 EISNER::TINKELMAN "Bob Tinkelman - CCA" 133 lines 21-APR-1988 06:15 -< 6200 Fact Sheet - from DEC Press Release >- -------------------------------------------------------------------------------- Product Name: VAX 6200 Series Function: The VAX 6200 Series is Digital's new line of compact, mid-range VAX computer systems designed for easy CPU, memory, and I/O growth. It offers low ownership costs and high reliability. VAX 6200 Series systems can serve as interactive departmental or company-wide timesharing systems, and can be integrated into an existing information system with Digital's industry-leading networking products such as DECnet, Ethernet, VAXclusters, and Local Area VAXclusters. A tightly coupled multiprocessor architecture makes the VAX 6200 Series ideal for interactive applications such as computer-aided design and manufacturing, office automation, commercial data processing, and many other engineering, scientific, and general-purpose applications. Key Features and Benefits: o Easy, on-site processor and memory growth through the VAX 6200 Series with no increase in floor space o 14-slot, 100 Mbyte/second system interconnect can accommodate a mix of up to four CPU modules, 256 Mbytes of memory, and six VAXBI adapter modules o Two standard VAXBI channels with up to six total VAXBIs (using an expander cabinet) enabling the system I/O to grow with applications o Up to 256 Mbytes of shared ECC memory--one of the largest available on any mid-range system--allows large programs to be resident in main memory, providing fast execution o Provides from 2.8 to 11 times the performance of Digital's VAX 11/780 computer o The new Version 5 of the VMS operating system, supporting symmetric multiprocessing, dynamically balances workloads across all available CPUs, provides parallel programming tools, and allows Computer Interconnect (CI)-based systems to participate in Local Area VAXclusters o CMOS technology increases reliability by reducing parts count and running at lower temperatures o CMOS low power requirements mean low operating costs o Compact cabinet occupies just 6 square feet of floor space; fits into almost any office computing environment o Optional memory battery backup o Processor and memory upgrade packages, providing a singular, modular growth path from the entry-level VAX 6210 to the top-of-the-line VAX 6240 Sample Configurations: VAX 6210 system building block o 60"h x 30"w x 30"d cabinet o 1 CMOS VAX CPU o 2 VAXBI channels and 10 available VAXBI slots o 1 internal TK50 streaming tape load device o 32 Mbytes of ECC shared memory, expandable to 256 Mbytes o Ethernet controller o VMS operating system license o DECnet End Node license o One-year warranty on hardware and software o Performance: 2.8 times VAX 11/780 o Price: $170,500 VAX 6220 building block o 60"h x 30"w x 30"d cabinet o 2 CMOS VAX CPUs o 2 VAXBI channels and 10 available VAXBI slots o 1 internal TK50 streaming load device o 64 Mbytes of ECC shared memory, expandable to 256 Mbytes o Ethernet controller o VMS operating system license o DECnet End Node license o One-year warranty on hardware and software o Performance: up to 5.5 times VAX 11/780 o Price: $293,400 VAX 6230 system building block o 60"h x 30"w x 30"d cabinet o 3 CMOS VAX CPUs o 2 VAXBI channels and 10 available VAXBI slots o 1 internal TK50 streaming tape load device o 64 Mbytes of ECC shared memory, expandable to 256 Mbytes o Ethernet controller o VMS operating system license o DECnet End Node license o One-year warranty on hardware and software o Performance: up to 8.3 times VAX 11/780 o Price: $388,800 VAX 6240 system building block o 60"h x 30"w x 30"d cabinet o 4 CMOS VAX CPUs o 2 VAXBI channels and 10 available VAXBI slots o Internal TK50 streaming load device o 128 Mbytes of ECC shared memory, expandable to 256 Mbytes o Ethernet controller o VT320 console terminal o VMS operating system license o DECnet End Node license o One-year warranty on hardware and software o Performance: 11 times VAX 11/780 o Price: $537,400 VAX 6200 entry-level VAXcluster o 2 VAX 6210 VAXcluster systems o VMS operating system licenses o DECnet Full Function licenses o VAXcluster licenses o 2 LA 100 console terminal/stands o Storage building block o Star coupler o HSC70 storage controller o Disk interface o SA482 storage array with 2 RA82 disks o Tape interface o TA79 tape drive o VAX Performance Advisor software licenses o One-year warranty on hardware and software o Price: $576,500 Availability: May 20, 1988 ### CORP/'88/314 ================================================================================ Note 120.9 VAX 62xx question(s) 9 of 25 EISNER::BLOOM "Bob Bloom, Raytheon Company" 12 lines 21-APR-1988 17:35 -------------------------------------------------------------------------------- > How about prices? Or delivery? From the Digital Store 20-APR-1988 VAX 6210 system building block $170,500 VAX 6220 Building block $293,400 VAX 6230 system building block $388,800 VAX 6240 system building block $537,400 VAX 6200 entry-level VAX cluster $576,500 Availablity: May 20, 1988 ================================================================================ Note 120.10 VAX 62xx question(s) 10 of 25 EISNER::DEC_NEWS_1 "News about DEC from DEC" 188 lines 25-APR-1988 23:01 -< VAX6200 Pricing (19-April-1988) >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. 20-Apr-88 04:05 PM Page: 1 ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! PRICING INFORMATION 19-April-1988 ******************************************************************************* HARDWARE ORDERING INFORMATION* ================ VAX 6210 SYSTEMS ================ U.S. DSMC Model Number Description List Price HW/SW 62AMA-AE/AJ 6210 VMS SYS PDUP $170,500 $ 871 62AMA-DE/DJ 6210 VMS SYS PPL 131,600 871 62ACA-AP/AT 6210 CLUST SYS PDUP 191,700 965 62ACA-DP/DT 6210 CLUST SYS PPL 146,600 965 SV-6A47A-AK/AN 6210 PRECONF SYS PDUP 219,300 1,140 SV-6A47A-AL/AM 6210 PRECONF SYS PPL 180,400 1,140 ================================== VAX 6200 ENTRY VAXcluster SYSTEMS ================================== U.S. DSMC Model Number Description List Price HW/SW SV-6A47B-AK/AN 6200 ENTRY CLUST PDUP $576,500 $2,957 SV-6A47B-AL/AM 6200 ENTRY CLUST PPL 479,100 2,957 HS62A-AK/AN 6200 FULL DATA REDUN OPT PDUP 138,800 617 HS62A-AL/AM 6200 FULL DATA REDUN OPT PPL 134,800 617 ================ VAX 6220 SYSTEMS ================ U.S. DSMC Model Number Description List Price HW/SW 62BMA-AE/AJ 6220 VMS SYS PDUP $293,400 $1,086 62BMA-DE/DJ 6220 VMS SYS PPL 242,500 1,086 62BCA-AP/AT 6220 CLUST SYS PDUP 316,000 1,181 62BCA-DP/DT 6220 CLUST SYS PPL 257,400 1,181 SV-6A47A-BK/BN 6220 PRECONF SYS PDUP 359,200 1,425 SV-6A47A-BL/BM 6220 PRECONF SYS PPL 308,300 1,425 ================ VAX 6230 SYSTEMS ================ U.S. DSMC Model Number Description List Price HW/SW 62CMA-AE/AJ 6230 VMS SYS PDUP $388,800 $1,226 62CMA-DE/DJ 6230 VMS SYS PPL 326,000 1,226 62CCA-AP/AT 6230 CLUST SYS PDUP 414,500 1,321 62CCA-DP/DT 6230 CLUST SYS PPL 340,900 1,321 SV-6A47A-CK/CN 6230 PRECONF SYS PDUP 471,700 1,635 SV-6A47A-CL/CM 6230 PRECONF SYS PPL 408,800 1,635 ================ VAX 6240 SYSTEMS ================ U.S. DSMC Model Number Description List Price HW/SW 62DMA-AE/AJ 6240 VMS SYS PDUP $537,400 $1,366 62DMA-DE/DJ 6240 VMS SYS PPL 470,600 1,366 62DCA-AP/AT 6240 CLUST SYS PDUP 563,700 1,461 62DCA-DP/DT 6240 CLUST SYS PPL 486,400 1,461 SV-6A47A-DK/DN 6240 PRECONF SYS PDUP 635,200 1,846 SV-6A47A-DL/DM 6240 PRECONF SYS PPL 568,400 1,846 ========================================= VAX 6200 PERFORMANCE ENHANCEMENT PACKAGES ========================================= U.S. DSMC Model Number Description List Price HW/SW 62BUA-AE 6210-6220 VMS UPG PDUP $147,400 $215 62BUA-DE 6210-6220 VMS UPG PPL 133,000 215 62CUA-AE 6220-6230 VMS UPG PDUP 114,500 140 62CUA-DE 6220-6230 VMS UPG PPL 100,200 140 62DUB-AE 6230-6240 VMS UPG PDUP 178,300 140 62DUB-DE 6230-6240 VMS UPG PPL 173,500 140 ========================= VAX 6200-SPECIFIC OPTIONS ========================= U.S. DSMC Model Number Description List Price HW/SW MS62A-AB 32 MB ECC 6200 MEMORY $23,900 N/A DWMBA-BA/BB 6200 VAXBI EXP. CAB 24,600 $ 75 DWMBA-CA 6200 VAXBI EXP. CHANNEL 11,200 30 DWMUA-AA/AB 6200 VAXBI/UNIBUS CAB 29,200 229 H7231-N 6200 BATT. BACKUP OPTION 1,800 19 KDB50-C RA-DISK CTRL 8,800 71 CK-DEBNA-LD CAB KIT 3 FT CBL 6200 300 N/A ============================ NEW VAX 8550 SYSTEMS PRICING ============================ New VAX 8550 systems pricing is effective April 19, 1988. New U.S. Old U.S. Model Number Description List Price List Price SV-8E47B-EK/EN 8550 PRECONF SYS PDUP $453,000 $558,600 SV-8E47B-EL/EM 8550 PRECONF SYS PPL 404,000 510,300 855CD-AP/AT 8550 CLUSTER SYS PDUP 422,000 528,150 855CD-DP/DT 8550 CLUSTER SYS PPL 369,000 474,600 855BC-AE/AJ 8550 VMS SYS PDUP 400,000 506,100 855BC-DE/DJ 8550 VMS SYS PPL 351,000 456,750 855BC-BE/BJ 8550 ULTRIX SYS PDUP 400,000 506,100 855BC-EE/EJ 8550 ULTRIX SYS PPL 351,000 456,750 NOTE: Ordering information for the Start-up Service Solution Package will be available shortly. SOFTWARE ORDERING INFORMATION The processor types to be used in the new software order number format are as follows: VAX 6210 = 2 VAX 6220 = 3 VAX 6230 = 4 VAX 6240 = D VAX 6200 Entry VAXcluster System = Use new license rating U.S. Model Number Description List Price QA-001AA-H5 VMS Media and Documentation Kit $4,410 - TK50 stand-alone back-up H-Kit for the VAX 6200 series - Must be ordered separately Announcement Sales Kits containing the following have been sent to your Digital Representative: ED-31036-46 VAX 6200 Series Information Sheet EA-31027-46 Digital's VAX Systems Brochure EJ-31686-46 VAX Systems Positioning Charts ED-31539-46 VAX Systems and Options Catalog, Special Supplement for VAX 6200 Series EJ-31455-46 VAX Performance Bulletin: Featuring VAX 6200 Series Systems EA-31277-48 VMS Brochure EA-31276-48 VMS Information Sheet ED-31368-48 VAXcluster Software Information Sheet ED-31375-48 VAX FORTRAN Information Sheet ED-31607-76 ClusterWide Licensing Information Sheet EJ-31606-76 ClusterWide License Model Number and Price Wheel OTHER RELATED LITERATURE EB-25949-46 VAX Hardware Handbook, Vol. 1 EB-28348-46 VAX Hardware Handbook, Vol. 2 EB-29228-46 VAXBI Options Handbook EB-29863-48 VAX Software Buyers' Guide EB-30905-48 VMS System Software Handbook EB-29813-48 VMS Software Languages and Tools Handbook EB-25780-44 VMS Software Information Management Handbook EJ-31385-46 VAX 8800 Series Performance Summary (Internal Use Only) ED-31265-46 VAX Systems and Options Catalog (April-June '88) ED-31240-46 VAX Systems and Options Special Supplement (featuring VAX 8800 Series) ED-29730-46 VAX 8250 Information Sheet ED-29731-46 VAX 8350 Information Sheet ED-29385-46 VAX 8530 Information Sheet ED-29386-46 VAX 8550 Information Sheet ED-31586-46 VAX 8800 Series Information Sheet ED-29934-41 MicroVAX 3500 and 3600 Information Sheet A/V MATERIALS FROM RECENT ANNOUNCEMENTS 13434 VAX 8810/8820/8830/8840/8842 Technical Slides/Script 13435 VAX 8810/8820/8830/8840/8842 Executive Overview (slides/script) Engineering with the VAX 8820/8830/8840 (Video Testimonial) ================================================================================ Note 120.11 VAX 62xx question(s) 11 of 25 EISNER::DEC_NEWS_1 "News about DEC from DEC" 113 lines 25-APR-1988 23:02 -< VAX 6200 Questions and Answers (Digital News) >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. 20-Apr-88 04:06 PM Page: 1 ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! QUESTIONS AND ANSWERS - 19-April-1988 ******************************************************************************* PRODUCT POSITIONING Q: DOES THE VAX 6200 SERIES REPLACE THE VAX 8000 SERIES? A: No. The VAX 8000 series will continue to meet the needs of a wide range of users, especially those that require large uniprocessors for single-stream applications. Q: DOES THE VAX 6200 SERIES HAVE THE SAME PROCESSOR AS THE MicroVAX 3500/3600? A: The VAX 6200 series uses the same CMOS-based CPU as the MicroVAX 3500/3600. However, the speed of the VAX 6200 series CPU is 80ns, slightly faster than the 90ns CPU in the MicroVAX 3500/3600. Also, I/O and memory interconnects are significantly faster in the VAX 6200 series to accomodate multiprocessing and higher bandwidth I/O. Q: IS THIS THE VAX 8400? A: We have never used the term "VAX 8400," although others such as external press have. PRODUCT SPECIFICATIONS Q: WHEN WILL THE VAX 6200 SERIES BE AVAILABLE IN VOLUME? A: NOW! The VAX 6200 series is available immediately from inventory, and will begin shipping in volume during Q4/FY88. In all cases, availability for a specific order depends on the date that the order was placed. Q: ARE VAX 6200 SERIES SYSTEMS PLANNED WHICH INCORPORATE MORE THAN FOUR PROCESSORS? A: We have no current plans to announce VAX 6200 series systems with more than four processors. Q: WHAT IS THE HIGH-SPEED SYSTEM INTERCONNECT ON THE VAX 6200 SERIES AND HOW DOES IT COMPARE WITH THE VAXBI? A: The high-speed system interconnect on the VAX 6200 series provides I/O bandwidth of up to 100 MB/sec. This high throughput capability is used to support communication between processors, memory and I/O channels. The VAXBI provides up to 13 MB/sec of I/O bandwidth and is used for high-speed peripherals and communications options. Q: DO THE SYSTEMS SUPPORT EXISTING UNIBUS PERIPHERALS? A: The VAX 6210 and VAX 6220 provide limited support for the UNIBUS. The UNIBUS channel is announced and will be available in Q1/FY89. Systems that are configured with a UNIBUS cannot support VAXBI expansion beyond the two VAXBI channels that are standard with the system. Also, only a few UNIBUS options are supported, as described in the announcement article in this issue and in the VAX System and Option Catalog Supplement. There is no UNIBUS support for the VAX 6230 or VAX 6240. To upgrade beyond a VAX 6220 configuration, the UNIBUS must be removed from the system and VAXBI options substituted. Q: CAN ANY UNIBUS SYSTEMS BE UPGRADED? A: VAX 6210 system can be upgraded to a VAX 6220 system. Any customers using UNIBUS devices should have an account plan that addresses long- term growth. Q: ARE THE SYSTEMS FULLY SUPPORTED IN A VAXcluster SYSTEM, INCLUDING LOCAL AREA VAXcluster PHASE II? A: Yes. VAX 6200 series systems are available in VAXcluster system configurations. The VAX 6200 series is fully compatible with the VAXcluster architecture, including Local Area VAXcluster Phase II. Q: WHICH DISK DRIVES ARE SUPPORTED BY THE SYSTEMS? IS THE RA70 SUPPORTED? A: VAX 6200 series systems support the RA60/RA81/SA482 disk drives. The RA70 is not currently supported. Q: ARE VAX 6200 SERIES SYSTEMS FAULT-TOLERANT? WHAT HAPPENS IF ONE PROCESSOR FAILS? A: If one of the CPUs in a VAX 6240 fails, for example, the entire system fails. However, the system will reboot without the failed CPU, so that disruptions are minimized and operations can continue until the system can be serviced. Also, high availablity configurations can be obtained by combining multiple VAX 6200 series systems in a VAXcluster system. SOFTWARE Q: WILL ULTRIX-32 AND VAXELN BE AVAILABLE ON THE NEW SYSTEMS? A: ULTRIX-32 support for the VAX 6210 and VAX 6220 will be available in the next functional release of the operating system. VAXELN support for the VAX 6200 series is being investigated. PERFORMANCE Q: HOW MANY USERS ARE SUPPORTED ON THE VAX 6200 SERIES? A: This is application dependent. For example, ALL-IN-1 supports 60 active users on the VAX 6210, 104 active users on the VAX 6220, 148 active users on the VAX 6230 and 184 active users on the VAX 6240. ================================================================================ Note 120.12 VAX 62xx question(s) 12 of 25 EISNER::HASSINGER "Bob Hassinger" 1 line 27-APR-1988 10:54 -< Someone please translate "PDUP" and "PPL" in prices >- -------------------------------------------------------------------------------- ================================================================================ Note 120.13 VAX 62xx question(s) 13 of 25 EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 6 lines 27-APR-1988 13:19 -< You want to lease software? >- -------------------------------------------------------------------------------- PDUP -- Paid Up, you pay once and that is it PPL -- Periodic Payment Lic., you pay each month PPL is only available on BI machines, the last I knew. Seton ================================================================================ Note 120.14 VAX 62xx question(s) 14 of 25 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 14 lines 28-APR-1988 12:05 -< Why the large upgrade price differences? >- -------------------------------------------------------------------------------- >< Note 120.10 by EISNER::DEC_NEWS_1 "News about DEC from DEC" > >PRICING INFORMATION 19-April-1988 >VAX 6200 PERFORMANCE ENHANCEMENT PACKAGES >========================================= > U.S. DSMC >Model Number Description List Price HW/SW >62BUA-DE 6210-6220 VMS UPG PPL 133,000 215 >62CUA-DE 6220-6230 VMS UPG PPL 100,200 140 >62DUB-DE 6230-6240 VMS UPG PPL 173,500 140 Why does it cost $73,300 more to go from an 6230 to 6240 as opposed to a much lesser amount for 6220 to 6230 upgrade? Is this some marketing voodoo or does that upgrade require the addition of some seemingly *very* expensive hardware? ================================================================================ Note 120.15 VAX 62xx question(s) 15 of 25 EISNER::OSUDAR "John Osudar" 4 lines 28-APR-1988 12:18 -------------------------------------------------------------------------------- 6210 to 6220 adds 32MB memory (@ about $25K-$30K) 6220 to 6230 adds no memory 6230 to 6240 adds 64MB memory (@ about $50K-$60K) From what I can see, there don't appear to be any other differences. ================================================================================ Note 120.16 VAX 62xx question(s) 16 of 25 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 62 lines 28-APR-1988 16:01 -< Anomalies with 62xx upgrade prices? >- -------------------------------------------------------------------------------- >< Note 120.15 by EISNER::OSUDAR "John Osudar" > > 6210 to 6220 adds 32MB memory (@ about $25K-$30K) > 6220 to 6230 adds no memory > 6230 to 6240 adds 64MB memory (@ about $50K-$60K) > From what I can see, there don't appear to be any other differences. There is still something fishy here. Take the prices of the four processors, then determine the difference in each one from the processor "below" it. Then, where appropriate, deduct the cost for additional memory that the systems have. We should then have a theoretical "average" price for adding another processor. We can then look at the "Performance Enhancement" prices and see if they are reasonable. 6240 6230 6620 6210 -------- -------- -------- -------- VMS SYS PPL 470,600 326,000 242,500 131,600 Minus cost of lesser processor 326,000 242,500 131,600 ------- ------- ------- "Extra" cost 144,600 83,500 110,900 for this proc Minus 32/64 MB 47,800 0 23,900 @ $23,900 per 32 MB ------- ------- ------- "Cost per added 96,800 83,500 87,000 processor" Now if you average these figures you get $89,100. This would appear to be what it costs for another 6200 processor. Now lets take a look at the upgrade kits: 6240 6230 6620 6210 -------- -------- -------- -------- VMS UPG PPL n/a 173,500 100,200 133,000 Minus Memory Cost 47,800 0 23,900 "Cost per added -------- -------- ------- processor" 125,700 100,200 109,100 Now subtract the $89,100 89,100 89,100 89,100 cost per processor from above -------- ------- ------- DEC's Mystery Money: 36,000 11,100 20,000 So the bottom line, assuming that memory and processor are the only differences amongst the various models, is that there can be a pretty big difference in price depending on what model you start out with and what you end up with. However, these numbers are so whacky that I'll bet somewhere along the line, as you move up you have to add some VAXBI expansion stuff (DWMBA-xA), at $10 or $20 K per. But then why do the preconfigured systems I started out with not exhibit this seeming anomaly? ================================================================================ Note 120.17 VAX 62xx question(s) 17 of 25 EISNER::MAYHEW "Bill Mayhew" 6 lines 28-APR-1988 17:42 -< DEC moves in mysterious ways... >- -------------------------------------------------------------------------------- I don't know this for certain, but I would bet that some of the "mystery money" goes into DEC's VMS Licensing pocket. Also, vaguely related to your point, note that one of the lower upgrades (I forget if it's the 10-20 or 20-30) requires addition of an additional cabinet. ================================================================================ Note 120.18 VAX 62xx question(s) 18 of 25 EISNER::KILLEEN "Jeff Killeen" 0 lines 28-APR-1988 19:52 -< WHAT IF YOU DO THE 6220 TO 6230 UPGRADE TWICE ;-) >- -------------------------------------------------------------------------------- ================================================================================ Note 120.19 VAX 62xx question(s) 19 of 25 EISNER::ERSKINE "Chris Erskine" 1 line 29-APR-1988 06:32 -< It gives you a 6230-2 which does not have supported software >- -------------------------------------------------------------------------------- ================================================================================ Note 120.20 VAX 62xx question(s) 20 of 25 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 4 lines 30-APR-1988 11:09 -< Upgrade pricing >- -------------------------------------------------------------------------------- I can't be sure without digging out my notes, but I think the upgrades also include an RA82 disk at each level. Alan ================================================================================ Note 120.21 VAX 62xx question(s) 21 of 25 EISNER::HAMPARIAN "Don Hamparian - Battelle" 10 lines 15-JUL-1988 04:19 -< Problems hooking printer to DMB on 6220 >- -------------------------------------------------------------------------------- We are helping to install a VAX 6220 which is going to replace a VAX 780. We had to buy to DMB boards because we are going to have both a line printer and a laser printer on the parallel ports. We have an old LP-26-EC line printer that we want connect to one of the parallel ports. We have a BC27A cable to do the job but once we hook it up the printer and Vax do not see "eye to eye." The Vax never sees the printer come on line and the print symbiont keeps saying printer offline. Any suggestions? ================================================================================ Note 120.22 VAX 62xx question(s) 22 of 25 EISNER::KENNEDY "Terry Kennedy" 15 lines 15-JUL-1988 17:40 -< Suggestions >- -------------------------------------------------------------------------------- > Any suggestions? 1) Does the LP26 (with the same cable) work on another system with an LP11 or DMF32 controller? If not, then: a) The printer is broken b) The cable is bad c) Someone has changed the switches inside the printer 2) I have an LP25 w/ BC27 cable, formerly on an LP11, now on a DMF32. My understanding is that the DMB is a BI-ized DMF, so it should work. 3) Do a SHO DEV L?A0:/FULL. What does it say? 4) Beware - I have been getting VMS V5 bugchecks from the YCDRIVER (LP on DMF32) - see 224.* in VMS for more info. ================================================================================ Note 120.23 VAX 62xx question(s) 23 of 25 EISNER::HAMPARIAN "Don Hamparian - Battelle" 2 lines 18-JUL-1988 11:24 -< Thanks >- -------------------------------------------------------------------------------- Re: 120.22 Thanks for the ideas.. I will pursue. ================================================================================ Note 120.24 VAX 62xx question(s) 24 of 25 EISNER::HAMPARIAN "Don Hamparian - Battelle" 2 lines 25-JUL-1988 10:24 -< Resolution of Printer Problem >- -------------------------------------------------------------------------------- It turns out the LP26 had a long line driver card (?) in it but the 6220/DMB/BC27A requires a short line driver card. ================================================================================ Note 120.25 VAX 62xx question(s) 25 of 25 EISNER::MCFARLAND "Sysmanageri Non Carborundum" 10 lines 15-OCT-1989 22:48 -< Emulated instructions may affect performance. >- -------------------------------------------------------------------------------- < Details, please, about the new VAX 62xx series ("Calypso") announced < today (19-Apr-1988)? No one seems to have mentioned that the VAX 6000 series, like the MicroVAX 3xxx and MicroVAX II on which the CPU is based, does not implement the entire VAX instruction set in hardware. Some of the instructions are emulated in software. Bob  ================================================================================ Note 121.0 The LN03 OPC is dead -- Long live the LN03 OPC 12 replies EISNER::MAYHEW "Bill Mayhew" 27 lines 19-APR-1988 16:57 -------------------------------------------------------------------------------- We, and our clients, have been having problems for the past several months with premature failures of LN03 "OPC" cartridges (part of the User Maintenance Kit -- the major part, actually). The symptom is that about halfway into the cartridge's rated life, a "bare spot" appears on the printed page -- about an inch or so up from the bottom, and two-thirds of the way across a normal ("portrait") page. This happens on a variety of different LN03 machines, all using genuine DEC toner and maintenance kits. We've been pursuing this on and off with Field Service for some time and the best answer we've gotten so far is "a bad lot of OPC cartridges". (The bad lot seems to span shipments running from last fall up through, at least, January of this year.) So far we've been successful in getting our distributor (we're a distributor-affiliated OEM) to replace the life-shortened cartridges, and we hear that the same is true with people who ordered them originally direct from DECdirect. There are also some reports that the problem has been seen elsewhere across the country. I'm interested in hearing if there's any consensus that this is indeed the problem, or if there's some other explanation. There were rumors of FCOs and such, but our Field Service office checked our machine and says that it is "up to date" -- the FCO consisted of the use of a second replaceable filter, which began shipping with User Maintenance Kits last fall, so of course we have it. There is also a belief that the problem may be due to some clogging of the screen over the fan at the top left corner of the back of the printer (viewed from the back). ================================================================================ Note 121.1 The LN03 OPC is dead -- Long live the LN03 OPC 1 of 12 EISNER::LEDERMAN "Bart Z. Lederman" 5 lines 19-APR-1988 19:39 -< I've seen it. >- -------------------------------------------------------------------------------- The problem I see is not as bad as yours: but most of the way through the rated life we do see weak spots near the bottom of the page. Ours are DEC kits ordered through DEC. It has been near the end of rated life, and the print is weak but not blank, so we have not raised a fuss. ================================================================================ Note 121.2 The LN03 OPC is dead -- Long live the LN03 OPC 2 of 12 EISNER::HASSINGER "Bob Hassinger" 15 lines 20-APR-1988 11:17 -< I think we are seeing the same thing - DEC has a problem >- -------------------------------------------------------------------------------- We have had a NUMBER of problems that sound similar. The DEC service people have blamed them them on a variety of things but for the most part the people we see know very little at this level. I suspect what you report could be a major source of what we are seeing also. With all the problems I have been seeing, this issue (reliability and cost of operation because of repairs and replaced kits) has started it's way up my list to the level where it could be major issue - i.e. making a fuss via DECUS and within our company at high levels, considering going to other products and most likely recommending others do so too. If the problems are "a bad lot" of OPCs then DEC had better say so soon and compensate us too if they want to keep us as satisfied LN03 customers. Bob H ================================================================================ Note 121.3 The LN03 OPC is dead -- Long live the LN03 OPC 3 of 12 EISNER::CARTER_A "Tony Carter" 3 lines 20-APR-1988 13:20 -< Seeing spots >- -------------------------------------------------------------------------------- We are in the same situation as Bart in .1, the spot appears near the end of the expected lifetime of the cartridge. We have seen it in cartridges we got in January of 1987 and November of 1987. ================================================================================ Note 121.4 The LN03 OPC is dead -- Long live the LN03 OPC 4 of 12 EISNER::MAYHEW "Bill Mayhew" 20 lines 20-APR-1988 15:11 -------------------------------------------------------------------------------- Re: .2, action to take... Bob, I agree. Basically the only way we got what little satisfaction we have on the issue, thus far, was to make one heckuva lot of noise at our local field service management over a period of three months, and to urge our customers to do the same (they were calling and complaining to us, since we resell the cartridges to them). You can guess what kind of a grade "field service management" got on the annual Field Service Survey which crossed my desk, last week. Re: .3: Tony (and Bart)... what do you think of as the "expected lifetime"? We see the problem at around 5000 pages on an OPC that's supposed to go 10,000 (or 8,000 at least). It shows up a lot less on LN03s used for program listings (which have lots of whitespace) than on those used for office work such as word processing (which have pages that are more completely filled with text). ================================================================================ Note 121.5 The LN03 OPC is dead -- Long live the LN03 OPC 5 of 12 EISNER::LEDERMAN "Bart Z. Lederman" 13 lines 20-APR-1988 19:43 -< I "believe" DEC >- -------------------------------------------------------------------------------- > Tony (and Bart)... what do you think of as the "expected lifetime"? > We see the problem at around 5000 pages on an OPC that's supposed I expect an OPC to last 10,000 pages: after all, I have to be believe DEC, don't I? I see the faint spot starting to show up probably something over 7000 to 8000 pages, with still useable output when the maintenence light comes on at 10000 pages. I'm also very careful to see that the OPC is not exposed to bright light during installation, isn't touched by fingers, etc. I hope it helps. ================================================================================ Note 121.6 The LN03 OPC is dead -- Long live the LN03 OPC 6 of 12 EISNER::CARTER_A "Tony Carter" 8 lines 22-APR-1988 13:37 -< OPC lifetime >- -------------------------------------------------------------------------------- > Tony (and Bart)... what do you think of as the "expected lifetime"? I, too ,expect the lifetime to be in the area of 10K pages. Typically we see a lifetime of 8-10K pages. We have LN03+'s and they are used for text and plots. I would expect that plots would contain less dots than typical text pages, so that may help our cause. It is always "the spot" that causes me to change the cartridge, the rest of the page usually shows little degradation. ================================================================================ Note 121.7 The LN03 OPC is dead -- Long live the LN03 OPC 7 of 12 EISNER::BRUCKER "Ken Brucker" 12 lines 4-MAY-1988 13:37 -< The "half life" of the LN03 >- -------------------------------------------------------------------------------- > < Note 121.0 by EISNER::MAYHEW "Bill Mayhew" > > -< The LN03 OPC is dead -- Long live the LN03 OPC >- > > The symptom is that about halfway into the cartridge's rated life, > a "bare spot" appears on the printed page -- about an inch or so > up from the bottom, and two-thirds of the way across a normal > ("portrait") page. This is exactly what we have seen as well. The one item that isn't mentioned here (but is refered to in 128.*) is the *atrocious* price that DEC charges for this item, especially when it only lasts 2/3 it's 'normal' life. ================================================================================ Note 121.8 The LN03 OPC is dead -- Long live the LN03 OPC 8 of 12 EISNER::MAYHEW "Bill Mayhew" 11 lines 4-MAY-1988 14:23 -< We think the cost is OK >- -------------------------------------------------------------------------------- I don't find the cost atrocious, particularly. The supplies cost for an LN03 (toner + maintenance kits, at recommended intervals, divided by pages printed) is somewhere around 3.5 cents per page, which is competitive with other 8ppm printers last I looked. DEC is, at least, providing "warranty replacements" for OPCs that have failed prematurely, so the premature failure doesn't have any direct effect on operating costs, either. However, the problem itself does not seem to be being addressed in any visible way, nor is the warranty-replacement workaround well-known... THAT'S where my beef lies. ================================================================================ Note 121.9 The LN03 OPC is dead -- Long live the LN03 OPC 9 of 12 EISNER::MAXWELL "Gary Maxwell" 8 lines 5-MAY-1988 22:34 -< Boy, this all sounds familiar >- -------------------------------------------------------------------------------- Gee, folks, I just read all these notes, and realized that both of our TI Omnilaser 2115's experience the identical faded spot into their OPC lifecycles. We get our OPC through TI, which if the engine is indeed the same as the LN03 (I'm not sure) the OPC's are all made by the same people. Are yours made from a light green plastic frame? Jes' wonderin'... ================================================================================ Note 121.10 The LN03 OPC is dead -- Long live the LN03 OPC 10 of 12 EISNER::MAYHEW "Bill Mayhew" 7 lines 5-MAY-1988 23:36 -< I believe they use the same print engine >- -------------------------------------------------------------------------------- I believe the TI is indeed the same machine internally; if memory serves, it even looks the same in the ads. That's a Ricoh engine, and presumably Ricoh also makes the OPC. But it's very interesting that another machine with the same engine is in fact having the same problem; mayhaps DEC should pursue this with Ricoh, if DEC cares... (a dubious hypothesis, at best). ================================================================================ Note 121.11 The LN03 OPC is dead -- Long live the LN03 OPC 11 of 12 EISNER::MAYHEW "Bill Mayhew" 26 lines 20-MAY-1988 12:19 -< The beat goes on... >- -------------------------------------------------------------------------------- A client with around a half-dozen LN03s is continuing to have the premature-OPC-fade problem and called in DEC Field Service. Now, with backup from Atlanta, DEC IS CLAIMING THAT A LIFETIME OF 5000 PAGES PER OPC IS NORMAL (despite the fact that the copy fades well before the "replace OPC" light comes on). This, despite the fact that the published specs call for a lifetime of 10,000 pages. Customers have bought LN03s based on DEC's published specifications and based on certain life-cycle cost analyses, which, if DEC's new position is allowed to stand, are tossed into a cocked hat. I also noted in one of the trade journals in the past couple of weeks that Oki Electric and Ricoh have been hit with a $270 million class action suit for conspiring to defraud and deceive customers of Okidata's LL6 laser printer, charging the companies with selling the LL6 with a defective "developer kit" in the laser engine, which is Ricoh-manufactured. I don't know if the LL6 uses the same engine as the LN03, or what this "developer kit" is, but it sounds suspicious. The article suggests that the suit is being supported by American Express because its Buyers Assurance Program (which doubles the warranty period on AmEx-Card-purchased items) had received numerous complaints about the LL6, though Amex is not named as the plaintiff. The article appears in the May 9 issue of Computer & Software News. ================================================================================ Note 121.12 The LN03 OPC is dead -- Long live the LN03 OPC 12 of 12 EISNER::LEDERMAN "Bart Z. Lederman" 14 lines 20-SEP-1988 19:22 -< Related problem with Talaris >- -------------------------------------------------------------------------------- I talked recently with a friend who is using a TALARIS printer based on the same Ricoh works. He uses the DEC LN03 toner ant maintenance kits, which are apparently cheaper than others available to him. He has seen the blank strips appear on the OPC, but worse than we've seen: they go all the way across the page. He's tracked it down to the fact that the drawer holding the OPC doesn't seal against light as well as it should, especially if the printer is exposed to daylight, and it kills the OPC. I don't know if the LN03 is better or not: our printer is well away from windows. He also does one thing which is not in the maintenance manual, but which makes a lot of sense. He vacuums out the paper dust and other debris regularly.  ================================================================================ Note 122.0 Can a bad RA81 crash a good HSC50? 4 replies EISNER::HUTTO "Brent Hutto" 20 lines 21-APR-1988 16:44 -------------------------------------------------------------------------------- We have an 8600 with one HSC50 controlling ten RA81s and two TA78s. The HSC50 has crashed several ( 3 - 5 ) times in the last week or so. When it crashed, it gave up an error message concerning the "k.ci". During the same time period, one of our RA81s was accruing errors. Field Service insisted on: 1) reformatting the disk, 2) replacing the servo board, 3) replacing the microprocessor board and 4) replacing the HDA before agreeing to even look at the HSC50!!! They mumbled something about error codes and bad HDAs crashing HSCs, etc, etc. Unless I don't even conceptually understand HSCs ( which is the only level at which I claim to understand them ) there doesn't seem to be any likely way that one RA81 can wipe out the HSC itself. If this impression is correct, then I'M REALLY, REALLY MAD AT FIELD SERVICE for jerking us around for a week. If anyone can think of a mechanism by which a bad disk can cause a good HSC50 to fail and report a k.ci error, please enlighten me so that I can think that our Field Service guys know at least a little bit about what they're saying. ================================================================================ Note 122.1 Can a bad RA81 crash a good HSC50? 1 of 4 EISNER::KENNEDY "Terry Kennedy" 14 lines 21-APR-1988 23:39 -< Temporary insanity (of controller) >- -------------------------------------------------------------------------------- > If anyone can think of a mechanism by which a bad disk can cause a > good HSC50 to fail and report a k.ci error, please enlighten me... Caveat: I don't have an HSCx0, but I do have a number of RA-series drives on UDA50's, which amounts to the same thing for purposes of this discussion. Yes, an RA-81 can cause a crash of the controller, be it UDA, HSC channel, KDA, KDB, etc. The interface is electrically isolated, so you can't 'zap' the controller. However, you *can* mumble garbage protocol packets at the controller. In the case of the UDA, this will usually cause the UDA microcode to hang. However, normal host- initiated recovery procedures *should* work - you haven't actually damaged the controller, just driven it temporarily insane... ================================================================================ Note 122.2 Can a bad RA81 crash a good HSC50? 2 of 4 EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 10 lines 22-APR-1988 10:00 -< GIGO here as well >- -------------------------------------------------------------------------------- This is exactly what seemed to happen at our sight, as either an RA series disk or a TA/TU78 tape began sending lots (I mean LOTS) of errors to the HSC the HSC would crash. After talking with Colorado FS said that it appears that yes, GIGO. This was confirmed by CDC when they took over our hardware service. We have since moved to HSC350, and had many fewer problems. (We have also had fewer hardware errors as well...) Seton ================================================================================ Note 122.3 Can a bad RA81 crash a good HSC50? 3 of 4 EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 11 lines 27-APR-1988 16:26 -< Bye, bye HSC (with RA60, this time)... >- -------------------------------------------------------------------------------- It happened again. This afternoon I was attempting to run the SPM DISKSPACE utility on an RA60 which then began spewing errors. Down goes the HSC. Other one attempts to pick up the load. Down goes the other HSC. (Now the cluster is waiting for access to some disk somewhere...) One HSC back up. Lets repeat the sequence. I broke the cycle by taking the offensive disk off line, completely. This time, at least, no VAX crashed. This is a major improvement, to me at least. Seton ================================================================================ Note 122.4 Can a bad RA81 crash a good HSC50? 4 of 4 EISNER::FULLER_S "Stuart Fuller" 19 lines 28-APR-1988 14:26 -< It's the drive, not the HSC >- -------------------------------------------------------------------------------- What happens is that the host (the VAX, in this case) issues a command to an HSC to cause a disk drive to do something. In the course of executing the command, then HSC encounters an error from the drive (typically a data error, but other hardware faults in the drive could cause this). While the HSC is handling this fault (be it running diags, or doing ECC correction, or retrying the error...), the host is still waiting, with a timeout counter running. When the timeout counter runs out, the host assumes that the HSC has gone "insane", and will tell the HSC to reboot itself. This reboot action (in HSC lingo) as a "Host clear requested by k.ci". Meanwhile, the host has failed the disks over to the other HSC, and it retries the command on the other HSC. The other HSC encounters the same fault with the drive, and the cycle is repeated. Bottom line: the disk drive is broke. Fix it. Stu  ================================================================================ Note 123.0 MXV11B question 9 replies EISNER::NORTON "Bill Norton" 12 lines 22-APR-1988 14:01 -------------------------------------------------------------------------------- Does anyone know of a way to disable the RAM on a MXV11-B multifunction board? We need the boot ROMS on it for the 1173 half-size CPU, but we'd like to use modern (ie. large,parity) memory cards, rather than the 128Kb non-parity memory on the MXV. If there's no way to disable the MXV's RAM, how about putting the boot ROMs in a ROM-only card? Is there one that has the same window-mapping access method that the MXV11-B2's require on the MXV? ================================================================================ Note 123.1 MXV11B question 1 of 9 EISNER::KILLEEN "Jeff Killeen" 17 lines 22-APR-1988 17:15 -< SO YOU WERE SUCKERED INTO BUYING ONE OF THOSE ALSO >- -------------------------------------------------------------------------------- > Does anyone know of a way to disable the RAM on a MXV11-B > multifunction board? NO - THERE IS NO WAY > We need the boot ROMS on it for the 1173 half-size CPU, but we'd > like to use modern (ie. large,parity) memory cards, rather than > the 128Kb non-parity memory on the MXV. > > If there's no way to disable the MXV's RAM, how about putting the > boot ROMs in a ROM-only card? Is there one that has the same > window-mapping access method that the MXV11-B2's require on the MXV? Buy a MRV11 ROM card. The MXV11-B2 ROMs work fine on that card and the MRV11 documentation tells you how to set the jumpers to have it work as a boot card. (I have 6 out in the field) ================================================================================ Note 123.2 MXV11B question 2 of 9 EISNER::KOZAM 19 lines 22-APR-1988 22:14 -< Yes, you can disable MXV11-B RAM >- -------------------------------------------------------------------------------- < Note 123.1 by EISNER::KILLEEN "Jeff Killeen" > > NO - THERE IS NO WAY [to disable MXV11-B RAM] There certainly is! It isn't just flipping a switch, but disabling MXV11-B RAM is fully described in Micronote # 019 (10-JAN-85). It is aptly titled "Disabling RAM on the MXV11-BF". The modification requires 2 modifications: 1. Cut pin 13 of chip E17 2. Install a wire between pin 13 of E17 and pin 14 of E17 The micronote has a picture showing exactly where the chips are located, just in case you don't have a print set. If you don't have the set of Micronotes, I can send you a copy of note # 019 (it's only 3 pages long). Marc Kozam ================================================================================ Note 123.3 MXV11B question 3 of 9 EISNER::KILLEEN "Jeff Killeen" 12 lines 22-APR-1988 23:23 -< FS ISSUE >- -------------------------------------------------------------------------------- > There certainly is! It isn't just flipping a switch, but > disabling MXV11-B RAM is fully described in Micronote # 019 (10-JAN-85). > It is aptly titled "Disabling RAM on the MXV11-BF". Yup I was aware of this. I should have said there is no way to do this without cutting a chip pin on the board. The problem I ran into with this was Field Service would *NOT* sign off they would continue to maintain the board after this modification had been made. Since the "street" price of the MRV11 is about $350 and costs $250 *LESS* per year to maintain it was not worth the problems. ================================================================================ Note 123.4 MXV11B question 4 of 9 EISNER::KENNEDY "Terry Kennedy" 2 lines 25-APR-1988 00:25 -< One coulld always find 'Brand X' F.S. w/ a better attitude... >- -------------------------------------------------------------------------------- > ...this was Field Service would *NOT* sign off they would continue to > maintain the board after this modification had been made. ================================================================================ Note 123.5 MXV11B question 5 of 9 EISNER::NORTON "Bill Norton" 9 lines 26-APR-1988 15:57 -< Clarification please >- -------------------------------------------------------------------------------- >2. Install a wire between pin 13 of E17 and pin 14 of E17 Which part of the severed pin 13 gets the wire - the stub still on the chip, or the hole in the board that used to have pin 13 in it? This is really what I was after - the 11/73-MXV combo has sat idle for the las two years and I wanted a no-cost way to put it to use with a large memory board. ================================================================================ Note 123.6 MXV11B question 6 of 9 EISNER::KOZAM 16 lines 27-APR-1988 19:23 -< Clarifications >- -------------------------------------------------------------------------------- >2. Install a wire between pin 13 of E17 and pin 14 of E17 > Which part of the severed pin 13 gets the wire - the stub still > on the chip, or the hole in the board that used to have pin 13 in > it? It's the stub still on the chip that gets the wire. What you're doing is applying +5 Volts from pin 14 to pin 17. Otherwise, pin 13 could be sitting at an unknown logic state. About your FS not signing off - I bet their attitude would change right away if they were told that they can stay on, as long as they don't mind not being paid for that board. Just an opinion. Marc Kozam ================================================================================ Note 123.7 MXV11B question 7 of 9 EISNER::NORTON "Bill Norton" 1 line 28-APR-1988 08:56 -< Now - which one is E17? >- -------------------------------------------------------------------------------- ================================================================================ Note 123.8 MXV11B question 8 of 9 EISNER::ERSKINE "Chris Erskine" 1 line 29-APR-1988 06:29 -< The one between E16 and E18 >- -------------------------------------------------------------------------------- ================================================================================ Note 123.9 MXV11B question 9 of 9 EISNER::NORTON "Bill Norton" 1 line 29-APR-1988 09:10 -< Diagram please... >- --------------------------------------------------------------------------------  ================================================================================ Note 124.0 MSV11-SA help needed No replies EISNER::GARDNER "Tim Gardner" 12 lines 25-APR-1988 10:57 -------------------------------------------------------------------------------- I am looking for some information on the MSV11-SA memory board. Specifically, I need to know if the board has a configurable END address. DECdirect clearly states that the starting address is settable in 128Kb increments, but I have a (bizzare) application that requires that I set the total memory at something less that the full 2Mb. Since Digital has offered to sell MSV11-SA boards at DECUS for a rather significant (54%) discount, we might be able to use these boards in our product line. Thanks for you help! tg  ================================================================================ Note 125.0 mVAX test meanings 1 reply EISNER::HAHN 21 lines 25-APR-1988 11:08 -------------------------------------------------------------------------------- Re: mVAX ka630- a.v1.3 On booting the system goes through variou test sequences KA630-A.V1.3 Performing normal system tests. 7..6..5..4..3.. Loadin system software 2..1.. What are tests 2 and 1 Pierre ================================================================================ Note 125.1 mVAX test meanings 1 of 1 EISNER::PROVOST 37 lines 25-APR-1988 11:46 -< uVAX test error msgs >- -------------------------------------------------------------------------------- > < Note 125.0 by EISNER::HAHN > > -< mVAX test meanings >- > What are tests 2 and 1 I quote below from chapter 7, table 7-1 "LED Status and Error Messages" in the MicroVAX Maintenance Guide, AZ-GM3AA-MN. I hope this helps. Tom =========================================================================== 2 SEARCHING FOR BOOTSTRAP DEVICE NOTE: Be sure to try the remedies in the Troubleshooting section of the Owner's Manual before exploring these possibilities. NOTE: Please check signal and power cables before assuming drives or controllers to be defective. Once cables have been checked, examine power-up LEDs on individual devices and refer to section 7-3. 1. RQDX controller, RD5n drive, RX50 drive, or interconnect cable defective or not properly connected. 2. TQK50 controller, TK50 drive, or interconnect cable defective or not properly connected. 3. MRV11D module 4. DEQNA module, interconnect cables, fuse, ethernet transceiver, or thernet cables. 5. KA630-A module. 1. BOOTSTRAP DEVICE FOUND 1. Q22-bus bootstrap device 2. Signal cable to bootstrap device (defective or not connected) 3. Power cable to bootstrap device (defective or not connected) 4. KA630-A module 0 TESTING COMPLETED  ================================================================================ Note 126.0 VAX System Positioning Announcements 2 replies EISNER::DEC_NEWS_1 "News about DEC from DEC" 4 lines 25-APR-1988 23:04 -------------------------------------------------------------------------------- This topic will be used to introduce Digital Equipment Corporation VAX System Positioning announcements. Misty ================================================================================ Note 126.1 VAX System Positioning Announcements 1 of 2 EISNER::DEC_NEWS_1 "News about DEC from DEC" 147 lines 25-APR-1988 23:05 -< VAX 6200 Series Positioning >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. 20-Apr-88 04:08 PM Page: 1 ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! VAX 6200 SERIES POSITIONING - 19-April-1988 ******************************************************************************* o High-end performance in the smallest package yet o Price/performance leader in multistream applications o Exceptional expansion in every key dimension: CPU performance, I/O performance and memory o VAX 6200-based VAXcluster systems: an added dimension in availability and growth The following messages position the VAX 6200 series within the VAX family and against the competition. o High-end performance in the smallest package yet - Highest performance per square foot of any VAX system - Up to 11 times VAX-11/780 performance in a single, compact cabinet o Price/performance leader in multistream applications - IBM 4381-class performance in the IBM 9370 price range - Four price and performance levels: VAX 6210, 6220, 6230 and 6240 - Excellent multistream performance across a wide range of applications and environments o Exceptional expansion capability in all key dimensions - CPU performance: from 1 to 4 processors - I/O performance: up to 6 VAXBI channels - Memory expansion: up to 256 MBytes o VAX 6200-based VAXcluster systems for high availability - VAXcluster systems complement the VAX 6200 series multiprocessor architecture by providing high availability for critical applications. - VAX 6200 Entry VAXcluster System provides a VAXcluster system solution built around the new VAX 6200 series - Start-Up Service Solution Package and Full Data Redundancy Option allow you to customize basic VAXcluster systems VAX FAMILY POSITIONING MicroVAX Series: Exceptional Price/Performance, Office Packaging o Work group, small department or distributed solution o Performance from .9 up to 2.7 times a VAX-11/780 o Price/performance leader in VAX family o Flexible office packaging o VAXserver and LAVc configurations for distributed computing VAX 8250/8350: Entry-Level VAXBI Systems o Departmental solution o Performance from 1.2 up to 2.3 times a VAX-11/780 o Large system capability at entry-level prices o Full VAXcluster support, large disk capacity o Symmetric multiprocessing support (VAX 8350) under VMS V5.0 VAX 6200 Series: Mid-Range Family in a Single Cabinet o Departmental or enterprise-wide solution o Performance from 2.8 to 11 times the VAX-11/780 o Price/performance leader for multistream applications o Flexible, well-balanced CPU performance, I/O capacity and memory o System growth through in-cabinet upgrades or VAXcluster systems o Highest performance per square foot in the VAX family VAX 8550: Single-Stream Performance in a Compact Package o Departmental or enterprise-wide solution o Price/performance leader for large single-stream applications o Compact, cost-effective solution where expansion is not a requirement VAX 8800 Series: High-Performance VAX computing o Enterprise, data center or high-performance computing solution o Performance from 6 up to 22 times a VAX-11/780 o The highest performance VAX system in a single cabinet o System growth through in-cabinet upgrades or VAXcluster systems o Maximum single-stream performance, maximum number of supported users o Supports high-performance computational as well as general timesharing environments VAXcluster Series: High Availability and High Data Rates o Enterprise, data center and critical applications o Highest performance and growth paths o High availability for maximum productivity o High I/O rates for large database applications VAX FAMILY POSITIONING ENTRY-LEVEL SYSTEMS | MID-RANGE AND HIGH-END FOR WORK GROUPS AND | SYSTEMS FOR LARGE SMALL DEPARTMENTS | DEPARTMENTS AND ENTERPRISES | Micro- | VAX VAX VAX- VAX VAX | 6200 VAX 8800 cluster Series 8250/8350 | Series 8550 Series Series | Entry-Level X - | X X - - Price | | Single-Stream X - | - X X X Performance | | Multistream X X | X - X X Performance | | CPU Upgrade - X | X - X X Capabilties | | Largest CPU - - | - - X X Upgrade | Increments | | Maximum I/O - X | X - X X Capabilities | | Compact X X | X X - - Packaging | | Office X - | - - - - Packaging | Available | | Ethernet/LAVc X X | X X X X Expansion and | Dist. Env. | | CI VAXcluster - X | X X X X Expansion and | High Avail. | ================================================================================ Note 126.2 VAX System Positioning Announcements 2 of 2 EISNER::DEC_NEWS_1 "News about DEC from DEC" 158 lines 25-APR-1988 23:06 -< VAX MultiProcessor Positioning >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. 20-Apr-88 04:08 PM Page: 1 ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! POSITIONING MULTIPROCESSOR, - 19-April-1988 ******************************************************************************* UNIPROCESSOR AND VAXcluster SYSTEMS o Multiprocessors, uniprocessors and VAXcluster systems are complementary solutions offering extraordinary flexibility and complete compatibility o Propose multiprocessor systems for price/performance and in-cabinet growth o Propose VAXcluster systems for high availability and high I/O requirements o Propose uniprocessor systems for single-stream performance MULTIPROCESSING -- A NEW VAX ADVANTAGE Multiprocessors offer the advantage of being able to execute more than one application at a time across multiple processors. As a result, more "throughput" (or total work completed) can be achieved. High throughput, along with the upgrade capability of the new VAX 6200 series, adds up to a very competitive VAX solution for many applications. The ideal multiprocessor workload will include numerous discrete tasks that can be divided among the available processors. This is known as a "multistream workload." In a typical multistream workload, many users have simultaneous access to the system, and each processor performs a mix of compute and I/O tasks. General timesharing, office automation and process control are good examples of multistream workloads. Refer to the Markets/ Applications section of this issue for details. FEATURES/BENEFITS OF TIGHTLY COUPLED MULTIPROCESSORS Feature Benefit o Multiple processors reside - High performance/small package in a single cabinet - Growth capability, cost-effective expansion - System can reboot and continue to function if one processor fails o Multiple processors function as - Multiprocessing is transparent a single system. to end users Feature Benefit o Multiple processors function as - System management cost is minimized a single system. - Dynamic balancing of workloads o Multiple applications can execute - Improved system throughput for simultaneously. multistream workloads - Improved price/performance for multistream workloads o Multiprocessor can be a node - More flexibility in configuring in a VAXcluster system VAXcluster systems UNIPROCESSORS -- LARGE CPUs FOR LARGE JOBS Customers who have large single-stream applications will prefer VAX uniprocessor systems, such as the VAX 8530, VAX 8550 and VAX 8810 systems, since they can complete a single task more quickly than a multiprocessor of comparable power. In a pure single-stream application, the system processes only one task at a time. System performance is determined by the size and power of the CPU assigned to the task. Applications that fit this description include finite element analysis, seismic and mapping applications and simulation. Some single-stream applications (such as finite element analysis) can be modified to take advantage of the multiprocessor architecture. This process, called decomposition, divides a single-stream application into a number of subtasks that can be executed simultaneously across multiple processors in parallel (parallel processing). For single-stream applications that cannot be decomposed, a uniprocessor will continue to be the best solution. MULTIPROCESSORS AND VAXcluster SYSTEMS ARE COMPLEMENTARY VAXcluster systems are actually another form of multiprocessing, also called closely coupled multiprocessing. VAXcluster systems offer several important advantages over tightly coupled multiprocessors such as the VAX 6200 series. Each system has its own memory and system software, providing a high level of availability. VAXcluster systems can also handle very high data rates, which could overrun a tightly coupled multiprocessor system. Finally, VAXcluster systems offer the greatest ability to expand beyond a single cabinet. Thus, VAXcluster systems continue to be the system of choice where high availability is required, high data rates are anticipated, or where a high degree of growth is required. SUMMARY Tightly coupled multiprocessors provide large amounts of processing power in a small package. Uniprocessors fill the need for high-performance single- stream computing. VAXcluster systems composed of tightly coupled multi- processors and uniprocessor systems can provide the benefits of all three approaches. MULTIPROCESSOR, UNIPROCESSOR AND VAXcluster SYSTEMS Three Approaches to Providing Approximately 6 x VAX-11/780 Attributes: Key Benefits: Tightly Coupled Multiprocessors Price/Performance Leader (multistream) Single Cabinet Optimized for Multistream Workloads Multistream Performance: In-Cabinet Expansion Capability 5.5 x VAX-11/780 High Performance in Small Package Single-Stream Performance: Easy System Management 2.8 x VAX-11/780 Upgrade Path: to 11 x VAX-11/780 Attributes: Key Benefits: Uniprocessor Configuration Maximum Single-Stream Performance Single Cabinet Predictable Performance Multistream Performance: Easy System Management 6 x VAX-11/780 In-Cabinet Expansion Capability Single-Stream Performance: 6 x VAX-11/780 Upgrade Path: to 22 x VAX-11/780 (multistream) Attributes: Key Benefits: Closely Coupled Multiprocessors High Availability Multiple Cabinets High Data I/O Rates Multistream Performance: Flexible Configurations for 5.6 x VAX-11/780 Mixed Workloads Single-Stream Performance: Growth Across Multiple Dimensions 2.8 x VAX-11/780 Upgrade Path: In-Cabinet or In-Cluster  ================================================================================ Note 127.0 SI93C disk drives 1 reply EISNER::BORCHARDT "RODNEY P. BORCHARDT" 3 lines 27-APR-1988 14:33 -------------------------------------------------------------------------------- I have been looking into System Industries SI93C disk drives. Does anyone have anything good or bad to say about the reliabilty of the drives or the dsa interface. ================================================================================ Note 127.1 SI93C disk drives 1 of 1 EISNER::CONROY "Alan Conroy" 8 lines 15-JUL-1988 20:20 -< SI disks and controllers >- -------------------------------------------------------------------------------- SI repackages other companies' disk drives and so the answer to the question about reliability depends upon the real manufacturer of the drive (I don't know what the SI numbers mean for disk drive models, but if you have one, you can look inside and usually find the manufacturer name and model number). If you know the actual manufacturer, you will probably get a better response. As far as the SI controllers go, I wouldn't take one if they were handed out for free. My advice is to skip SI and go Emulex (and I don't even like Emulex any more, so that should tell you what I think of SI!)  ================================================================================ Note 128.0 LN03 AC/AD KITS 1 reply EISNER::PINSLEY "Howard W. Pinsley" 3 lines 27-APR-1988 20:53 -------------------------------------------------------------------------------- I am curious about where people are getting ln03 AC and AD maintenance kits. We are looking to find a low cost vendor (by definition not DEC). I'd be interested to hear who you use and what the cost is. ================================================================================ Note 128.1 LN03 AC/AD KITS 1 of 1 EISNER::MAYHEW "Bill Mayhew" 4 lines 27-APR-1988 23:07 -------------------------------------------------------------------------------- The guts of both kits are made by Ricoh (as is the print engine). We looked at buying Ricoh-branded units from a largish office supply distributor who tends to sell things at very competitive prices, but their price was not enough below DEC's to justify it.  ================================================================================ Note 129.0 Keyswitch Crashes 8650... 2 replies EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 11 lines 28-APR-1988 13:02 -------------------------------------------------------------------------------- Have I got one for you... We were just attempting to switch our 8650 over to RDC mode and had a minor problem... The key got stuck and wiggling it (just a little bit) to free the key caused the system to dump... How impolite of the the system! (I am just glad that it was lunch time and some people were away from the terminal...) Anyone else had a problem like this? Seton ================================================================================ Note 129.1 Keyswitch Crashes 8650... 1 of 2 EISNER::FULLER_S "Stuart Fuller" 5 lines 28-APR-1988 14:32 -------------------------------------------------------------------------------- What did the system dump with? That is, what was the bugcheck reason? Could it have been coincidental? Stu ================================================================================ Note 129.2 Keyswitch Crashes 8650... 2 of 2 EISNER::DROPPERS "Seton Droppers PBS:(703)739-5100" 7 lines 29-APR-1988 12:49 -< Powerfail... >- -------------------------------------------------------------------------------- Stopped with a powerfail. This had occured before, as I found out after asking around. DEC is replacing both keyswitches and the logic board supporting them. Seton  ================================================================================ Note 130.0 How to use VAX 6200s 4 replies EISNER::DEC_NEWS_1 "News about DEC from DEC" 12 lines 30-APR-1988 19:31 -------------------------------------------------------------------------------- Sorry folks, this has taken much longer than I expected. After he got back I was able to get a couple of things out of the suitcase, but then we needed to go shopping, take the kids to the baseball games, register them for school, it just went on and on. I finaly told them all that I would be persona-non-grata, the bringer of out of date news, if I did not get on with it... Here are copies of a bunch of VAX 6200 announcements I found lurking under the dirty socks. They seem to be DEC discussions of how you can use this wonderful new machine they built... Misty ================================================================================ Note 130.1 How to use VAX 6200s 1 of 4 EISNER::DEC_NEWS_1 "News about DEC from DEC" 41 lines 30-APR-1988 19:33 -< Introducing The VAX 6200 Series (19-April-1988) >- -------------------------------------------------------------------------------- Well, the first one is really an announcement... I think we all know about it by now, but I'll include it for completeness. Misty INTRODUCING THE VAX 6200 SERIES - 19-April-1988 ******************************************************************************* o A new series of mid-range VAX systems o Combining the latest CMOS technology with an innovative multiprocessor architecture o A line of fully compatible VAX computers that offers the customer new levels of price/performance and growth capacity On April 19, 1988, Digital will publicly announce a new series of mid-range VAX systems: The VAX 6200 series. This new series--comprised of the VAX 6210, VAX 6220, VAX 6230 and VAX 6240 --is built on three key technologies: Digital's CMOS VLSI VAX processor, a symmetric multiprocessing hardware and software environment, and the VAXBI I/O interconnect. With these technologies, the VAX 6200 series offers impressive system expansion capabilities and sets a new standard in price/performance and cost of ownership for a mid-range system. The VAX 6200 series is an integral new member of the VAX family. It provides a strong complement to the recently announced VAX 8800 series and to the existing MicroVAX 3000 series. VAX 6200 Performance Enhancement Packages, also being announced in this issue, provide a seamless growth path for your customers. The VAX 6200 series is also an important new component in Digital's VAXcluster strategy--providing even greater flexibility for VAXcluster system configurations. The VAX 6200 series is AVAILABLE IN VOLUME NOW! Demos are available in your local Area Centers for Technology. ================================================================================ Note 130.2 How to use VAX 6200s 2 of 4 EISNER::DEC_NEWS_1 "News about DEC from DEC" 83 lines 30-APR-1988 19:37 -< VAX 6200 in the Financial Services >- -------------------------------------------------------------------------------- VAX 6200 IN THE FINANCIAL SERVICES - 19-April-1988 ******************************************************************************* o Expandable mid-range system for departmental and large work group computing o Excellent system for many multistream application environments Computing in the Financial Services Industry typically involves multiple users running a range of applications as well as multiple users running copies of the same application. The applications may be batch or inter active. In either case, they tend to be I/O intensive and can be characterized as multistream, i.e., there are numerous discrete processes or jobs which can be run concurrently. With Digital's new multiprocessors and the symmetric multiprocessing support in VMS V5.0, multiple jons are dynamically distributed among the available processors. Multistream workloads share one processor on a uniprocessor system. WHEN THE VAX 6200 SHOULD BE CLUSTERED For applications that require high availability, we recommend that the VAX 6200 series system be clustered. For applications such as funds transfer, be certain that the software provides adequate recovery features to meet the customer's requirements. WHEN THE VAX 6200 IS NOT THE SYSTEM OF CHOICE Some financial institutions have applications that require greater system throughput than the VAX 6200 series provides. For those applications which require very high levels of single system performance, sell the VAX 88X0 Series. For those applications which can be clustered and the customer prefers a single cabinet cluster configuration, sell the VAX 8842. For customers who prefer to purchase large VAX systems clustered as a single "mainframe-like" system, sell the VAX 897X systems. For large modeling applications and demanding analytics development, a uniprocessor system such as the VAX 8550 or 8810 will, in many instances, provide better price/performance than a multiprocessor system. However, if the system is shared by a number of users and will frequently be running more than one process, the VAX 6200 series may still be preferable. Market data feed handling for the trading floor may be best supplied in many environments by networked MicroVAX systems, particularly the high availability MIRA configurations. POSITIONING THE VAX FAMILY IN THE FINANCIAL INDUSTRY A major competitive advantage for Digital in financial services is the wide range of compatible VAX systems which provide solutions platforms that span from the smallest bank, brokerage or insurance branch office to the large enterprise-wide resource centers of the largest financial institutions. LOW-END SYSTEMS--such as the MicroVAX 2000, MicroVAX II and the MicroVAX 3500/3600 are generally the system of choice for small-to-medium retail bank branches, retail brokerage and insurance agency offices. They are also the ideal system for many work groups in regional and global/international centers and are frequently the preferred system for use as a communication server. MID-RANGE SYSTEMS--are typically needed to support the processing requirements of large branch offices and large work groups. Digital's mid- range systems often provide the right level of processing power for departmental/regional computing needs. Often they also serve as ideal systems for incremental growth to VAXcluster systems at both regional and global/international centers. HIGH-END SYSTEMS--such as clustered 88XXs and 897Xs are the system of choice for the institution-wide data and information computing systems at global/ international centers and for some large regional centers. With the introduction of the VAX 6200 series, enhanced multi streamed throughput is available to Financial Services customers across the VAX mid-range. ================================================================================ Note 130.3 How to use VAX 6200s 3 of 4 EISNER::DEC_NEWS_1 "News about DEC from DEC" 134 lines 30-APR-1988 19:42 -< VAX 6200 in Engineering Applications (19-April-1988) >- -------------------------------------------------------------------------------- VAX 6200 IN ENGINEERING APPLICATIONS - 19-April-1988 ******************************************************************************* o Flexible configurations for wide range of performance o Throughput grows almost linearly with added processors o Excellent host system for mixed (and varying) interactive and batch computing demands o Growing list of supporting CMPs and SCMPs For a wide variety of engineering applications, the VAX 6200 series will span the performance range from MicroVAX 3600 to just below the VAX 8820/8800, typically as a host system for batch or interactive computing demands. The VAX 6200 series is superb in high-throughput, high-productivity environments. It has high-performance consistency over a range of jobs. Many engineering shops (especially with mixed workloads) think of their applications as "performance-oriented." These are, in fact, frequently "high thoughput" applications. In verifiable single-stream, performance-oriented applications that cannot be decomposed, Digital processors (such as the VAX 8550 or 8810) may be more appropriate. If it can take advantage of vectors, consider proposing a Floating Point Systems Processor plus a VAX 6200 series processor. VAX 6200 SERIES IN ESG MARKETS Mechanical Engineering Market Host system for interactive computing such as: engineering file and database access, technical calculations, sketching documentation and electronic mail/information exchange. Host system for batch computing, such as: stress, thermal, modal, fluid, aero, acoustic and kinematic analyses for engineering design decision trade- offs and evaluations. Parallel Processing: applications such as finite element analysis (FEA) have good potential for parallel processing. Batch oriented, compute-intensive FEA programs show, on average, a 50% performance boost with a second processor. ESG Cooperatvie Marketing Program vendors currently offering parallel VMS versions of their applications include: o ABAQUS from HKSI o I-DEAS from SDRC Other CMP FEA applications will be supported in the near future. Earth Resources Engineering Market VAX 6200 series machines are significant for the exploration and production departments of major oil companies. The applications most naturally suited for these new systems are: o Gridding and contouring software from CMP vendors such as Dynamic Graphics, Zycor and Radian o Seismic modelling from CMP Sierra Geophysics o Reservoir simulation from CMPs: ECL Ltd., J. S. Nolan and Scientific Software Intercomp Electronics Design Automation Market The tightly coupled VAX 6200 series will run applications such as SPICE very effectively in parallel. A SPICE circuit simulation job has been shown to run 60 to 70% faster with the addition of a second processor. The VAX 6200 series will also provide increased throughput in a multi-user environment. See the accompanying chart for some representative applications. Computer-Aided Software Engineering (CASE) Market The CASE environment includes: requirements specification, analysis, design, prototyping, coding, test and maintenance. Frequently, these applications are multi-user, which makes them natural for the VAX 6200 series. Third-party tools address front-ends--requirements specification, analysis, design and prototyping--to run on our platforms from VAXstations through VAX 8800 systems. The VAX 6200 series should be a powerful complement to VAXstations. VAX 6200 series is well suited to simulation and modelling of software in these applications: finite-element analysis, thermal analysis, vibration analysis (e.g., simulation of earthquake propagation), fourier transforms, and mathematically intensive software applications. The table that follows shows the peak throughput achieved by a VAX 6220 and VAX 6240 compared to a single-processor VAX 6210. All configurations had an HSC70 with RA82 disks. The results are based on multistream batch tests, where multiple copies of the same benchmark are run simultaneously. The PEAK throughput (jobs/hour) is compared across systems. VAX 6220 and VAX 6240 Multistream Throughput Relative to VAX 6210 The results show that throughput increased in an almost linear manner as processors and memory were added. APPLICATION VAX 6210 VAX 6220 VAX 6240 Mechanical: ABAQUS 1.0 1.98 N/A ANSYS (4 benchmarks) 1.0 1.91 to 1.95 3.73 to 3.88 NASTRAN (8 benchmarks) 1.0 1.76 to 1.98 2.56 to 3.91 PATRAN (8 benchmarks) 1.0 1.67 to 1.97 2.94 to 3.86 Electronics: SCICARDS 1.0 2.00 4.00 SPICE 1.0 1.99 4.00 Earth Resources: ECLIPSE 1.0 1.96 3.82 SIMBEST (SSI) 1.0 2.00 4.00 UTPETRO 1.0 2.00 4.00 VESPA 1.0 1.99 4.00 ================================================ OVERALL 1.0 1.67 to 2.00 2.56 to 4.0 ABAQUS is a trademark of Hibbitt, Karlson & Sorensen, Inc. I-DEAS is a trademark of Structured Dynamics Research Corporation SPICE is a registered trademark of the University of California at Berkley ================================================================================ Note 130.4 How to use VAX 6200s 4 of 4 EISNER::DEC_NEWS_1 "News about DEC from DEC" 64 lines 30-APR-1988 19:44 -< Positioning VAX6200 in Manufacturing (19-April-1988) >- -------------------------------------------------------------------------------- POSITIONING VAX 6200 IN MANUFACTURING - 19-April-1988 ******************************************************************************* o WHAT MANUFACTURING APPLICATIONS - Manufacturing Planning and Control System (MPCS) - Shop Floor Information Systems (SFIS) - Process Control (SCADA) o PERFORMANCE CHARACTERIZATION - CDIS Order-Entry Multi-User Workload - EMS Shop Floor Information System Workload CIM positions the VAX 6210, VAX 6220, VAX 6230 and VAX 6240 in the Manufacturing Computing Model as follows. PERFORMANCE CHARACTERIZATION The following information positions the VAX 6200 series using workloads based on actual CMP applications. A full report describing the details of the configurations, the methodology used and the details of the results will be published in the future. Until then, use the results reported on here only as a guide for characterizing the general performance capabilities of the systems and applications under test. GRAPHICAL SYNOPSIS The results of the Order-Entry/Distribution (CDIS) and the Shop Floor Information System (EMS) workloads substantiates the message that the VAX 6200 series provide an efficient and effective platform for scaling/ expanding the system to match the workload. The behavior of these workloads on the VAX 6230 and VAX 6240 indicated that VMS was efficiently distributing the computational work proportionally across each CPU for the given system. In other words, VMS V5.0 delivers the functionality that makes multi- processor systems operate efficiently. CDIS WORKLOAD CDIS is a fully integrated on-line distribution package developed by Computer Data Information Systems of Westlake Village, California. EMS WORKLOAD SFIS is a mid-range shop floor management application developed and licensed by Effective Management Systems of Milwaukee, Wisconsin. The workload reported on here is part of an integrated Shop Floor Information System (SFIS) which integrates data generated by factory data collection devices. It also provides a finite capacity schedule simulation application. EMS can manage the plant-wide manufacturing database. The results are calculated by taking the transactions per hour achieved using the MicroVAX II system and reporting how the other systems performed relative to the performance of MicroVAX II, expressed in transactions per hour. In the second graph, the VAX 6210 system is used as the basis for relative performance.  ================================================================================ Note 131.0 Hardware and Software Service Bundled 2 replies EISNER::DEC_NEWS_1 "News about DEC from DEC" 8 lines 30-APR-1988 19:51 -------------------------------------------------------------------------------- Well this one is a problem. I will put it here, but it crosses both the Hardware and Software lines. DEC is announcing that they will bundle Software Service with Hardware Service, and let Field Service administer the whole ball of wax. My apologies to those of you that are so immersed in Hardware that the mere sight of the word `Software' is offensive, but its the best I can do... Misty ================================================================================ Note 131.1 Hardware and Software Service Bundled 1 of 2 EISNER::DEC_NEWS_1 "News about DEC from DEC" 132 lines 30-APR-1988 19:52 -< Comprehensive Service Solution (19-April-1988) >- -------------------------------------------------------------------------------- COMPREHENSIVE SERVICE SOLUTION - 19-April-1988 ******************************************************************************* for the VAX 6200 SERIES o Comprehensive one-year hardware and software warranty with 24-hour/7-day coverage o Committed on-site hardware response time of 2 hours or less o Industry-leading "predictive maintenance" tools o Extended warranty and post-warranty total system service also available Comprehensive system support -- around the clock -- is included with every VAX 6200 series system. Digital continues its leadership in service and support by providing an integrated one-year hardware and software warranty with every VAX 6200 series system. This comprehensive warranty can be extended for up to three years through an extended warranty or through Digital's integrated post-warranty support program. Warranty and post-warranty support can be tailored to provide the customers with the appropriate level of service they need. The initial one-year warranty demonstrates the quality and reliability of our equipment, and provides customers with a full year of the best service in the industry at no charge. DESCRIPTION OF INTEGRATED SUPPORT Digital provides complete hardware installation and an integrated hardware and software warranty on all VAX 6200 series systems. This full one-year warranty covers all hardware, options, Digital-licensed operating system, and layered products included in the initial purchase. An especially important new feature of this warranty is the around-the-clock support. For hardware, this warranty includes DECservice support. Software support is provided at the same level as that provided with MicroVAX 3500/3600 systems. Customers can also take advantage of the price protection ensured by extending the warranty to cover the system for a total of two or three years. ONE-YEAR WARRANTY FEATURES The integrated hardware and software warranty includes: o 24-hour/7-day hardware and software coverage, including on-site corrective hardware service and telephone software support o Committed response time of 2 hours or less o Preventive maintenance o Installation of Hardware Field Change Orders o Remote diagnostic tools o Access to the Digital Software Information Network (DSIN) o Right-to-Use/Copy software updates o Conformance to SPD OPTIONAL FIRST-YEAR SERVICES In addition to the standard features, the following optional services are also available. o Additional Document Distribution Service (KZMC) per product for customers who require additional software documentation o Optional software installation for the operating system and layered products o Startup Service Packages (SSPs) The following services will be offered on these systems as soon as the service is available: o Software Update Installation Service o Warranty Uplift to System Support Plus NOTE: Neither initial media/documentation nor updates (SMS) are included with the warranty service. Both are available as an optional purchase per product. Also, the customer has the opportunity, at purchase, to select an extended warranty providing up to three full years of integrated hardware and software support. OPTIONAL POST-WARRANTY SERVICES Customers can continue the same high level of total system support enjoyed during warranty by tailoring a support program to meet their individual needs following the warranty period. Specifically, customers can select: o Integrated DECservice hardware support and Software Product Service (for kernel software only) o A la carte software services per layered product o Technology--Digital understands the technological environment which is rapidly, and radically, changing the type of service customers need today. We also use state-of-the-art technology to deliver that service, through predictive service tools such as VAXsim and SPEAR. These serve as the cornerstones of what experts have labeled "the most innovative use of technology" in the industry. o Expertise--is provided through a global support organization of 15,000 professionals, 450 service locations in 64 countries, 14 Customer Support Centers, and unmatched remote diagnostic technologies. And Digital has the best and most skilled professionals in the industry today, who work in partnership with management to ensure customers' business success far into the future. o Support--including software support that provides users with the most up-to-date information from the most experienced professionals -- including 24 hours a day telephone support from Colorado Springs. Our resource capabilities include 24-hour/7-day support, a worldwide logistics operation that can deliver the right part on time, wherever it is needed, and the financial resources necessary to continue to invest millions in the future of service technologies. SERVICE THAT'S MORE THAN REPAIR Today's Digital Field Service organization can help provide you with the Digital Difference Field Service's portfolio of service tools and technologies allows us to provide "predictive service," identifying problems BEFORE they occur. Tools, such as VAXsim for constant system integrity monitoring, and SPEAR to analyze and diagnose system problems, allow Digital to deliver service that is effective, cost-efficient and ensures customer satisfaction. This enables customers to depend on maximum availability and uninterrupted applications with their Digital computing solutions. ================================================================================ Note 131.2 Hardware and Software Service Bundled 2 of 2 EISNER::DEC_NEWS_1 "News about DEC from DEC" 62 lines 30-APR-1988 19:53 -< Integrated System Support Services (19-April-1988) >- -------------------------------------------------------------------------------- INTEGRATED SYSTEM SUPPORT SERVICES - 19-April-1988 ******************************************************************************* o Integrated "System Support" services for all VAX family systems o New, high level service provides comprehensive support 24-hours/7-days a week o Full availability across all VAX processors by Q1/FY89 Over the past year, Field Service has supported many new Digital systems with a one-year integrated hardware and software warranty. We are now announcing a new portfolio of integrated hardware and software "System Support" services to extend this integrated support beyond the warranty period. This provides the comprehensive, single source of support desired by Digital customers today. DESCRIPTION The new System Support portfolio will provide three levels of service: Basic System Support, Digital System Support and a new high-level System Support Plus service. These services are designed to enhance customer satisfaction and reduce cost of ownership by integrating key components of traditional hardware and software service into one easy-to-purchase package. In addition, three separate Software Product Services are being announced to supplement the System Support products: Layered Software Product Support, Media and Documentation Distribution Service and Software Update Installation Service. o System Support Plus--High-level, priority support 24-hours a day, 7-days a week for customers who need to maximize their hardware and software system availability and efficiency. o Digital System Support--This package is designed to provide customers with full on-site hardware service and advisory software support by extending the standard hours of hardware coverage to 16-hours a day, Monday through Friday. o Basic System Support--This level of service is for customers requiring a base level of hardware service, and software advisory support, 9-hours a day, Monday through Friday. SUPPLEMENTAL SOFTWARE PRODUCT SERVICES To supplement the customers integrated System Support service agreement, Field Service is also introducing the following services for software products. These new services are available only as supplements to the System Support services. o Layered Software Product Support--Extends the System Support features to layered products. o Media and Documentation Distribution Service--For customers with Layered Software Product Support (or under a one-year integrated warranty), this service provides distribution of media and documentation. o Software Update Installation Service--This service is for customers who want Digital to do the installation of their new software updates, and provide new version impact planning and analysis.  ================================================================================ Note 132.0 Versatec -BI/Unibus woes No replies EISNER::RENES "Stuart Renes, DFWLug Chair" 22 lines 5-MAY-1988 10:47 -------------------------------------------------------------------------------- We have acquired a Versatec 123 controller board and associated LWDRIVER device driver for our Versatec 7244 plotter. It all works just fine in PRINT mode, i.e., sending ASCII files to the plotter through a VMS queue (LWA0:). What it DOESN'T do is work in transparent DMA plot mode... garbage gets sent out and the device finally times out. After experimenting with burst mode settings on the 123 card, I'm beginning to wonder if the VAX BI/Unibus may not like this particular board in DMA mode (600kb/sec.?). Note that in PRINTER mode, it works in PIO (programmed I/O) mode. Anyone out there have experiences like this with BI/Unibus DMA class boards to share? Thanks in advance. Stuart.  ================================================================================ Note 133.0 PDP-11 / VAX Shared Memory 3 replies EISNER::SEASTREAM "Steven J. Seastream" 6 lines 6-MAY-1988 10:31 -------------------------------------------------------------------------------- Has anyone out there ever tried (successfully or unsuccessfully) an application involving shared memory among a Unibus PDP-11 and any VAX? We use PDP-11/44's for real-time data acquisition and control and are considering using VAXes as front-ends to our 11's. However, we are concerned about the speed of data transfer across Decnet and are interested in implementing a shared-memory system. ================================================================================ Note 133.1 PDP-11 / VAX Shared Memory 1 of 3 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 14 lines 8-MAY-1988 11:23 -< It has been done >- -------------------------------------------------------------------------------- >> Has anyone out there ever tried (successfully or unsuccessfully) >> an application involving shared memory among a Unibus PDP-11 >> and any VAX? Shortly after the 11/780 was announced, a paper was presented at DECUS describing this. They put both a CPU and memory on the VAX Unibus. As I recall, their biggest problem had to do with the fact that the Unibus can only have a single arbitrator. I'm sure that I have the paper here someplace, but doubt that I could find it in less than a week of searching. :-) Alan ================================================================================ Note 133.2 PDP-11 / VAX Shared Memory 2 of 3 EISNER::NAGY "Frank J. Nagy, VAX Guru" 3 lines 8-MAY-1988 17:18 -------------------------------------------------------------------------------- As I remember, the work Alan is describing was done at SLAC (Stanford Linear Accelerator Center) in Palo Alto and the paper should be available as a Technical Note from the Publications Office. ================================================================================ Note 133.3 PDP-11 / VAX Shared Memory 3 of 3 EISNER::PROVOST 41 lines 9-MAY-1988 08:47 -< Paper in PROCEEDINGS >- -------------------------------------------------------------------------------- > < Note 133.1 by EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" > > -< It has been done >- >>> Has anyone out there ever tried (successfully or unsuccessfully) >>> an application involving shared memory among a Unibus PDP-11 >>> and any VAX? >Shortly after the 11/780 was announced, a paper was presented >at DECUS describing this. They put both a CPU and memory >on the VAX Unibus. As I recall, their biggest problem >had to do with the fact that the Unibus can only have a >single arbitrator. > >I'm sure that I have the paper here someplace, but doubt >that I could find it in less than a week of searching. :-) See Proceedings of the Digital Equipment Computer Users Society Vol. 6, No. 4, Spring 1980 "A PDP-11 Front-end for a VAX-11/780" M. J. Browne, Charles Granieri, D. J. Sherden, Leon J. Weaver Stanford Linear Accelerator Center, Stanford California. Abstract An unpublicized feature of the VAX-11/780 is the provision for attaching a PDP-11 to the VAX UNIBUS Adapter. This can give significantly improved I/O performance for applications which are limited by overhead in the VAX I/O driver rather than by the transfer speed of the UNIBUS itself. We have implemented such a system using a PDP-11/04 as a "front-end" to a CAMAC data acquisition system. Both the PDP and the VAX have full access to the UNIBUS. That portion of the PDP address space which does not have UNIBUS memory can be mapped to buffers in the VAX memory, allowing the PDP to access VAX memory and to initiate DMA transfers directly to the VAX. The VAX also has full access to the PDP memory, providing a convenient means for developing and downloading the PDP software.  ================================================================================ Note 134.0 DV11 TIMING PROBLEM No replies EISNER::JOE 12 lines 11-MAY-1988 00:35 -------------------------------------------------------------------------------- Does anyone have any idea where I can get information on a DV11 Problem???????? We are running Digital Data Systems' PEP-70 Fast memory, and are experiencing a minor annoyance. On occasion, we experience a terminal lock-out from random terminals on our system. ( This is the Los Angeles Police Dispatch system, running on PDP-11/70's.) From the research I have done, we seem to not be up to snuff on our rev levels, most notably on the board that handles NPR's. Does any one have any info on timing problems, or where I can get the up to date ECO history of the DV11 Family?????  ================================================================================ Note 135.0 SID/XSID register layouts 2 replies EISNER::BRUCKER "Ken Brucker" 10 lines 11-MAY-1988 14:37 -------------------------------------------------------------------------------- Does anyone know what the SID and XSID register layouts are for the new VAXen that have been announced? Specifically I am looking for the layouts on the 3x00, 62x0, and 88x0 (polarstar) configurations. I have already tried CSC and they couldn't tell me much about the register layouts or what values to expect to find for the various CPU's and subtypes. Ken ================================================================================ Note 135.1 SID/XSID register layouts 1 of 2 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 33 lines 11-MAY-1988 16:28 -< KA650-AA SID >- -------------------------------------------------------------------------------- >> Does anyone know what the SID and XSID register layouts are for >> the new VAXen that have been announced? For the KA650 (3xxx series): SID (IPR 62): <31:24> Processor type = 10 (decimal) (CVAX chip) <23:8> Reserved for future use <7:0> Microcode revision level SYS_TYPE (physical address 20040004 hex): <31:24> System type code. This field reads as 1 for all single-processor Q-22 bus based systems. <23:16> Revision level. This field reflects the revision level of the KA650-AA firmware. <15:8> System subtype code. This field reads as 1 for the KA650-AA. <7:0> Reserved for use by Digital. These definitions are from the KA650 CPU Module Technical Manual, EK-KA650-UG-001. I would like to get some ACTUAL numbers for these fields from various systems. In particular, the fields reserved for Digital. Be sure to mention the system type (3200, 3500, etc.) Alan ================================================================================ Note 135.2 SID/XSID register layouts 2 of 2 EISNER::ECKERT "Jerry Eckert" 6 lines 31-JUL-1988 15:10 -< Polarstar SID format >- -------------------------------------------------------------------------------- For Polarstar systems: <31:24> Processor type = 17. <23:22> CPU number <21:16> Revision level <15:0> Serial number  ================================================================================ Note 136.0 Can you boot a VAxstation 2000 through the printer port? 2 replies EISNER::KOZAM 15 lines 11-MAY-1988 21:55 -------------------------------------------------------------------------------- Is it possible to boot up a Vaxstation 2000 using a terminal connected to its printer port? My system recognizes the terminal, prints the ROM based diagnostic tests, starts up uVMS, then dies, giving an error of "Unsupported CPU". My hunch is that uVMS sees the workstation hardware, then gets confused, since it is being booted up through the printer port. Why not use the workstation monitor? Well, the Vaxstation 2000 comes with a cute handle, which makes me want to take it with me when I travel. When I arrive, all I should need to do is hook up a terminal. Has anyone accomplished this? Marc Kozam ================================================================================ Note 136.1 Can you boot a VAxstation 2000 through the printer port? 1 of 2 EISNER::MAYHEW "Bill Mayhew" 21 lines 11-MAY-1988 23:13 -< You bet your pixels, you can! >- -------------------------------------------------------------------------------- Yes, it is possible, I have done it on several occasions, for much the same reason as you. (By "handle" you must, of course, mean the "cable retainer" :-) ). I use the DEC-recommended BCC08 "console cable", not the standard BCC05 "printer cable", to effect the connection. The difference is that the BCC08 jumpers pins 8 and 9 of the 9-pin connector together. This is the same trick used, incidentally, to get a Pro 350 to boot up to ODT. I have always commented-out the @STARTVWS line in SYSTARTUP.COM before I ran my machine this way. I've never gotten the error that you reported. I don't know if those two sentences are related... I think DEC is missing a nice niche market by not positioning the VS2K as a "portable VAX". If only there were a nice half-height good-capacity disk... Well, I am in contact with a user who has identified just such a beast, the MicroScience HH10-90, which supposedly can be found in the $1K range and is a roughly 80Mb disk. Instructions on formatting it can be found in one of the data libraries on CompuServe's VAX Forum (I think the MicroVAX library but I may be wrong). Unfortunately one sees _very_ little "market presence" for MicroScience. ================================================================================ Note 136.2 Can you boot a VAxstation 2000 through the printer port? 2 of 2 EISNER::KENNEDY "Terry Kennedy" 6 lines 12-MAY-1988 04:01 -------------------------------------------------------------------------------- > Unfortunately one sees _very_ little "market presence" for > MicroScience. Didn't Priam buy out Microscience's technology? I have had Priam 1/2 ht. drives in the clean room that looked *identical* to the newer Microsciences...  ================================================================================ Note 137.0 No MNP on -2574 No replies EISNER::TILLMAN "Brian Tillman, Smiths Industries." 3 lines 13-MAY-1988 09:34 -------------------------------------------------------------------------------- I just tried calling the -2574 number several times and, while I connected all right, no error control was established and my MNP commands messed up the autobaud, so I just got æææ...  ================================================================================ Note 138.0 Support for the DEQNA will disappear 6 replies EISNER::TILLMAN "Brian Tillman, Smiths Industries." 4 lines 23-MAY-1988 13:40 -------------------------------------------------------------------------------- I received a package from DEC today with some goodies in it. One of the goodies was a piece of paper announcing the discontinuation of support from the DEQNA. The paper said support would continue for at least a year, but you ought to consider switching to the DELQA. ================================================================================ Note 138.1 Support for the DEQNA will disappear 1 of 6 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 12 lines 23-MAY-1988 14:17 -< Possible upgrade for the DEQNA?? >- -------------------------------------------------------------------------------- >> I received a package from DEC today with some goodies in it. One of the >> goodies was a piece of paper announcing the discontinuation of support >> from the DEQNA. The paper said support would continue for at least a >> year, but you ought to consider switching to the DELQA. At one of the symposium sessions last week, I heard DEC say that they were looking into some kind of upgrade program for the DEQNA. It may be that they simply want to get rid of them, rather than support them forever (fat chance that they would!). Alan ================================================================================ Note 138.2 Support for the DEQNA will disappear 2 of 6 EISNER::CONROY "Alan Conroy" 9 lines 15-JUL-1988 20:26 -< DEC does it once again... >- -------------------------------------------------------------------------------- >At one of the symposium sessions last week, I heard DEC >say that they were looking into some kind of upgrade program >for the DEQNA. It may be that they simply want to get >rid of them, rather than support them forever (fat chance >that they would!). If anyone hears about such an upgrade program, I would certainly like to hear about it. It really *BURNS* me up to be caught with yet another piece of hardware that DEC sold us and now will not support... ================================================================================ Note 138.3 Support for the DEQNA will disappear 3 of 6 EISNER::STAMERJOHN "RW Stamerjohn" 6 lines 4-NOV-1988 16:54 -< Any new news? >- -------------------------------------------------------------------------------- Has anyone heard of any new news on any sort of upgrade program? The local office here says that there is no such thing, but I don't think that the question was very well researched. All comments/tips/suggestions/pointers appreciated. ================================================================================ Note 138.4 Support for the DEQNA will disappear 4 of 6 EISNER::ROECKEL 9 lines 13-FEB-1989 14:12 -< Upgrade Plan in the Works >- -------------------------------------------------------------------------------- > Has anyone heard of any new news on any sort of upgrade program? > The local office here says that there is no such thing, but I don't > think that the question was very well researched. Yes. We received a letter from our salesman that said DEC is going to offer an upgrade plan. Bascially, we were told to sit tight until furthur notice. ================================================================================ Note 138.5 Support for the DEQNA will disappear 5 of 6 EISNER::RAPPORT 19 lines 12-OCT-1989 14:08 -< any news folks? >- -------------------------------------------------------------------------------- < Note 138.4 by EISNER::ROECKEL > -< Upgrade Plan in the Works >- > Has anyone heard of any new news on any sort of upgrade program? > Bascially, we were told to sit tight until furthur notice. Is there an update to this.....? Ive got a few GPX stations with DEQNA's --5.2 upgrade says use DELQA or DEQSA (gesundheit) DESQA. signed, hopeful in Rochester. ================================================================================ Note 138.6 Support for the DEQNA will disappear 6 of 6 EISNER::MAYHEW "Bill Mayhew" 1 line 12-OCT-1989 15:39 -< See 163.* in this conference and 20.* in BUSINESS_PRACTICES >- -------------------------------------------------------------------------------- This topic should probably have been merged with 163. (Moderators?)  ================================================================================ Note 139.0 Other Mailer services for repairs 7 replies EISNER::HAHN 7 lines 24-MAY-1988 16:57 -------------------------------------------------------------------------------- The DEC mailer service is wonderful; however if you do not have an account with them (our University frowns on having "accounts" any place) the repair time is 6-8weeks which is not acceptable. What other places have you found that do board repairs. In my case it is a M8061 (RLV12) RL02 controller. Pierre ================================================================================ Note 139.1 Other Mailer services for repairs 1 of 7 EISNER::BOSTWICK 17 lines 24-JUN-1988 19:26 -< Here is one we have used >- -------------------------------------------------------------------------------- > What other places have you found that do board repairs. In my case > it is a M8061 (RLV12) RL02 controller. > Pierre We have used both DECmailer and an outfit called ESS at 414/255-4634. Both have been good. I don't believe you'll have any 'account' troubles with ESS - we have had good response and luck with them. They WILL do parts swap, parts sale, or repair your board. I know there are other outfits - this one is reasonably close, and has given us good service. jim ================================================================================ Note 139.2 Other Mailer services for repairs 2 of 7 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 2 lines 24-JUN-1988 23:25 -< THIS IS WHERE CDC GETS MOST OF THEIR BOARDS FIXED >- -------------------------------------------------------------------------------- > We have used both DECmailer and an outfit called ESS at > 414/255-4634. Both have been good. ================================================================================ Note 139.3 Other Mailer services for repairs 3 of 7 EISNER::KOZAM 12 lines 12-OCT-1989 19:18 -< DECmailer alternatives for RD53 >- -------------------------------------------------------------------------------- Has anyone used a mailer service to get an RD53 drive repaired? Here is what I found: Cost of used drive $ 399 Cost of DECmailer 685 Cost of Micropolis factor repair 325 DEC's cost of about $ 700 is out of the question with used drives going for $ 399. Micropolis (the maker of the RD53) wants $ 325, which is better than DEC, but barely better than just getting a used drive and using the broken one for spare parts. Anyone else out there going competent repairs at a good price? ================================================================================ Note 139.4 Other Mailer services for repairs 4 of 7 EISNER::KENNEDY "Terry Kennedy" 27 lines 13-OCT-1989 02:59 -< Repair prices / repair vs. used / RD53 vs. RD54 >- -------------------------------------------------------------------------------- > DEC's cost of about $ 700 is out of the question with used drives going > for $ 399. Micropolis (the maker of the RD53) wants $ 325, which is better > than DEC, but barely better than just getting a used drive and using the > broken one for spare parts. Remember that you can DECmailer/Factory Repair the logic board seperately from the whole drive. That reduces the cost to a *very* competitive price (As I recall, under $200). Of course, you need to be able to determine where the problem is *before* sending the drive logic (only) back. BTW, DEC's repair price on RM05 cards is lower than Imprimis' (the manu- facturer's) repair price. So, DEC isn't arbitrarily hiking the prices. I would vastly prefer a rebuild with warranty from the drive vendor over a used drive from "somebody". Drives are *not* something I'd buy used (at least not the 5.25" and smaller). By the way, if you have an RQDX3, the Maxtor XT-2190 (the RD54_ is avail- able at prices ranging from $1000 to $1395, *NEW*, quantity one). You might want to consider that as well. There are a lot of outfits which will repair drives at a flat rate (usually based on the capacity of the drive). I've not dealt with any of them, as I prefer to go with the manufacturer (and with lots of RD53 and RD54 lookalikes [over a hundred] I haven't had one fail yet, in about 3 years). You might want to check a publication called Service News, which has a large number of ads for these folks. ================================================================================ Note 139.5 Other Mailer services for repairs 5 of 7 EISNER::SAUNDERS "John Saunders, Pacer Software" 5 lines 15-OCT-1989 03:33 -< RD54 Power Requirements >- -------------------------------------------------------------------------------- When replacing an RD53 with an RD54 (or XT-2190), remember the difference in power requirements, especially the startup current. We got burned on this trying to configure a BA123 cabinet with three RD54s and an RD53. DEC would only support *two* RD54s, due to the the startup current. ================================================================================ Note 139.6 Other Mailer services for repairs 6 of 7 EISNER::KENNEDY "Terry Kennedy" 12 lines 15-OCT-1989 17:52 -< BA123 power supply oddities >- -------------------------------------------------------------------------------- > DEC would only support *two* RD54s, due to the > the startup current. That is a quirk of the switching supply in the BA123, and is a restriction for that cabinet only (the dual supplies must be paired RD54/RD53, RD54/RD53). However, if you really need more than the 450 Mb or so that this setup gives you, you should probably consider alternative disk configurations (such as 3rd- party ESDI / SCSI / SMD or DEC RA-type drives). Some of these configurations will fit in the existing BA123 (newer SCSI drives will get you up to 3.5 Gb with 2 drive bays and 1 dual slot), while others will require an external cab- inet. ================================================================================ Note 139.7 Other Mailer services for repairs 7 of 7 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 16 lines 16-OCT-1989 11:30 -< Preventing spin-up power surges >- -------------------------------------------------------------------------------- > However, if you really need more than the 450 Mb or so that > this setup gives you, you should probably consider alternative > disk configurations (such as 3rd-party ESDI / SCSI / SMD or > DEC RA-type drives). Some of these configurations will fit in > the existing BA123 (newer SCSI drives will get you up to 3.5 Gb > with 2 drive bays and 1 dual slot), while others will require > an external cabinet. Another advantage of some ESDI drives/controllers is that they don't spin up until the first time you access them. This prevents the power surge of multiple drives from happening all at once. I use the Andromeda ESDC controller (ESDI) and a pair of Imprimis (formerly CDC) 442 drives (formats to ~390 MB). The first drive spins up at boot time, the second one spins up when it is mounted in the startup command file.  ================================================================================ Note 140.0 Callback modem use with KERMIT 7 replies EISNER::NORTON "Bill Norton" 13 lines 24-MAY-1988 17:41 -------------------------------------------------------------------------------- A customer site has callback modems which we want to connect to from KERMIT sessions going through either a DMZ32 or DECserver 200. We would run KERMIT and connect to our modem port, make it dial the other system, and then what? What happens to KERMIT's connected port when that port rings? Is this concept workable through a served port? If so, how to set up the port's characteristics in the server, and with $set terminal? Has anyone done this? ================================================================================ Note 140.1 Callback modem use with KERMIT 1 of 7 EISNER::ERSKINE "Chris Erskine" 40 lines 25-MAY-1988 07:39 -< It works >- -------------------------------------------------------------------------------- >< Note 140.0 by EISNER::NORTON "Bill Norton" > > -< Callback modem use with KERMIT >- > > A customer site has callback modems which we want to connect to > from KERMIT sessions going through either a DMZ32 or DECserver 200. > > We would run KERMIT and connect to our modem port, make it dial > the other system, and then what? Normally after connecting to the other system, you login and use it. > What happens to KERMIT's connected port when that port rings? When the port is using the modem, the line is busy. > Is this concept workable through a served port? > If so, how to set up the port's characteristics in the server, and > with $set terminal? We do it all of the time. The terminal server has to be set for dynamic access and you have to create the LTA port using LATCP on the VAX. In Kermit, you then do a CONNECT LTAxx. You have to send a character to the modem to create the true link (Function of reverse LAT). You will also want to set the baud rate ... on the terminal server. I do not know what version of VMS you are using or if set terminal will effect the terminal server characteristics of the terminal server yet. It is best to have the terminal server set up for communications for the modem. You will want to also disable autobaud on it. > Has anyone done this? Yes. This link is from a terminal server modem although I am not using Kermit right now. If you have any problems or need some help you can give me a call. Chris ================================================================================ Note 140.2 Callback modem use with KERMIT 2 of 7 EISNER::NORTON "Bill Norton" 10 lines 25-MAY-1988 18:37 -< Clarification >- -------------------------------------------------------------------------------- Maybe I didn't make myself clear... The customer's modems are callback types where after it answers it verifies who you are and then *calls you* at a predefined number. If KERMIT is still connected to our LTA port, will the our modem answer the callback? Will the terminal server let our modem answer the callback, or does our reverse LAT connection to the port preclude this? ================================================================================ Note 140.3 Callback modem use with KERMIT 3 of 7 EISNER::ERSKINE "Chris Erskine" 1 line 26-MAY-1988 06:26 -< Never tried it but should work. >- -------------------------------------------------------------------------------- ================================================================================ Note 140.4 Callback modem use with KERMIT 4 of 7 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 4 lines 26-MAY-1988 16:47 -< I think it should work, also. >- -------------------------------------------------------------------------------- I would seem to me that even though you may be connected to the modem, DTR should be present, so he modem should answer. In fact, because you are connected, you should see any messages, like "RINGING" the modem wants to spit out. ================================================================================ Note 140.5 Callback modem use with KERMIT 5 of 7 EISNER::CONROY "Alan Conroy" 3 lines 15-JUL-1988 20:29 -< Slow... >- -------------------------------------------------------------------------------- We tried using Kermit through dial-out on the servers and found it horribly slow. Finally gave up and use direct-connect modems now. Anyone have an idea as to why we had these problems? ================================================================================ Note 140.6 Callback modem use with KERMIT 6 of 7 EISNER::ROECKEL 40 lines 19-JUL-1988 12:54 -< It works GREAT for us ! >- -------------------------------------------------------------------------------- > Maybe I didn't make myself clear... > > The customer's modems are callback types where after it answers > it verifies who you are and then *calls you* at a predefined number. > > If KERMIT is still connected to our LTA port, will the our modem > answer the callback? > > Will the terminal server let our modem answer the callback, or does > our reverse LAT connection to the port preclude this? We use KERMIT to access a Hayes 1200 on a DECserver 200. Here are the details of what we go through when accessing a system that has to call us back. 1. Run KERMIT. Issue the CONNECT command and connect to the LAT port we have set up using LATCP. 2. Type AT to test the connection and get the OK repsonse from the Hayes modem. 3. Enter the dial command and dial the number. (ATDT......) 4. Once connected, enter security info. The other system then tells us to hang up, so we use the HAYES escape command (+++) to get the attention of the modem, then tell the modem to hang up (ATH) 5. Wait until the modem senses ringing -- you actually see RING displayed on your screen. Then issue the HAYES answer command to tell the modem to answer the incoming call (ATA) Our modem has autoanswer disabled. 6. Once the modem answers, log in to the calling system. SIMPLE !! There is never a need to exit KERMIT. The HAYES escape command (+++) is VERY VERY handy. When you get hang up, you can use it to get the attention of the modem, then issue modem commands to hang up the line or whatever. Read the manual on how to use this command because there are very explicit rules!! Bruce ================================================================================ Note 140.7 Callback modem use with KERMIT 7 of 7 EISNER::ROECKEL 16 lines 19-JUL-1988 12:57 -< Port Speed Set Correctly? >- -------------------------------------------------------------------------------- > We tried using Kermit through dial-out on the servers and found it horribly > slow. Finally gave up and use direct-connect modems now. Anyone have an idea > as to why we had these problems? I may be way off base here, but did you check the DECserver port speed ? You should have autobaud disabled, Alternate Speed disabled, and have the port set for 1200 baud (or whatever you want the modem to work at.) I suspect that the port is set at 300 baud (or its set for autobaud, with the alternate speed set to 300 baud) and you are actually getting a 300 baud link. Bruce  ================================================================================ Note 141.0 Moving the LP27 and LP06 far away? 4 replies EISNER::ZIMMERMAN "John Zimmerman" 11 lines 25-MAY-1988 11:42 -------------------------------------------------------------------------------- I am looking to move a couple of my parallel printers (LP06 and LP27) a ways away, approx. 200 ft. The LP27 can be upgraded to a longline printer but this is a bit pricey (~4600). I have not been able to find a longline upgrade for the LP06ÿ. The only other solution I have been able to find is to use Emulex's new terminal server which contains a parallel port. Any suggestions? thanks jz ================================================================================ Note 141.1 Moving the LP27 and LP06 far away? 1 of 4 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 11 lines 25-MAY-1988 12:38 -< Less host overhead >- -------------------------------------------------------------------------------- >> -< Moving the LP27 and LP06 far away? >- >> The only >> other solution I have been able to find is to use Emulex's new >> terminal server which contains a parallel port. That solution has the advantage of eliminating the interrupt- per-character load on the host (assuming you are now using the LP11 interface). Alan ================================================================================ Note 141.2 Moving the LP27 and LP06 far away? 2 of 4 EISNER::KILLEEN "Jeff Killeen" 3 lines 25-MAY-1988 16:21 -< CALL BLACK BOX >- -------------------------------------------------------------------------------- Black Box sells units to do this for about 500-600 dollars. ================================================================================ Note 141.3 Moving the LP27 and LP06 far away? 3 of 4 EISNER::TINKELMAN "Bob Tinkelman - CCA" 15 lines 25-MAY-1988 16:37 -< Consider using a Serial-to-Parallel converter >- -------------------------------------------------------------------------------- One way is to place a "serial to parallel" converter in front of the printer. Then the problem is simply getting a serial channel near the printer. I gather that you would then use reverse LAT from on a nearby terminal server. Several companies make these. Be aware that there are two types parallel printers interfaces - Centronics and Dataproducts. They are almost the same, but different. In the PC world, and hence in most of the world, parallel means Centronics. DEC has traditionally used Dataproducts. As an example, see page 124 of the Winter, 1988, Black Box Catalog. This model is a little fancier than you need but costs only $395. It runs up to 38K bps and contains 8K of buffering and can be configured to run either serial-to-parallel or parallel-to-serial. Disclaimers: I am a Black Box "customer" only. I have not used this product. ================================================================================ Note 141.4 Moving the LP27 and LP06 far away? 4 of 4 EISNER::ERSKINE "Chris Erskine" 8 lines 26-MAY-1988 06:24 -< Another option >- -------------------------------------------------------------------------------- MBD Systems in Orange, CA also sell long line options for the both printers. Their interface can consist of a black (blue) box on one end and a board on the printer or 2 black boxes. I have used them a number of times and will do so again when needed. Their interface is compatable with the DEC long line options so you can match and mix. Chris - (an employee 10 yrs. ago but now a satisfied customer)  ================================================================================ Note 142.0 Speed Up a VAX 11/780 ? 3 replies EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sap" 9 lines 29-MAY-1988 15:50 -------------------------------------------------------------------------------- There is a company called NUMONIX that we bought a board from to speed up the CPU on our VAX 11/750. At DEXPO they announced a board that would speed up the CPU on a VAX 11/780. Has anyone used it or can offer any comments on this board ? ================================================================================ Note 142.1 Speed Up a VAX 11/780 ? 1 of 3 EISNER::KENNEDY "Terry Kennedy" 8 lines 30-MAY-1988 01:20 -< Faster than a 785? >- -------------------------------------------------------------------------------- > At DEXPO they announced a board that would speed up the CPU on > a VAX 11/780. Has anyone used it or can offer any comments on > this board ? Is the performance improvement greater than that which you get from a 785 upgrade? The 785 upgrades are running about $35,000 on the used market and have the advantage of being DEC-supported. Of course, if theis alternative is faster, then I'll probably get one of those... ================================================================================ Note 142.2 Speed Up a VAX 11/780 ? 2 of 3 EISNER::NAGY "Frank J. Nagy, VAX Guru" 3 lines 30-MAY-1988 09:41 -------------------------------------------------------------------------------- I noticed this product in a short announcement article (or ad) in one of the industry rags. As I remember, the speedup claimed was on the order of 40% or approximately 1/2 of the 785 speedup. ================================================================================ Note 142.3 Speed Up a VAX 11/780 ? 3 of 3 EISNER::FULLER_S "Stuart Fuller" 3 lines 30-MAY-1988 12:58 -< Cache upgrade? >- -------------------------------------------------------------------------------- As I understand it, the speedup consists of basically an enlargement of the cpu's cache and translation buffer. The cpu still runs at the same speed, but doesn't have to refer to main memory as often.  ================================================================================ Note 143.0 Let's talk DECtalk 3 replies EISNER::BENDER "Joel Bender" 20 lines 1-JUN-1988 22:23 -------------------------------------------------------------------------------- ----- The following note I moved from CONFERENCE_IDEAS. Thanks Terry, your right, a whole conference would be overkill. ----- I'm interested in talking with anyone who is using and/or is planning to use DECtalk. We current have a DTC01 (one line, stand alone) box on loan and we're planning on getting a multi-line "phone server" in the future. We (the Computer Section of Facilities Engineering, Cornell Univ.) have most of the EMCS systems in the major buildings on campus wired to a uVAX II and I would like to dial in and have it tell me the status of various building systems. Who else is using DECtalk? How do you manage "talking menus"? Do you use it in a "real-time" environment? Do you have any security measures? Ahh... here's a good question... Anybody using it in an AI application? ================================================================================ Note 143.1 Let's talk DECtalk 1 of 3 EISNER::RICE "Gary Rice" 10 lines 2-JUN-1988 14:09 -------------------------------------------------------------------------------- >Who else is using DECtalk? How do you manage "talking menus"? Do you use >it in a "real-time" environment? Do you have any security measures? You should contact Tom Hintz. (See the PC SIG Steering Committee list in the back of the Newsletters for phone & address). He has implemented an extensive series of DECtalk applications for doing automatic weather reporting for the local agricultural community in Florida. Gary ================================================================================ Note 143.2 Let's talk DECtalk 2 of 3 EISNER::KENNEDY "Terry Kennedy" 22 lines 2-JUN-1988 21:06 -------------------------------------------------------------------------------- >Who else is using DECtalk? How do you manage "talking menus"? Do you use >it in a "real-time" environment? Do you have any security measures? We've had a DTC01 here at the College for about 2 years. It was a 'novelty' purchase at the time. Currently it sits on my desk and is used to scare the occasional IBM rep who comes in. Seriously, we are beginning to look at a bunch of these for instructional use for the visually-impaired. The box has a huge amount of features, most of which we haven't explored fully. The supplied DECUS C examples were rapidly discarded and I have writtem a new interface library (but only for the functions I currently use). I've never hooked it up to a phone line, so I can't say how well that part works. In attempting to correct the unit's speech, I have overflowed the user dictionary several times. For example, it pronounces 'golf' as rhyming with 'bowl-f'. Of course, I have Version 2 ROM's - is your box newer? (it says the version when it turns on). Lastly, if you don't give it anything to say for a few weeks, it gets bored and turns on all the LED's on the back and refuses to talk. Again, this may be an older version. ================================================================================ Note 143.3 Let's talk DECtalk 3 of 3 EISNER::RENES "Stuart Renes, DFWLug Chair" 4 lines 3-JUN-1988 16:44 -< VOICE abstract >- -------------------------------------------------------------------------------- See the 1987 Fall Proceedings (DECUS), pp. 115-120 for an excellent article on VOICE, a computer controlled telephone information system that uses DECtalk at Cal Poly.  ================================================================================ Note 144.0 18-BIT VS 22-BIT QBUS DEVICES 13 replies EISNER::KILLEEN "Jeff Killeen" 12 lines 4-JUN-1988 12:22 -------------------------------------------------------------------------------- I just pulled apart a PDP-11/03 system. The question is how can I tell if a board will work on a 11/53/73/83 system. I have heard that some PDP-11/03 boards are 18-bit addressing only. For example.... DLV11 (M7940) RXV11 (M7946) REV11 (M9400-YA) After I change the 9270-A to a 9270-Q backplane Will they work on a 22-bit system? ================================================================================ Note 144.1 18-BIT VS 22-BIT QBUS DEVICES 1 of 13 EISNER::NAGY "Frank J. Nagy, VAX Guru" 8 lines 4-JUN-1988 12:44 -< 18/22-bit only issue for DMA devices >- -------------------------------------------------------------------------------- As I remember, the 18-bit vs 22-bit QBus issue only affects DMA devices. Thus the DLV11 will work in the Q22 systems since its a programmed-I/O device. The RXV11, however, is a DMA device - however, didn't RSX (at least) support the RXV11 on the Q22 bus by setting up a common region in the first 256K of memory? The driver would do DMA to/from buffers in this common region and then do memory-to-memory copies between the user buffers and the common region. I have no idea about the REV11, what in the work is it? ================================================================================ Note 144.2 18-BIT VS 22-BIT QBUS DEVICES 2 of 13 EISNER::KILLEEN "Jeff Killeen" 4 lines 4-JUN-1988 13:11 -< THE ANSWER IS >- -------------------------------------------------------------------------------- > what in the world is it? The bootstrap/terminator ================================================================================ Note 144.3 18-BIT VS 22-BIT QBUS DEVICES 3 of 13 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 53 lines 4-JUN-1988 14:17 -< Slight confusion >- -------------------------------------------------------------------------------- >> The question is how can I tell >> if a board will work on a 11/53/73/83 system. I have heard that some >> PDP-11/03 boards are 18-bit addressing only. For example.... >> >> DLV11 (M7940) >> RXV11 (M7946) >> REV11 (M9400-YA) After I change the 9270-A to a 9270-Q backplane >> >> Will they work on a 22-bit system? || As I remember, the 18-bit vs 22-bit QBus issue only affects DMA devices. || Thus the DLV11 will work in the Q22 systems since it's a programmed-I/O || device. Absolutely correct. || The RXV11, however, is a DMA device - however, didn't RSX (at || least) support the RXV11 on the Q22 bus by setting up a common region || in the first 256K of memory? The driver would do DMA to/from buffers || in this common region and then do memory-to-memory copies between the || user buffers and the common region. You are confusing the RXV11 (M7946) with the RXV21 (M8029). The RXV11 is the non-DMA controller for the RX01. The RXV21 is a DMA controller for the RX02. Thus, the RXV11 will work just fine in a 22-bit backplane. You are right about RSX (and RT-11) with the RXV21. || I have no idea about the REV11, what in the world is it? The REV11 is a terminator, bootstrap and DMA refresh board. It cannot be used in a 22-bit system because the termination is 18-bit only. For those of you with short memories, the REV11 DMA refresh capability was for refreshing memory boards that did not have an on-board refresh capability. The LSI-11 had a jumper option that allowed the microcode to perform the refresh function, but at a decrease in CPU speed. The REV11 took over this function by doing DMA refresh cycles, allowing the CPU to execute instructions at full (such as it was) speed. The whole issue became moot with the demise of memory boards that needed external refresh. Only a few early models were ever built. The whole subject of Q22 compatible options is addressed quite will in Micronote #5 (23-Apr-84) by Charlie Giorgetti. It lists all processors, backplanes, memories, controllers and bus-expansion cables that were available at that time. For options that have "restricted compatibility", it tells *why* they are restricted (as I did above for the REV11). In the next note, I will extract the list of "restricted compatibility" options. Alan ================================================================================ Note 144.4 18-BIT VS 22-BIT QBUS DEVICES 4 of 13 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 173 lines 4-JUN-1988 14:18 -< Q22 bus restricted options >- -------------------------------------------------------------------------------- This note is an extract from DEC's Micronote #5, "Q22 Compatible Options" by Charlie Giorgetti, dated 23-Apr-84. Please refer to the original Micronote for complete information, as this note does not list the "fully compatible" options. Alan ---------------------------------------------------------------- A Q22 compatible option is defined as a Q-bus option that will work without restriction in an extended Q-bus system, that is a 22-bit Q-bus system. The list below is of options that are not compatible in Q22 systems and the reason for the restriction. The requirements for a device to be Q22 compatible are the following: 1) Processors, memories, and DMA devices must all be capable of 22-bit addressing. 2) Devices must use backplane pins BC1, BD1, BE1, BF1 and DC1, DD1, DE1, DF1 for BDAL18-21 only. Processors, memories or DMA devices which are not capable of 22-bit addressing may generate or decode erroneous addresses if they are used in systems which implement 22-bit addressing. Memory and DMA DMA devices which implement only 16 or 18-bit addressing may be used in a 22-bit backplane, but the size of the system memory must be restricted to the address range of those devices. NOTE 16 or 18-bit DMA devices can potentially work in Q22 systems by buffering I/O in the 16- or 18-bit address space. Any device which uses the backplane pins list above for purposes other than BDAL18-21 is electrically incompatible with the Q22 bus and may not be used without modification. FULLY COMPATIBLE OPTIONS: (See original Micronote) RESTRICTED COMPATIBILITY OPTIONS: PROCESSORS: KDF11-A M8186 LSI-11/23 CPU Prior to etch rev. C, 18-bit addressing only, and use of BC1, BD1, BE1, BF1 for purposes other than BDAL18-21. KD11-HA M7270 LSI-11/2 CPU 16-bit addressing only, and use of BC1, BD1, BE1, BF1 for purposes other than BDAL18-21. KD11-F M7264 LSI-11 CPU 16-bit addressing only, and use of BC1, BD1, BE1, BF1 for purposes other than BDAL18-21. KXT11-A M8063 SBC-11/21 CPU 16-bit addressing only. BACKPLANES/BOXES: DDV11-B 6 x 9 backplane 18-bit addressing only BA11-M H9270 4 x 4 backplane 18-bit addressing only BA11-N H9273-A 4 x 9 backplane 18-bit addressing only BA11-VA H9281-A,B,C 2 x n dual-height backplane (n = 4, 8, 12) BA11-VA used the H9281-A 18-bit addressing only VT103 4 x 4 backplane 18-bit addressing only NOTE It is usually possible for the adventurous hacker to add the extra backplane wires to convert the above backplanes to 22-bit addressing. -- Alan MEMORIES: MMV11-A G653 8 KB core memory 16-bit addressing, Q-bus required on C/D backplane connectors MRV11-AA M7942 ROM module 16-bit addressing only MRV11-BA M8021 UV PROM-RAM 16-bit addressing only MRV11-C M8048 PROM/ROM module 18-bit addressing only MSV11-B M7944 8 KB bus refreshed RAM 16-bit addressing only MVS11-C M7955 32 KB RAM 18-bit addressing only MSV11-D,E M8044/M8045 8, 16, 32, 64 KB RAM 18-bit addressing only MXV11-A M8047 Multifunction module 18-bit addressing only on memory, the memory can be disabled to use just the serial lines OPTIONS: AAV11 A6001 D/A converter Use of BC1 for purposes other than BDAL18 ADV11 A012 A/D converter Use of BC1 for purposes other than BDAL18 BDV11 M8012 Bootstrap/Terminator CS Rev. E or earlier 18 bits only DLV11-J M8043 Serial line interface CS Rev. E or earlier incompatible with KDF11-A and KDF11-B [Don't ask me why, I'm just reporting it -- Alan] DRV11-B M7950 General purpose DMA interface (quad) 18-bit DMA addressing only KPV11-B,C M8016-YB,YC Power-fail/line-time clock/terminator Termination for 18 bits only KUV11 M8018 Writable Control Store For use with KD11-F processor only KWV11-A M7972 Programmable real-time clock Use of BC1 for purposes other than BDAL18 REV11 M9400 Terminator, DMA refresh, bootstrap Bootstrap for use with KD11-F and KD11-HA processors only. Termination for 18-bits only. DMA refresh may be used in any system. RKV11-D M7269 RK05 controller interface 16-bit DMA addressing only. [Can be "hot-wired" for 18-bit DMA. See me for details. -- Alan] RLV11 M8013 RL01/2 controller interface M8014 18-bit DMA only, use of BC1 and BD1 for purposes other than BDAL18 and BDAL19. Requires two adjacent slots with C/D interconnect. RXV21 M8029 RX02 floppy disk interface 18-bit DMA addressing only TEV11 M9400-YB 120-Ohm bus terminator Termination for 18 bits only VSV11 M7064 Graphics display 18-bit DMA addressing only BUS CABLE-CARDS: M9400-YD Cable connector 18-bit bus only M9400-YE Cable connector with 240-Ohm termination 18-bit bus only M9401 Cable connector 18-bit bus only ================================================================================ Note 144.5 18-BIT VS 22-BIT QBUS DEVICES 5 of 13 EISNER::KILLEEN "Jeff Killeen" 11 lines 4-JUN-1988 14:50 -< QUESTIONS - MORE QUESTIONS >- -------------------------------------------------------------------------------- > The REV11 is a terminator, bootstrap and DMA refresh board. > It cannot be used in a 22-bit system because the termination is > 18-bit only. For those of you with short memories, the REV11 Thanks Alan - you saved some hardware from the trash bin. However this now raises a question - if I replace the H9270-A backplanes in the BA11-M boxes I have with H9270-Q backplanes how do I terminate them? (Since the REV11 or TEV11 are 18-bit only). When DEC re-did the H9270 did they put termination in it? ================================================================================ Note 144.6 18-BIT VS 22-BIT QBUS DEVICES 6 of 13 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 53 lines 5-JUN-1988 13:20 -< The answer is in the Micronotes >- -------------------------------------------------------------------------------- >> However this now raises a question - if I replace the H9270-A backplanes >> in the BA11-M boxes I have with H9270-Q backplanes how do I terminate >> them? (Since the REV11 or TEV11 are 18-bit only). When DEC re-did the >> H9270 did they put termination in it? To the best of my knowledge, none of the H9270-x backplanes have termination. However, the only one I have handy is the original H9270 (no suffix). In such a short backplane, the termination on the CPU card should be sufficient. This subject is addressed quite thoroughly in DEC's Micronote #35 by Jack Toto (19-Jul-85), "Backplane Expansion/Termination". To answer your question directly, I will quote from the section on Single Backplane Termination Rules: 1. Then using a processor with 240 ohmn termination, the bus can accomodate up to 20 ac loads (total) before additional termination is required. If more than 20 ac loads are included, the far end of the bus must be terminated with 120 ohms, although termination of 240 ohms is optimum. Following the addition of at least the minimum termination, up to 35 ac loads may be present in a single backplane. 2. Then using a processor with 120 ohmn termination, the bus can accomodate up to 35 ac loads (total) before additional termination is required. If more than 35 ac loads are included, the far end of the bus must be terminated with 120 ohms. When this is done, up to 45 ac loads may be present. 3. In all cases, the bus can accomodate up to 20 (total) dc loads. The processor terminations are: KDF11-A 240 ohms 11/23 dual width KDF11-B 120 ohms 11/23+ quad width KDJ11-A 240 ohms 11/73 dual width KDJ11-B 120 ohms 11/73 quad width MicroVAX I 240 ohms MicroVAX II 240 ohms For many years I ran a system in a H9270 backplane with no additional termination. There were no problems. By the way, I hope that you realize that an H9270-A backplane can be upgraded to an H9270-Q by adding the extra four wires to all the slots. Alan ================================================================================ Note 144.7 18-BIT VS 22-BIT QBUS DEVICES 7 of 13 EISNER::KILLEEN "Jeff Killeen" 14 lines 5-JUN-1988 13:30 -< AIN'T DECUServe GREAT! >- -------------------------------------------------------------------------------- > For many years I ran a system in a H9270 backplane with > no additional termination. There were no problems. Thanks again Alan! I shall now build my portable PDP-11/73 (BA11-M & RD54-DA). > By the way, I hope that you realize that an H9270-A backplane > can be upgraded to an H9270-Q by adding the extra four wires > to all the slots. For the 32 wire wraps (4X4X2) I will spring for the H9270-Q and keep the old one as "spare". (about $175) THANKS AGAIN ALAN! ================================================================================ Note 144.8 18-BIT VS 22-BIT QBUS DEVICES 8 of 13 EISNER::KENNEDY "Terry Kennedy" 11 lines 5-JUN-1988 16:59 -< Wrapping is faster! >- -------------------------------------------------------------------------------- > For the 32 wire wraps (4X4X2) I will spring for the H9270-Q and keep the > old one as "spare". (about $175) Total effort is about 10-15 minutes to do the wrapping, compared with at least an equal time wrestling the backplane in & out. Not to mention that you can (usually) find wire-wrap wire a lot faster than a backplane! P.S. - If you do wrap it, do it as two vertical rows (A/B side, C/D side), and then connect them, rather than ABCD-ABCD-ABCD... This will reduce the noise. I've done this and it runs fine at 18Mhz - although the J11-RQDX2 combo makes the 11/03 power supply scream! ================================================================================ Note 144.9 18-BIT VS 22-BIT QBUS DEVICES 9 of 13 EISNER::KILLEEN "Jeff Killeen" 12 lines 5-JUN-1988 17:25 -< WHO INVESTS IN WIRE WRAP TOOLS ANY MORE? >- -------------------------------------------------------------------------------- > Total effort is about 10-15 minutes to do the wrapping, compared with at > least an equal time wrestling the backplane in & out. Not to mention that > you can (usually) find wire-wrap wire a lot faster than a backplane! Not if you don't have good wire wrap tools! > I've done this and it runs fine at 18Mhz - although the J11-RQDX2 > combo makes the 11/03 power supply scream! Are you telling me the 11/03 power supply is marginal? I didn't even think about this since it weighs more than a BA23 power supply. ================================================================================ Note 144.10 18-BIT VS 22-BIT QBUS DEVICES 10 of 13 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 10 lines 5-JUN-1988 22:38 -< Less power to you :-) >- -------------------------------------------------------------------------------- >> Are you telling me the 11/03 power supply is marginal? I didn't even >> think about this since it weighs more than a BA23 power supply. If you are referring to the BA11-M box (H780A power supply): Output Power (Combinations not to exceed 110 watts): +5 1.5 to 18.0 amps +12 0.25 to 3.5 amps Alan ================================================================================ Note 144.11 18-BIT VS 22-BIT QBUS DEVICES 11 of 13 EISNER::KOZAM 21 lines 5-JUN-1988 22:53 -< H780 (BA11-M) power rating and "screaming" >- -------------------------------------------------------------------------------- >Are you telling me the 11/03 power supply is marginal? I didn't even >think about this since it weighs more than a BA23 power supply. A quick comparison between the H7864 (BA23) power supply and H780 (BA11-M) is as follows. The H786 (BA11-N) is very similar to the H780. H780 H7864 +5 Volts 0-18A 36A (typical) +12 Volts 0-3.5A 7A (typical) As you can see, pound for pound, the H7864 is a much better deal. At the same time, I've never had difficulties with any of the power supplies being unable to deliver enough amps for normal use. (If you had intended to power an RD5X drive with the H780, then you'd be pushing things.) I think that the "screaming" which was referred to has to do with the design of the power supplies. "Switching" power supplies, such as these often emit a high-pitched tone, which is very annoying. I've only noticed this with the H780, especially when a heavy current is drawn. Marc Kozam ================================================================================ Note 144.12 18-BIT VS 22-BIT QBUS DEVICES 12 of 13 EISNER::KENNEDY "Terry Kennedy" 21 lines 6-JUN-1988 00:10 -< Power supply *can* be a problem >- -------------------------------------------------------------------------------- > Who has wire-wrap tools any more? You mean you remove the Unibus NPR jumpers with your teeth? 8-) > As you can see, pound for pound, the H7864 is a much better deal. Yep - the 780 can (according to specs) run *either* a J-11 *or* an RQDXn, but not both. Add memory and terminal I/O and it gets worse... However, it has worked for me - but you may not want to play 'you bet your boards'. > I think that the "screaming" which was referred to has to do with > the design of the power supplies. "Switching" power supplies, such as > these often emit a high-pitched tone, which is very annoying. I've only > noticed this with the H780, especially when a heavy current is drawn. The 780 has a design such that as the load increases, the switching frequency drops, lowering efficiency, causing increased current draw. Once you get past the knee, the supply *will* die. The BA23's supply has the sense to either pop the circuit breaker or burn the cable har- ness (sorry, I couldn't resist...). ================================================================================ Note 144.13 18-BIT VS 22-BIT QBUS DEVICES 13 of 13 EISNER::CETRON 8 lines 11-JUN-1988 14:26 -------------------------------------------------------------------------------- Last I recall Radio Sclock sold wire-wrap tools for 10$.....and it took me exactly 25.6 minutes to do our backplane (yes I time it, yes I won the bet, and no the dollars weren't enough to make me rich, but yes it did impress the hell out of the client :-) ) -ed  ================================================================================ Note 145.0 MICRO NOTES 2 replies EISNER::KILLEEN "Jeff Killeen" 4 lines 5-JUN-1988 15:36 -------------------------------------------------------------------------------- Is there anyway to order MicroNotes - including the old ones? I know of late they have stopped publishing them. ================================================================================ Note 145.1 MICRO NOTES 1 of 2 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 40 lines 5-JUN-1988 23:07 -< More on MicroNotes >- -------------------------------------------------------------------------------- >> Is there anyway to order MicroNotes - including the old ones? >> I know of late they have stopped publishing them. As far as I know, they are not available from DEC at this time. The SASE (Systems And Support Engineering) group is thinking of starting them again, so you might try writing to: Robert Cansler Principle Engineer - Micro-Systems Engineering Support Unit Digital Equipment Corporation 146 Main Street Maynard, MA 01754 As far as getting a copy of the existing ones, there is *no* copyright notice on any of the three editions that I have: Red Cover, 85 notes, 253 pages, 1-Apr-77 to 29-Oct-79 Yellow Cover, 21 notes, ~162 pages, 13-Apr-82 to 10-Jan-85 Blue Cover, 41 notes, 400 pages, 13-Apr-82 to 17-Sep-85. The Yellow edition is a proper subset of the Blue edition. There were also 30 to 40 additional notes that were issued between 1979 and 1982, some of which I have. I remember seeing note numbers up near 120. Effective with the Yellow edition, they started the numbering over at one. I believe that there was an earlier edition (with a red cover?) and, I suspect, another one between my Red and Yellow editions, but I have no information on them. Any clues from the audience? Due to the volume of pages, I would be reluctant to make copies unless there is no way that DEC will make them available. I think the best thing to do is write to Robert Cansler and ask for them. Alan ================================================================================ Note 145.2 MICRO NOTES 2 of 2 EISNER::KENNEDY "Terry Kennedy" 1 line 6-JUN-1988 00:13 -< They sent me a set about 7 months ago! >- -------------------------------------------------------------------------------- > As far as I know, they are not available from DEC at this time.  ================================================================================ Note 146.0 CD-ROM questions 2 replies EISNER::HASSINGER "Bob Hassinger" 25 lines 6-JUN-1988 15:00 -------------------------------------------------------------------------------- I am interested in information on DEC's CD-ROM product(s). There is some interest here in getting a CD-ROM device to access various CDs from a variety of sources. First question: Are things standard enough so you can count on accessing any of the CD-ROMs out there via DEC's PC version of the RRD50? How about from the microVAX/Qbus version (i.e. VMS access software is one issue I should think). Does anyone know if there would be a chance we could access the VMS distributions on CD-ROM via the PC version or if a VAX/VMS version will be necessary? Second: I just looked on the Electronic Store and in DECdirect and found the price of a reader plus interface for either the PC or the microVAX is running in the $1350 range but the Electronic Store is currently listing what seems to be the PC version for $495 (RRD50-EA)! Am I reading these right? If so what is going on with the PC version? There seem to be hints that a new product or products are in the wind. The low price for one and the obvious need that VMS software availability will generate for Qbus, Unibus and BI systems for another. Any clues about this or advice on how good an idea it would be to buy one of these now rather than wait a while? Bob H ================================================================================ Note 146.1 CD-ROM questions 1 of 2 EISNER::MAYHEW "Bill Mayhew" 25 lines 6-JUN-1988 15:49 -< Pssst! Has Digital got a deal for you! >- -------------------------------------------------------------------------------- < Note 146.0 by EISNER::HASSINGER "Bob Hassinger" > > I just looked on the Electronic Store and in DECdirect and > found the price of a reader plus interface for either the PC or the > microVAX is running in the $1350 range but the Electronic Store is > currently listing what seems to be the PC version for $495 (RRD50-EA)! Yup, you are reading those correctly. The PC-version price is a special sale. You may have noticed it was on sale even cheaper to DECUS attendees. For the first time I'm aware of, most of the "DECUS Special Sale" items are also available at sale prices through DECdirect or the E-store (though the prices are somewhat higher than the DECUS Special prices). I believe the sale runs out in a week or so -- June 19 pops to mind, and makes sense since I think the Estore's database is updated weekly, on Sundays. For an even bigger shock, have a look at the RRD50-AA, which professes to be the bare drive plus US country kit. It's priced at just $110 (not MY typo!), though the bare-drive itself (RRD50-A2) is $1050. Unfortunately I have not yet found a source for the Q-bus controller as a separate item; I have its Mxxxx number around somewhere but it is not a listed part and I have not geared myself up for a go-round with the PDQ/Unlisted Parts Service. ================================================================================ Note 146.2 CD-ROM questions 2 of 2 EISNER::HAHN 11 lines 6-JUN-1988 18:06 -< CD-ROM/mVAX/PC >- -------------------------------------------------------------------------------- I have the CD-ROM attached to my mVAX. I also have two of the 3 Library CD disks. It is very easy to pull anything from the CD ROM and copy it to my main disk, In order to save time when doing searches for titles files I have extracted a complete directory to my main disk. I had a long talk at Cinc. on the ability to "read" the new V5 Doc set when it comes out using a PC. If I run windows (under PCSA) then I should be able to use the DOC set withou having a work station. Pierre  ================================================================================ Note 147.0 RL02 PARTS 5 replies EISNER::KILLEEN "Jeff Killeen" 2 lines 6-JUN-1988 20:16 -------------------------------------------------------------------------------- Does anyone have any ideas on how to get light bulbs for RL02's ? ================================================================================ Note 147.1 RL02 PARTS 1 of 5 EISNER::KENNEDY "Terry Kennedy" 9 lines 7-JUN-1988 00:51 -< Send me MAIL >- -------------------------------------------------------------------------------- > Does anyone have any ideas on how to get light bulbs for RL02's ? Well, you call Unlisted Parts, find out you have to order 144 of them, and that the lead time is 120 days... Or, send me a mail message stating how many of the 143 I'm not using you want... P.S. - RL02 and RK07 bulbs are *not* the same! ================================================================================ Note 147.2 RL02 PARTS 2 of 5 EISNER::KOZAM 7 lines 7-JUN-1988 02:18 -< RL02 Bulbs == GE 73 Type >- -------------------------------------------------------------------------------- > Does anyone have any ideas on how to get light bulbs for RL02's ? RL02 bulbs are all GE 73 type (yes, that's General Electric). Not too long ago, I purchased a package of 10 from a local electrical supplier. The cost was $ 3.00 for the package. I don't recall if it was special order. Marc Kozam ================================================================================ Note 147.3 RL02 PARTS 3 of 5 EISNER::HAHN 5 lines 7-JUN-1988 09:00 -< get higher voltage bulb >- -------------------------------------------------------------------------------- I have now forgoten where after blowing 5 bulbs we purchased what I believe are higher voltage bulbs (they are dimmer) to put into RL01/2. It was a radio store. If you want I will pull out one of them and look. Pierre ================================================================================ Note 147.4 RL02 PARTS 4 of 5 EISNER::KENNEDY "Terry Kennedy" 5 lines 7-JUN-1988 18:52 -< Careful! >- -------------------------------------------------------------------------------- > ... we purchased what I believe are higher voltage bulbs... Careful - there are two types of drivers used for the RL02 front panel lamps. Your lamps probably are similar to those used in the RK07 and will blow the other driver type used in the RL02. ================================================================================ Note 147.5 RL02 PARTS 5 of 5 EISNER::CETRON 8 lines 11-JUN-1988 14:32 -< I see the light... >- -------------------------------------------------------------------------------- My field service rep will supply me with as many as I need - and usually many more - he HATES to come all the way just to replace a 'bloody' light bulb.... I think I have 40 spares... Do you want one Jeff.... -ed  ================================================================================ Note 148.0 Moderation of this conference 5 replies EISNER::PROVOST 3 lines 8-JUN-1988 13:08 -------------------------------------------------------------------------------- This TOPIC is for discussion of MODERATION of this CONFERENCE. Reply with your suggestions, complaints, etc. ================================================================================ Note 148.1 Moderation of this conference 1 of 5 EISNER::PROVOST 9 lines 8-JUN-1988 13:11 -< My initial contributions >- -------------------------------------------------------------------------------- I have entered DIR * as a table of contents in a reply to TOPIC 4.0. I have *'d the TOPICS which I believe should be on-going forever. I attempted to deal with DIR *.*, but quickly blew my quota. I welcome your suggestions as to what can be done to make this conference more efficiently valuable to you. Tom Provost ================================================================================ Note 148.2 Moderation of this conference 2 of 5 EISNER::MAYHEW "Bill Mayhew" 5 lines 8-JUN-1988 14:03 -< A couple moderation questions >- -------------------------------------------------------------------------------- What's the advantage of having a note that contains the table of contents, as opposed to just doing a DIR oneself? I haven't yet explored the keywording feature of VAX Notes, but this conference would seem to me like a natural for some keyword access scheme based on the product part-number (LN03, etc.) ================================================================================ Note 148.3 Moderation of this conference 3 of 5 EISNER::HOPKINS "Al Hopkins" 24 lines 8-JUN-1988 19:45 -< Keywords, KWIK, and Personal Magazines? >- -------------------------------------------------------------------------------- > I haven't yet explored the keywording feature of VAX Notes, but this > conference would seem to me like a natural for some keyword access > scheme based on the product part-number (LN03, etc.) While Tom's concept of identifying those Notes of lasting value is an interesting one, I concur with Bill. Dir or dir/title="search string" is sufficient assuming the author made the title topic significant. Because this is often *NOT* the case, I urge the use of KEYWORDS by the moderator(s). Many textual databases have a "Thesaurus" of accepted/approved KEYWORDS. While this *controls* the keywords it does not automatically enforce that all appropriate keywords are added nor does it address keyword personalization that an end user might suggest. While "MARK"ing is one mechanism for this, there is no easy method to "connect" the MARK with KEYWORDS. Also, while VTX has the concept of Personal Magazine to assist an end user in traversing many topics, there is no "DIRECTORY/MARK" or other such command to allow the end user to develop their own personalized approach to navigating their view of these conferences. ================================================================================ Note 148.4 Moderation of this conference 4 of 5 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 10 lines 2-FEB-1989 11:20 -< Needle in a Haystack >- -------------------------------------------------------------------------------- < I haven't yet explored the keywording feature of VAX Notes, but > < this conference would seem to me like a natural for some keyword > < access scheme based on the product part-number (LN03, etc.) > I agree. Is there any way that HARDWARE-HELP can be further divided into CPU_HELP, PRINTER_HELP, COMM_HELP, etc.? I'm trying to find info on a problem with an LN03, and it's tough to pick out among all the different topics; SEARCH is so slow I haven't been bothering with it. Some kind of subdivision or index scheme would be helpful. ================================================================================ Note 148.5 Moderation of this conference 5 of 5 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 45 lines 2-FEB-1989 12:16 -< XCOM WANTS TO KNOW - IS THIS A SOLUTION? >- -------------------------------------------------------------------------------- JOHN... Do these two examples solve your issue? EXAMPLE 1 $SEARCH DECUSERVE_TITLES LN03 -> yields.... DEC_SOFTWARE 12-JAN-1988 15 Digital font file format? (LN03) DESKTOP_PUBLISHING 3-NOV-1987 1 VAX SCRIPTPRINTER/LN03R DESKTOP_PUBLISHING 8-MAR-1988 1 Mixing TEK 4014 and text on LN03+ DESKTOP_PUBLISHING 25-MAR-1988 6 Do LN03+ and TeX mix? HARDWARE_HELP 12-FEB-1988 2 LN03 weirdness HARDWARE_HELP 25-MAR-1988 1 LN03 Toner Options HARDWARE_HELP 19-APR-1988 12 The LN03 OPC is dead -- Long live the LN03 OPC HARDWARE_HELP 27-APR-1988 1 LN03 AC/AD KITS HARDWARE_HELP 14-JUL-1988 0 Big Ben Increases Paper Capacity of LN03 HARDWARE_HELP 20-SEP-1988 2 LN03 saga HARDWARE_HELP 25-OCT-1988 0 LN03X-DE barcode cartridge for LN03 HARDWARE_HELP 9-JAN-1989 2 LN03r : software required? LATEST_RELEASE_INFOR 30-OCT-1987 4 LN03R/Scriptprinter (see 91.* for LPS40) OFFICE_AUTOMATION 23-NOV-1987 1 PRO/WPSPLUS --> LN03 PRO_300_SERIES 9-JUL-1987 0 P/OS LA75/LN03 Plus Update VMS 13-NOV-1988 6 FORM LIBRARY causing page ejects w/LN03Plus $SEARCH/MATCH=AND DECUSERVE_ALL_TITLES HARDWARE, LN03 -> yields.... HARDWARE_HELP 26-OCT-1987 15.12 Using an LN03 with an IBM PC/XT HARDWARE_HELP 1-AUG-1988 40.36 Bar codes for LN03 and LA75 printers HARDWARE_HELP 25-SEP-1987 57.2 LN03 and TeX font loading HARDWARE_HELP 28-SEP-1987 57.4 TeX for LN03 on L&T SIG tape HARDWARE_HELP 28-SEP-1987 57.6 Better LN03 Programmer Manual HARDWARE_HELP 28-SEP-1987 57.7 Well, that is a different story. ABOUT the LN03R... HARDWARE_HELP 28-SEP-1987 57.8 Sounds like expected bad news for the LN03 HARDWARE_HELP 27-JUL-1988 57.17 LN03 (basic) should print 66 lpp @ power-up HARDWARE_HELP 2-AUG-1988 57.26 Anyone with an LN03A-A2 who can provide info? HARDWARE_HELP 12-FEB-1988 2 LN03 weirdness HARDWARE_HELP 25-MAR-1988 1 LN03 Toner Options HARDWARE_HELP 19-APR-1988 12 The LN03 OPC is dead -- Long live the LN03 OPC HARDWARE_HELP 4-MAY-1988 121.7 The "half life" of the LN03 HARDWARE_HELP 27-APR-1988 1 LN03 AC/AD KITS HARDWARE_HELP 14-JUL-1988 0 Big Ben Increases Paper Capacity of LN03 HARDWARE_HELP 20-SEP-1988 2 LN03 saga HARDWARE_HELP 25-OCT-1988 0 LN03X-DE barcode cartridge for LN03 HARDWARE_HELP 9-JAN-1989 2 LN03r : software required?  ================================================================================ Note 149.0 System overheating detection and reaction 8 replies EISNER::LAMB "Charles P. Lamb" 13 lines 8-JUN-1988 19:36 -------------------------------------------------------------------------------- This topic is intended as a place to discuss relative merits of, and known problems with, various schemes for detection of and reaction to overheating (and possibly other environmental problems) of systems. Obviously, vendor names will be mentioned; please remember to keep the content of your notes technical and non-commercial. Many of us have unreliable cooling systems and must deal with the problem of the cooling system failing when no one is around to notice, like on weekends. The only commercial product I know of to gracefully bring down a VAX when overheating is detected is a hardware/software package made by IntraComputer. This seems to work quite well in supplying ample warning messages before the temperature becomes high enough to force it to shut down the system. However it costs circa $4K. I hope there are some other systems out there with capabilities at least as good as this at a lower cost. ================================================================================ Note 149.1 System overheating detection and reaction 1 of 8 EISNER::LEDERMAN "Bart Z. Lederman" 6 lines 8-JUN-1988 19:57 -< Try "Hardcopy"? >- -------------------------------------------------------------------------------- There is a home-made overtemperature detection system in one of the slick industry magazines (I belive "Hardcopy" this month) which looks nice for those who want to "roll their own". All I can tell you is that in our site all cabinets have thermostats wired to a central alarm panel: be we have a 24 hour manned operation. ================================================================================ Note 149.2 System overheating detection and reaction 2 of 8 EISNER::MAYHEW "Bill Mayhew" 22 lines 8-JUN-1988 21:35 -< Some other overtemp detection approaches >- -------------------------------------------------------------------------------- It seems to me that I read that one of Digital's fancy Power Distribution Systems had an option to help support this need. One of my customer sites (used to be my employer, where I developed this thing) had a different solution. The building environmental systems are controlled by a network of LSI-11s which are tied to (now) an 11/73 Ultrix host which is in the same computer room as the VAX. There's a 24-hour guard located right outside that room, who is tied into a security monitoring computer which is also tied into the 11/73. It's a simple thing for the guard to be notified automagically as the temperature rises. Now that the VAXen at the site have TCP/IP, it should be not too tough to have messages sent to any/all VAX users, too -- and to, for example, initiate a batch job or equivalent that shuts down the system. $4K sounds like an expensive solution. I can think of some simpler ones, like a garden-variety thermostat set up to connect Data Terminal Ready to Data Set Ready or Carrier Detect on a "spare" serial port -- thereby looking like an "incoming phone call" to a program that has allocated that port and is watching it. When the "call" comes in, it alerts everyone, and does whatever else you want to do that's fancy. Total cost would be well under $50 plus an interesting program and a spare port. ================================================================================ Note 149.3 System overheating detection and reaction 3 of 8 EISNER::MERRIMAN "George Merriman - CCA/NY" 6 lines 8-JUN-1988 23:16 -< See DEC Professional for example >- -------------------------------------------------------------------------------- > $4K sounds like an expensive solution. I can think of some simpler > ones, like a garden-variety thermostat set up to connect Data Terminal > Ready to Data Set Ready or Carrier Detect on a "spare" serial port This is exactly the solution described in the recent DEC Professional, which is probably the article refered to in .1. ================================================================================ Note 149.4 System overheating detection and reaction 4 of 8 EISNER::COY "Dale E. Coy" 22 lines 12-JUN-1988 20:52 -< You get what you pay for >- -------------------------------------------------------------------------------- We're using a box called SAM (I think that's the one referred to earlier). We bought the 1-system box, which was much less than the $4K quoted. It monitors under-floor temp, above-floor temp, and humidity. As an Electrical Engineer, I was impressed with the hardware design. As a System Manager, I was impressed with the .COM files provided. As a System Analyst, I was impressed with the overall quality of the thinking and system design. At a controllable temp, it starts sending msgs to OPER. At a higher temp, it sends messages to all users to "Tell the System Manager". At a higher temp, it does @SHUTDOWN and then pulls the plug (trips the PDS off, in our case). If all else fails, the hardware pulls the plug. All temperature thresholds are settable via RS-232 port on the box. It's seldom that I see a truly nice, elegant hardware/software solution to a problem like this. In my opinion, it's worth the price. The DEC PRO solution is a nice "do it yourself" approach, and you could certainly improve on it. But how much "tinkering" can you really do for $2K (or $4K for their "big box)? ================================================================================ Note 149.5 System overheating detection and reaction 5 of 8 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 12 lines 15-JUN-1988 00:57 -< Brownout protection >- -------------------------------------------------------------------------------- >> This topic is intended as a place to discuss relative merits of, and >> known problems with, various schemes for detection of and reaction to >> overheating (and possibly other environmental problems) of systems. I would like to see something that, in addition to the over-temperature shutdown, would shut down the system if the line voltage drops below some minimum for more than some minimum time. While Los Angeles does not usually have brownout problems, we did suffer one when a 5KV line dropped nearby but didn't trip the circuit breaker. Alan ================================================================================ Note 149.6 System overheating detection and reaction 6 of 8 EISNER::COY "Dale E. Coy" 3 lines 15-JUN-1988 19:10 -< A PDS would probably do that >- -------------------------------------------------------------------------------- Many PDS (Power Distribution Systems) have the capability to do shutdowns (trip the breaker - not gently) on overvoltage and undervoltage, as well as phase-reversals, etc. ================================================================================ Note 149.7 System overheating detection and reaction 7 of 8 EISNER::CLAPP 23 lines 16-JUN-1988 11:13 -< $200 partial solution >- -------------------------------------------------------------------------------- > However it costs circa $4K. I hope there are some other systems > out there with capabilities at least as good as this at a lower cost. About a year ago, we bought several "DUOfone Home Monitor" units from Radio Shack (part 43-165) on sale for less than $200 each. They plug into an AC outlet and a phone line and can be setup to monitor several parameters. If a threshold is exceeded, they cycle through a list of (up to) four phone numbers - dialing them and using a voice synthesizer to announce the nature of the problem. It takes more than just answering one of the listed phones to clear the alarm - you actually have to ring the unit back. Parameters monitored: Temperature - single sensor on unit, programmable trip points AC power failure - if it persists beyond 5 minutes (unit has batteries) 3 sets of dry sensor contacts - configurable for NO or NC contacts Microphone - Several seconds of LOUD noise will trigger. They hedge on this, but you can try to put it right next to a smoke detector. I'm not sure if this is in the current catalog. I have seen a similar unit from another vendor recently - don't remember details. It seems to be well thought out and good value for the price. If you have anyone near a phone who can respond fairly quickly, it is much better than nothing. ================================================================================ Note 149.8 System overheating detection and reaction 8 of 8 EISNER::TIHOR "By day System Manager, by night wily " 4 lines 26-JUN-1989 22:22 -< SAM on a network >- -------------------------------------------------------------------------------- We also use a SAM. I have written an application that serves out the temperature, power, flood, smoke, and humidity data over the network via TCP/IP to service our UNIX systems (the VMS systems are remotely controled via DECnet.)  ================================================================================ Note 150.0 Will a MicroVAX II DRV11-WA work in s MicroVAX 3600? 6 replies EISNER::TILLMAN "Brian Tillman, Smiths Industries." 6 lines 16-JUN-1988 16:34 -------------------------------------------------------------------------------- Hopefully, this is a simple question with a simple answer... Does anyone know *for sure* if a DRV11-WA from a MicroVAX II can be used in a MicroVAX 3600? We've been told the Q-BUS in the 3600 is different from the Q-BUS in the II and the DRV won't play. We want to verify this. ================================================================================ Note 150.1 Will a MicroVAX II DRV11-WA work in s MicroVAX 3600? 1 of 6 EISNER::RICE "Gary Rice" 15 lines 16-JUN-1988 22:39 -------------------------------------------------------------------------------- > Does anyone know *for sure* if a DRV11-WA from a MicroVAX II can > be used in a MicroVAX 3600? We've been told the Q-BUS in the 3600... In a word: NO. According to a developer at Vector General (who makes graphics boards and tubes for 11/70's and Microvaxes, the Q-bus has 2 problems in the 3000 series boxes: 1) The re-design of the "top" of the board 2) The bus has a different interleaving scheme (that is, you can't just plug the board in any slot...) Gary ================================================================================ Note 150.2 Will a MicroVAX II DRV11-WA work in s MicroVAX 3600? 2 of 6 EISNER::NORTON "Bill Norton" 4 lines 17-JUN-1988 10:19 -< Should work OK... >- -------------------------------------------------------------------------------- If you're not worried about FCC compliance and aesthetics of the enclosure, and plug the card into the QBUS half of the slot, it's likely to work just fine. I've seen an original QBUS RRD50 plugged into a 3600 with no problem. ================================================================================ Note 150.3 Will a MicroVAX II DRV11-WA work in s MicroVAX 3600? 3 of 6 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 47 lines 17-JUN-1988 12:50 -< A DRV11-WA will work, but with restrictions >- -------------------------------------------------------------------------------- >> Does anyone know *for sure* if a DRV11-WA from a MicroVAX II can >> be used in a MicroVAX 3600? Well, yes and no. There are several issues. Let's look at them. || 1) The re-design of the "top" of the board This one is purely physical. The DRV11-WA is a dual-height board, while the BA213 cabinet is designed for quad-height boards with the new 0.95" bracket at the handle end. You can either use the board as-is (messing up the air flow and looking ugly), or you could get one of DEC's adapter kits (see the Electronic Store for info) that is designed just for that application. >> We've been told the Q-BUS in the 3600 >> is different from the Q-BUS in the II and the DRV won't play. We >> want to verify this. || 2) The bus has a different interleaving scheme (that is, || you can't just plug the board in any slot...) The Q-Bus is the same Q-Bus we have come to know and love. The difference is the backplane layout. Some backplanes run the bus in a "serpentine" fashion down the backplane and others (BA23 and BA123) do so after dedicating the first few slots to the C/D interconnect. The BA213 backplane runs the Q-Bus straight down the A/B slots and the C/D interconnect straight down the C/D slots. There is another note in this conference that explains this in detail. HOWEVER, you are not home free yet! The problem with the DRV11-WA is that it has a mode of operation that allows the external device to do Read/Modify/Write operations to memory locations. The MicroVAX 3xxx CPU DOES NOT ALLOW Q-BUS DEVICES TO DO R/M/W OPERATIONS TO MEMORY. There is a rumor that a future version of the CPU board *will* allow this, but has not been announced yet. From the above, it appears that you CAN use your DRV11-WA with the MicroVAX 3500/3600 as long as you don't use R/M/W cycles and can take care of the physical mounting problems. If you need more information, I will try to get it, but this is the essential part of it. Alan ================================================================================ Note 150.4 Will a MicroVAX II DRV11-WA work in s MicroVAX 3600? 4 of 6 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 5 lines 17-JUN-1988 13:58 -< THESE ARE MOST LIKELY DUMB QUESTIONS >- -------------------------------------------------------------------------------- > locations. The MicroVAX 3xxx CPU DOES NOT ALLOW Q-BUS DEVICES > TO DO R/M/W OPERATIONS TO MEMORY. There is a rumor that a Isn't this what DMA is? If so how does DEC's DMA boards work? ================================================================================ Note 150.5 Will a MicroVAX II DRV11-WA work in s MicroVAX 3600? 5 of 6 EISNER::KILGALLEN "Larry Kilgallen" 13 lines 17-JUN-1988 14:43 -< A Software Description of Read-Modify-Write >- -------------------------------------------------------------------------------- When hardware does the equivalent of: MOVL src,memory-address or MOVL memory-address,dst that counts as DMA. The particular type of prohibited DMA (read- modify-write) is when hardware does the equivalent of: BISL2 src,memory-address such that the operation requires the initial contents of memory AND changes those contents. ================================================================================ Note 150.6 Will a MicroVAX II DRV11-WA work in s MicroVAX 3600? 6 of 6 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 0 lines 17-JUN-1988 14:51 -< UNDERSTAND - THANKS >- --------------------------------------------------------------------------------  ================================================================================ Note 151.0 Fix for Able Quadrasync/E 1 reply EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 30 lines 17-JUN-1988 15:05 -------------------------------------------------------------------------------- This note is copied to here from the RSX conference because the fix is a hardware one. --- Alan <<< EISNER::DUA0:[NOTES$LIBRARY]RSX.NOTE;1 >>> -< RSX >- ================================================================================ Note 78.0 Anybody remember Able QuadrAsync patch? 3 replies EISNER::OSUDAR "John Osudar" 19 lines 15-JUN-1988 18:24 -------------------------------------------------------------------------------- The request attached below comes from a colleague of mine who is not currently on DECUServe. Can anybody help? ------------------------------------------------------------------------ Remember when RSX-11M V3.2 came out? Back then, when DZs cost a fortune and DHs cost even more, Able sold a nice board called a QuadrAsync that had four DL11-type interfaces on it. It was a single Unibus board with a single Unibus load for four terminals. Unfortunately, there was a slight difference between the QuadrAsync and a genuine DEC DL11. The old half duplex terminal driver didn't care, but the full duplex driver did. Able, if I remember right, fixed the design pronto, but early vintage QuadrAsyncs needed a patch to module TTYL to work correctly. Over the years, we took DZs out of our VAX, put them in PDP-11s, removed the QuadrAsyncs, and lost the patch to TTYL. Now we have to put a QuadrAsync back into a PDP-11, and I can't find the patch. Does anyone remember that patch? ================================================================================ Note 151.1 Fix for Able Quadrasync/E 1 of 1 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 38 lines 17-JUN-1988 15:18 -< The fix for the Quadrasync/E >- -------------------------------------------------------------------------------- >> Unfortunately, there was a slight difference between the QuadrAsync and >> a genuine DEC DL11. The old half duplex terminal driver didn't care, >> but the full duplex driver did. Able, if I remember right, fixed the >> design pronto, but early vintage QuadrAsyncs needed a patch to module >> TTYL to work correctly. The fix for the Quadrasync/E (The DL11-E 4-line equivalent) is contained in Able Engineering Order #00142. The reason is "To maintain software compatibility with RSX11-M v3.2." It is five pages of diagrams, but the text is as follows: Make the following etch cuts on the component side of the board: A) At pin 1 of U97 B) At pin 2 of U97 C) The etch running between pins 3 & 4 of U97 D) At pin 8 of U94 Add the following wires on the component side of the board: A) U5-1 to U91-5 B) U5-2 to U91-3 C) U5-3 to U97-1 D) U5-4 to U91-9 E) U5-5 to U91-11 F) U5-6 to U97-2 G) U5-8 to U97-14 H) U5-9 to U94-5 J) U5-10 to U94-3 (Note: There is no step "I") K) U9-13 to U94-11 L) U9-12 to U97-13 M) U9-1 to U9-2 to U94-9 If you need a copy of the E.O., please call me. Alan Frisbie (213) 256-2575  ================================================================================ Note 152.0 PS/2 Keyboard problem 1 reply EISNER::GRIFFIN 24 lines 18-JUN-1988 08:17 -------------------------------------------------------------------------------- -< IBM PS/2 Keyboard problem >- I have a problem with IBM PS/2 Model 30's. We have several of these machines connected to a PC LAN (Novell), and from the file server a TCP/IP gateway to our network of MicroVAXen. The terminal emulator provided does an excellent emulation of a VT220. However, one of the software packages we are using on our MicroVAX uses the underlined character attribute of the VT220, and the emulator changes this to dark blue characters. These characters are very difficult to read on a black background. The emulator does provide a way of changing the screen colors on the PC. But this is just background info. The problem is that the emulator requires the use of the key combinations Alt-Esc and Alt-Enter(Return). There is apparently a difference from the old IBM PC series and the PS/2 hardware, as it works just fine on a PC/AT that is on the same LAN, but it will not work on a PS/2. A couple of hints: IBM added a Ctrl and an Alt key on the right side of the keyboard on the PS/2. This may have something to do with it. Also, the software does not have a way to save default screen setups. We thought of this, but it will not be implemented until the next version of the emulator. Our corporate systems people have been no help. Anyone have an idea? ================================================================================ Note 152.1 PS/2 Keyboard problem 1 of 1 EISNER::KENNEDY "Terry Kennedy" 33 lines 19-JUN-1988 02:30 -< Right way & the wrong way >- -------------------------------------------------------------------------------- > ...software packages we are using on our MicroVAX uses the underlined > character attribute of the VT220, and the emulator changes this to dark > blue characters. These characters are very difficult to read on a black... This is the function of the way IBM designed the original display adapters (MDA & CGA). The attribute byte bits for the monochrome-only attributes maps to the color attributes on the CGA. It just so happens that the blue/underline problem is the one people hit most often. Remember, IBM assumed that the CGA systems were going to be used for games, not business! > The problem is that the emulator requires the use of the key combinations > Alt-Esc and Alt-Enter(Return). There is apparently a difference from the > old IBM PC series and the PS/2 hardware, as it works just fine on a PC/AT > that is on the same LAN, but it will not work on a PS/2. Unless the emulator is using dead-key codes (Ctrl-Home, for example), it is just plain *wrong* to sniff for keyboard data bytes directly. The *only* supported (by IBM) method is to look for the meta keycodes (Shift, Ctl, Alt) as status bits in the BIOS data segment at 40H. Talking to the keyboard INT directly gets you into trouble (*always*) - remember early versions of XY- Write Word Processing? > A couple of hints: IBM added a Ctrl and an Alt key on the right side of... Exactly - And *neither* generats the old Alt & Ctl scancodes! IBM reserves the right to change *all* of the scancodes - the only reason they didn't is that there was no valid reason to. However, the BIOS maps both to the same code (2332, convoluted) when you use the 'correct' INT. If you want to Mail me a message, we can discuss the package and possible solutions offline. Terry Kennedy  ================================================================================ Note 153.0 VAX TEMPERATURE MONITORS 1 reply EISNER::EARLE "David Earle" 6 lines 18-JUN-1988 13:14 -------------------------------------------------------------------------------- I am currently looking into purchasing a temperature monitoring system for a VAX 11/785. I have read about the Gryphon Temperature Sentry, but would like to know about others. Would like to know if other shops have them; pros and cons; installation and programming, reliability, pricing, etc. ================================================================================ Note 153.1 VAX TEMPERATURE MONITORS 1 of 1 EISNER::TIHOR "By day System Manager, by night wily " 2 lines 26-JUN-1989 22:24 -< See 149.* >- -------------------------------------------------------------------------------- See 149.*  ================================================================================ Note 154.0 Power to BA213 C/D slots? No replies EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 11 lines 21-JUN-1988 01:42 -------------------------------------------------------------------------------- I just saw an article in Computer Technology Review about the BA213 backplane/cabinet used with the MicroVAX 3xxx. It stated that the C/D backplane slots DO NOT HAVE power supplied to them. The article is by two people at Dilog, which makes me tend to believe it. On the other hand, it will cause enough problems for existing boards that I rather doubt that DEC would do such a thing. Does anybody know for sure? Alan  ================================================================================ Note 155.0 Are DEBNA and TBK50 same board? 8 replies EISNER::OSUDAR "John Osudar" 5 lines 21-JUN-1988 19:06 -------------------------------------------------------------------------------- My 6220 just arrived today. It looks like the BI Ethernet interface (DEBNA) and BI TK50 console storage interface (TBK50) are identical boards, except for different ROM's. Does anybody know if that's really tru ================================================================================ Note 155.1 Are DEBNA and TBK50 same board? 1 of 8 EISNER::KENNEDY "Terry Kennedy" 7 lines 21-JUN-1988 21:31 -------------------------------------------------------------------------------- > Does anybody know if that's really true? Unlikely, unless somebody changed the interface on the console TK50. TK50's are (sort-of) SCSI interface, and Ethernet is, well, Ethernet. Of course, if there is an intelligent I/O panel anything is possible. You may want to look at the 2-5-2 Etch code on the solder side of the board to check (or have Field Service check if you're paranoid). ================================================================================ Note 155.2 Are DEBNA and TBK50 same board? 2 of 8 EISNER::HASSINGER "Bob Hassinger" 7 lines 22-JUN-1988 09:54 -< Maybe... >- -------------------------------------------------------------------------------- Every time I complain and ask why I have to take a TK50 on the 62xx rather than a TK70 I get hand waving about the design of a BI board that only supports the TK50 and is somehow involved in another interface too. This sounds a lot like what they have seemed to say - odd though... Bob H ================================================================================ Note 155.3 Are DEBNA and TBK50 same board? 3 of 8 EISNER::JENKINSON 1 line 22-JUN-1988 13:25 -< We were told they were the same >- -------------------------------------------------------------------------------- We were told they are the same board, just differnet firmware as you surmised ================================================================================ Note 155.4 Are DEBNA and TBK50 same board? 4 of 8 EISNER::OSUDAR "John Osudar" 8 lines 22-JUN-1988 18:27 -< Looks like same board indeed >- -------------------------------------------------------------------------------- Looks like you're right, the DEBNA and TBK50 are indeed the same board but with different firmware. Even the BI options book shows hints of this -- the DEBNA section block diagram shows "Tape Controller" in the middle, and mentions BI registers for tape (as "unused" registers). Since the processor on the T1034 board is a MicroVAX, I guess anything is possible! Still, it seems curious that DEC would waste the Ethernet interface IC's on a TK50 controller that doesn't need them... ================================================================================ Note 155.5 Are DEBNA and TBK50 same board? 5 of 8 EISNER::KENNEDY "Terry Kennedy" 7 lines 22-JUN-1988 20:54 -------------------------------------------------------------------------------- > ...Still, it seems curious > that DEC would waste the Ethernet interface IC's on a TK50 controller > that doesn't need them... It may have been designed as a combo board (sort of like the DMF32- serial and parallel on one card) and for some reason they decided not to use both (perhaps there wasn't enough room in the ROMS for both control programs) ================================================================================ Note 155.6 Are DEBNA and TBK50 same board? 6 of 8 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 6 lines 22-JUN-1988 22:22 -< Maybe thats where the DEBNTs went >- -------------------------------------------------------------------------------- I had heard that the "old" BI enet board, the DEBNT, was in fact designed as a combo board that did both enet and tape. Suposedly, the problem was that they couldnt get both parts to work reliably with any kind of thruput. I also heard that a DEBNA is really just a recycled DEBNT with some changes that get rid of the tape support. Sort of hangs together... ================================================================================ Note 155.7 Are DEBNA and TBK50 same board? 7 of 8 EISNER::OSUDAR "John Osudar" 2 lines 23-JUN-1988 16:21 -< sounds likely >- -------------------------------------------------------------------------------- Makes (some) sense... especially since one of the edges of the DEBNA (and TBK50!) board says "DEBNT" on it! ================================================================================ Note 155.8 Are DEBNA and TBK50 same board? 8 of 8 EISNER::TERRAZAS 13 lines 27-JUN-1988 09:57 -< DEBUA problems in the past >- -------------------------------------------------------------------------------- At one time DEC was working with a Utah firm to make the combo board a reality, according to a friend who works with that high-priced workstation firm, before they could complete the task to spec, DEC and this company had a minor falling out and DEC decided to make some changes to the DEBNT and call it DEBNA. The resulting board is not very useful in a combo mode, so the company decided to drop it. They are, however using their own DEBNT with their drivers in-house as a combo TK50-DECNET controller. All of this comes from the old days when you weren't allowed to run TK50s off of Unibusses off of BIs. MikeT  ================================================================================ Note 156.0 BA23 Power Supply Harness 1 reply EISNER::STAMERJOHN "RW Stamerjohn" 15 lines 23-JUN-1988 10:24 -------------------------------------------------------------------------------- Help, We need the main wiring harness which connects power supply (30-21749-01) to backplane in a Microvax I/BA23 configuration. Original cable part # is 70-20450-01, has been replaced by part # 17-01311-01 (FCO EQ-01427-01). Does any body know where we can get this part (the new version). We are pursuing getting part from DEC, but order process will take 2-3 weeks. Ralph p.s. We seem to have been bitten by problem discussed in 11.14. Can the replacement part 17-01311-02 (MV II) be used as substitute for the 17-01311-01 version? ================================================================================ Note 156.1 BA23 Power Supply Harness 1 of 1 EISNER::KENNEDY "Terry Kennedy" 23 lines 24-JUN-1988 02:32 -< SOme ways out... >- -------------------------------------------------------------------------------- > Help, We need the main wiring harness which connects power supply... Well, if you're willing to bite the bullet on price, DEC Field Service can emergency order it for you *if* you also place a per-call request for it to be installed. [Presuming you don't have Field Service contract coverage - if you do, they should also be able to get it in a day or so.] If they say they cannot, there is something wrong. If price is more important than DEC support, just junk the harness and the connectors on the power supply and backplane and replace with #14 or 16 stranded cable, available at electrical supply houses everywhere. Do *not* attempt to solder to the connectors - they are oxidized from the harness failure. Regarding ordering from DEC, 2-3 weeks is *very* optomistic - since it is from Unlisted Parts, the quote alone will take a week, and the lead time is currently 3 months. Besides, I am told by reliable sources that even the FCO assembly has had the connectors derated and has been superseded. I have been talking to Part No. Assistance to find out if the new part is in fact supplied in the FCO kit. Currently the answer is not known. I *do* know that the latest cable has individual wires and *white* connectors. The part given in the FCO has *yellow* connectors which will not carry the load. The part that part fixed was a molded (ribbon-style) cable assembly.  ================================================================================ Note 157.0 Xyplex Terminal servers 5 replies EISNER::BRUCKER "Ken Brucker" 12 lines 23-JUN-1988 13:42 -------------------------------------------------------------------------------- We are currently evaluating the Xyplex MAXserver and cluster controller devices to use as a corporate standard terminal server into our DEC systems. Does anyone have any comments, good or bad, about dealings with Xyplex or their servers? We are currently using a few DECserver 100s and 200s and are considering swapping out all of the DECservers and replacing them with the Xyplex boxes. Anybody out there have some experience with Xyplex? Thank you! Ken Brucker ================================================================================ Note 157.1 Xyplex Terminal servers 1 of 5 EISNER::OSUDAR "John Osudar" 32 lines 23-JUN-1988 17:04 -< we evaluated Xyplex, among others >- -------------------------------------------------------------------------------- We recently evaluated Xyplex terminal servers in preparation for adding 96 lines at our site. (f.y.i. we also evaluated Emulex Performance 4000 LAT server and considered DECserver 500) The Xyplex seems to be well-designed, both in hardware and software. It does use Xyplex's own protocols, and they presently do not support access to any of DEC's PDP-11 operating systems. (I believe they have some TCP/IP capabilities available as an option.) Xyplex claims high throughput in both directions, and that seems to be true. Their management software is quite nice, provides considerable information on status of ports, interfaces, servers, etc., and is VMS screen-oriented. Their software is compatible with (can run alongside) other Ethernet protocols on the same interface. Pricing was relatively competitive -- somewhat higher than Emulex, but lower than Ds500 for comparable line configurations. Their hardware is compact and nicely packaged. From our point of view, Xyplex's drawbacks included: use of their own protocols instead of LAT -- which requires additional host software, including device drivers and detached processes (ACP or equivalent); the lack of RSX-11M-Plus support (which LAT provides); and the fact that with only eight lines per "server", the efficiency of Ethernet usage is reduced from what you'd get with 32 lines per server (Emulex) or up to 128 (Ds500). These factors were especially important to us due to our operating environment -- our terminals may need to connect to any of several dozen systems, almost all under the management of other divisions; we have a number of PDP-11's running RSX; and our site-wide Ethernet already has a lot of traffic on it. Other sites might not object as much (or might even find some of these features to be benefits!) (Incidentally, we ended up buying three 32-line Emulex P4000's, which just arrived on Tuesday.) ================================================================================ Note 157.2 Xyplex Terminal servers 2 of 5 EISNER::KILGALLEN "Larry Kilgallen" 17 lines 23-JUN-1988 17:18 -------------------------------------------------------------------------------- My wife reports that Xyplex has been very helpful with telephone support over the years, including when they have no contractual obligation to do so (calls from someone not on the list of designated users, etc.) For my part, the fact that Xyplex uses their own protocol is a distinct advantage, since their protocol is optimized to offload work from the VAX (including on INPUT, where any other method provides more than one VAX interrupt per QIO*). The LAT protocol is in fact more general, aimed at supporting VAX, PDP-11 and TOPs-10/Tops-20. This generality is a major reason why it is so much less efficient than Xyplex. * Except, of course, for single character QIOs, where the humble DZ-11 may outperform them all if the term "perform" can be used in connection with single-character QIOs. ================================================================================ Note 157.3 Xyplex Terminal servers 3 of 5 EISNER::DOLAR 26 lines 26-JUN-1989 15:06 -< Xyplex - the next chapter >- -------------------------------------------------------------------------------- I am also a reasonably satisfied Xyplex user. My big issue was remote connectivity; and Xyplex's options blew away DEC's (MUXserver) option. I use 8 port xyplex servers, their gateways, The MAXserver, and their proprietary protocol, all of which I am satisfied with. Although this is an old conference, I am posting a note because I am concerned with Xyplex's long term direction as it impacts my configuration. Major developments with the Xyplex world include: 1. LAT support - takes a different card in the maxserver and raises questions about long-term viability of their proprietary protocol. 2. LAT patents - the only bright spot in my niche :-} . 3. Product support - While Xyplex has been exceptional among vendors in providing support for their products, they did send out a letter indicating the dropping of support for their original linear coax product, and significant price increases on some of their proprietary protocol ethernet gear. These are both ominous developments. While I understand Xyplex's need to stay current with the market, I appreciate the technical elegance of their proprietary protocol, and don't think I would survive without their management software. Any comments? ================================================================================ Note 157.4 Xyplex Terminal servers 4 of 5 EISNER::KILGALLEN "Larry Kilgallen" 19 lines 27-JUN-1989 11:32 -< They are dropping the major advantage they had ! >- -------------------------------------------------------------------------------- <<< NODE::DISK$PAGESWAPPER:[000000]IO.NOTE;3 >>> -< Pageswapper Input/Output Discussion >- ================================================================================ Note 1042.11 modems and multiplexors 11 of 11 NODE::US129532 "Larry Kilgallen" 11 lines 23-OCT-1988 20:51 -< Xyplex is Dropping their block mode input >- -------------------------------------------------------------------------------- I am sorry to report that Xyplex says effective with Version 3.0 of their software, they will no longer do DMA into the VAX based on QIO completion. They are switching to a character-based scheme using the terminal class driver in a manner similar to LAT and several other Ethernet interconnect schemes. Their main reason for abandoning the approach was the maintenance effort of tracking VMS terminal driver features. Possibly this was exacerbated by their efforts to sell into the Unix market, where nobody cares. ================================================================================ Note 157.5 Xyplex Terminal servers 5 of 5 EISNER::TINKELMAN "Bob Tinkelman - CCA" 9 lines 30-SEP-1989 11:49 -< Deep per-port buffering >- -------------------------------------------------------------------------------- I heard from a client that Xyplex has announced a new (or enhanced?) server box with a lot of per-port buffering, 16K/port? I don't have specific model references. Does anyone know the details, or better yet have any experience with this? One of our major interests would be the ability to buffer input from 9600 asynch "broadcast streams" - i.e. ones where these is no flow-control available.  ================================================================================ Note 158.0 DHQ11 Documentation Problem? 3 replies EISNER::NORTON "Bill Norton" 3 lines 24-JUN-1988 09:22 -------------------------------------------------------------------------------- I've been hearing tantalizing reports of incorrect documentation for a switch on the DHQ11 card. Does anyone know the details? ================================================================================ Note 158.1 DHQ11 Documentation Problem? 1 of 3 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 6 lines 24-JUN-1988 23:24 -< NO PROBLEMS WITH PDP'S >- -------------------------------------------------------------------------------- > I've been hearing tantalizing reports of incorrect documentation > for a switch on the DHQ11 card. Does anyone know the details? That would be news to me since I now have installed 5 of them with no problems and only DEC documentation to go on. ================================================================================ Note 158.2 DHQ11 Documentation Problem? 2 of 3 EISNER::NORTON "Bill Norton" 6 lines 27-JUN-1988 09:52 -< Further information >- -------------------------------------------------------------------------------- According to one service person, switch 1 on the the address-selection switchpack E19 is *not* 'unused'. In the open position the card acts like a DHU, and in the closed position the card acts like a DHV. He cites this as a possible cause for random crashes on a 11/73 running RSX11M+ V4.0. Anyway, his fix was to flip all the DHQ's to DHV mode. Time will tell if the crashes go away. ================================================================================ Note 158.3 DHQ11 Documentation Problem? 3 of 3 EISNER::NORTON "Bill Norton" 2 lines 29-JUN-1988 17:37 -< Atlanta says... >- -------------------------------------------------------------------------------- The service guy says Atlanta told him "for reliable operation, the switch must be *closed* so the card operates in DHV mode."  ================================================================================ Note 159.0 VT330 TERMINAL EMULATORS 4 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 2 lines 25-JUN-1988 10:40 -------------------------------------------------------------------------------- I am looking for a *FULL* VT330 terminal emulator for both the MAC and the PC. I need the multiple screen feature most of all. Any thoughts? ================================================================================ Note 159.1 VT330 TERMINAL EMULATORS 1 of 4 EISNER::RICE "Gary Rice" 12 lines 26-JUN-1988 09:17 -< Try this one >- -------------------------------------------------------------------------------- >I am looking for a *FULL* VT330 terminal emulator for both the MAC and >the PC. I need the multiple screen feature most of all. Any thoughts? A group here at McDonnell Douglas did a study of the Emulators available for the PC (didn't include the MAC). Their emphasis was graphics so I can't say that the multi-window function was considered, but they came up with the name "Walker, Rich & Quinn" as the "best" vendor of a VT300 series emulator. I don't have their number handy, but they advertise regularly in the trade press. Gary ================================================================================ Note 159.2 VT330 TERMINAL EMULATORS 2 of 4 EISNER::MAYHEW "Bill Mayhew" 15 lines 26-JUN-1988 18:32 -< WRQ's Reflection series >- -------------------------------------------------------------------------------- Walker Richer and Quinn's product is indeed good, but it is not a full 330 emulator, though it does (according to them) provide full 330 ReGIS support (i.e. it does ReGIS like a 330 would). I'm not sure if it does multi-windows. They're at 206-324-0350 in any case, and I have been VERY satisfied with their customer support except for the standard phone-lines-busy problems. When I found a bug in their VT2xx emulation, they practically knelt down, kissed my feet, and made me a beta tester for life. I have no relationship with them except that a friendly colleague works there. A variation on their product also provides an interesting looking ability to backup your PC to your VAX, without the requirement for DECnet or any other software besides base-level VMS, MS/PC-DOS, and their product (which is called Reflection N, by the way, N=2 ==> VT220, N=4 ==> VT240 w/VT330 ReGIS, "plus" after either model being the one that can do PC backups and some other stuff). ================================================================================ Note 159.3 VT330 TERMINAL EMULATORS 3 of 4 EISNER::KILGALLEN "Larry Kilgallen" 8 lines 27-JUN-1988 08:24 -< More information about the requirement, please >- -------------------------------------------------------------------------------- > I am looking for a *FULL* VT330 terminal emulator for both the MAC and > the PC. I need the multiple screen feature most of all. Any thoughts? Is it the "single async line" aspect of multiple screen that you crave, or some other aspect of the way that VT330s work? If your MAC/MS-DOS machine were on LocalTalk/Ethernet, you could have multple sessions at the same time without doing VT330 emulation. (Or is it the graphics that concern you most?) ================================================================================ Note 159.4 VT330 TERMINAL EMULATORS 4 of 4 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 14 lines 27-JUN-1988 08:58 -< THIS IS NOT A DUAL SESSION PROBLEM >- -------------------------------------------------------------------------------- > Is it the "single async line" aspect of multiple screen that you > crave, or some other aspect of the way that VT330s work? If your > MAC/MS-DOS machine were on LocalTalk/Ethernet, you could have multple > sessions at the same time without doing VT330 emulation. (Or is > it the graphics that concern you most?) None of the above. The VT330 has multiple pages of memory. We have an application in which the user has to review/fill out four data entry screens to complete a single transaction. Rather than repaint the screen forms over a 2400 baud (2400 BPS for you Alan) telephone lines we load up four pages of the VT330 and just keep switching between pages. We now have this application running on VT330's and HDS 2200's but some of our users also want a PC on their desk.  ================================================================================ Note 160.0 CDC 9762 SHIPPING INFO 2 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 7 lines 29-JUN-1988 12:47 -------------------------------------------------------------------------------- I justed de-installed two CDC 9762 disk drives and I have no manuals. Does anyone how to lock down the heads? Does anything else need to be done to ship these? These are the drives the RM02/03's are based on - so whatever shipping proceedure works for RM's should work for these. ================================================================================ Note 160.1 CDC 9762 SHIPPING INFO 1 of 2 EISNER::NORTON "Bill Norton" 6 lines 29-JUN-1988 17:34 -< Locking pin location >- -------------------------------------------------------------------------------- The head lock pin on an RM03 is under the cover behind the door. You have to remove the cover. The pin is at the front of the voice coil frame, and you just move it from the storage hole to the locking hole. I don't remember if the motor gets restrained or not. ================================================================================ Note 160.2 CDC 9762 SHIPPING INFO 2 of 2 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 25 lines 29-JUN-1988 22:28 -< More CDC disk drive info >- -------------------------------------------------------------------------------- >> I justed de-installed two CDC 9762 disk drives and I have no manuals. >> Does anyone how to lock down the heads? Does anything else need to be done >> to ship these? According to the RM05 User Guide, there are a few other items (that can probably be ignored). Remember, this is for an RM05, so there may be some differences. 1) The front deck hold-down screws (bolts) between the upper frame and the deck casting. Some versions also had one or more wood/foam blocks between the deck casting and unit frame. 2) Two shipping screws in the motor plate. 3) Tape over the logic chassis latch. 4) Rear deck hold-down screws (bolts), possibly under a pair of plastic caps. 5) A strap and block around and between the power supply and the blower assembly. I hope this helps. Alan  ================================================================================ Note 161.0 KMV1A driver on MicroVaxII/3600 6 replies EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 19 lines 29-JUN-1988 13:22 -------------------------------------------------------------------------------- We have a KMV1A running on a µVII with a driver that was written with the assistance of the "KMV1A MicroVMS Development Tools". We want to upgrade the CPU to a 3600, which means we will no longer be using MicroVMS. 1) If we just pull the driver, will it run on the 3600, even though we (presumably) will have no tools to do further development? 2) Hmmm how can I phrase this... What would be a reasonable thing for a vendor to do if they elimainted something like MicroVMS, and there where certain tools that where only available under that product? How quickly would they be interested in releasing such? -Charlie (toe-ing the line and needing this info yesterday) (Attention Mods, I also posted this in VMS, primarily to get greater coverage due to urgency). ================================================================================ Note 161.1 KMV1A driver on MicroVaxII/3600 1 of 6 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 20 lines 29-JUN-1988 22:39 -< You should be OK >- -------------------------------------------------------------------------------- >> We want to upgrade the CPU to a 3600, >> which means we will no longer be using MicroVMS. MicroVMS IS VMS. The only difference is that the MicroVMS kit didn't have all the files that were ususally needed only on a large VAX. As of version 5.0, even that distinction will go away. You don't have to worry about MicroVMS turning into VMS, just the usual stuff at major updates. >> 1) If we just pull the driver, will it run on >> the 3600, even though we (presumably) will have no tools >> to do further development? As with any other privileged program, you will have to re-assemble and link the driver. Be sure to read the v5.0 VMS Device Support manual for the changes necessary to run under v5.0. They are relatively minor. As for your tools, they probably run without changes, unless they are privileged. Alan ================================================================================ Note 161.2 KMV1A driver on MicroVaxII/3600 2 of 6 EISNER::BYRNE_C "Charlie Byrne" 15 lines 6-SEP-1988 12:40 -< KMV1A hangs occasionally >- -------------------------------------------------------------------------------- Our Hong Kong office has installed the KMV1A in a 3600 and are experiencing some odd problems. They are using the DEC supplied driver and the DEC supplied X.25 Link Level (HDLC) package. The are only doing transmission, no reception. They report that on an unloaded system, at semi-random intervals (about once per 1000 frames), the output stops, the KMV just send flags. The process issuing the QIO goes into LEF wait. The QIO's are issued continuously, each one issed following completion of the prior, synchronized by I/O completion AST. The speed is 19.2. I'm not yet sure if they are lsoing frames. I have a similar report from another site of ours using KMV1A for RECEIVE only. Any help on this would be appreciated whilst I await further details from overseas. ================================================================================ Note 161.3 KMV1A driver on MicroVaxII/3600 3 of 6 EISNER::BYRNE_C "Charlie Byrne" 3 lines 8-SEP-1988 11:59 -< Problem fixed >- -------------------------------------------------------------------------------- The problem refered to in .2 has apparantly been fixed. Hong Kong reports thay they got a patch from DEC CSS for the KMV X.25 Link Level s/w which has apparently done the trick. ================================================================================ Note 161.4 KMV1A driver on MicroVaxII/3600 4 of 6 EISNER::BYRNE_C "Charlie Byrne" 31 lines 5-DEC-1988 23:29 -< KMV1A availability/cabling/origin >- -------------------------------------------------------------------------------- Speaking of KMV1A's: 1) The quoted lead time is 30-60 days. If you are hoping to get one, plan on more like 60 (or more) than 30 or 45. I got mine in about 70 days, and then only after we had done much escalating and using all the pull we could as a major account. 2) Included with my KMV1A-SF (for a BA213 in a MV3600) was the following curious note: "Due to late availability of a right angle cable, a BC004 cable has been included with your KMV1A-SA/SF. Please use this cable instead of the BC22F cable." "When the correct right angle cable will be available will be announced through the DEC-Direct and Options Catalogue". Well it's nice that they gave us a 25' straight cable, but if you try to use this on a BA213, you will have the minor problem of not being able to close the door on the MV3600. FCC etc will not be amused. FS is coming to install the boards tomorrow, and I'm not sure if they are responsible for connecting the cable (probably not). What we will probably have to do is run a ribbon cable off the board, under the chassis, and connect it to the BCC04 in back. 3) Where are these things made? I believe X.25, which is probably run on a lot of KMV's, is supported from Reading, England. The spelling of "Catalogue" in the release note would lead me to believe that the note originated outside the United States. ================================================================================ Note 161.5 KMV1A driver on MicroVaxII/3600 5 of 6 EISNER::MCCORMICK_B "Bob McCormick" 4 lines 6-DEC-1988 17:43 -< Its the KING's English >- -------------------------------------------------------------------------------- Charlie: That's the way they spell it with the king's english in O'Canada, too. But I think ur right ... don't think that KMV is a Kanata (Ontairo) product ... Reading sounds Right! ================================================================================ Note 161.6 KMV1A driver on MicroVaxII/3600 6 of 6 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 1 line 13-DEC-1988 09:28 -< PSI definitely from Reading; Hardwware ??? >- --------------------------------------------------------------------------------  ================================================================================ Note 162.0 QBUS <-> UNIBUS adapter? 5 replies EISNER::BRUCKER "Ken Brucker" 6 lines 30-JUN-1988 20:15 -------------------------------------------------------------------------------- Does anyone know of a QBUS to UNIBUS adapter or vice versa? We would like to setup a single test system with both bus types that our developers can play on instead of obtaining two systems to test the respective device drivers out. Ken Brucker ================================================================================ Note 162.1 QBUS <-> UNIBUS adapter? 1 of 5 EISNER::HAHN "DECUServe Vice Chair" 1 line 30-JUN-1988 22:28 -< ex >- -------------------------------------------------------------------------------- check with ABLE for their QUNIVERTER. Pierre ================================================================================ Note 162.2 QBUS <-> UNIBUS adapter? 2 of 5 EISNER::KENNEDY "Terry Kennedy" 6 lines 1-JUL-1988 00:42 -< It's just not the same... >- -------------------------------------------------------------------------------- Note that a Unibus with a Q-bus perverter on it is *not* the same thing as a real Q-bus. Things which come to mind are timings (older unbuffered devices may get data lates on such adapters) as well as the problem of address-space mapping. You should be able to run a good number of boards on such an adapter, but it isn't the best wa to do testing... ================================================================================ Note 162.3 QBUS <-> UNIBUS adapter? 3 of 5 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 12 lines 1-JUL-1988 02:12 -< U on Q, or Q on U, or U + Q >- -------------------------------------------------------------------------------- If you do look at Able, also look at their Microverter. Some of their other cards let you go either way, but this beast was to make a U on a Q system and, since it provided the missing UMRS, lets you run a 4 meg system with U dma controllers! I have fond memories of booting an 11/73 from a Unibus SI-9400 controller with CDC-9766s (RM05s). The boot card was even an old 11/45s M-9301-YB! RSX thought it was a U system, and that the SI controller was an RH11, and used the UMRS to get into 22 bit space. Old used U hardware is so cheap, now, that it would seem best to simply have 2 seperate systems for whatever you are doing. ================================================================================ Note 162.4 QBUS <-> UNIBUS adapter? 4 of 5 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 6 lines 1-JUL-1988 23:01 -< Another vendor >- -------------------------------------------------------------------------------- >> Does anyone know of a QBUS to UNIBUS adapter or vice >> versa? Besides Able, Dilog also makes one. Alan ================================================================================ Note 162.5 QBUS <-> UNIBUS adapter? 5 of 5 EISNER::LYMAN "Rich Lyman" 8 lines 15-JUL-1988 18:45 -< Watch for missing bandwidth >- -------------------------------------------------------------------------------- Also, beware of losing bandwidth going from bus to bus. I had to put a Unibus card on the Qbus and did some tests doing DMA across configurations. Staying on either the Unibus or Qbus would give approximately one Megabyte/second. Going across the converter chopped that down to about 500 Kbytes per second.  ================================================================================ Note 163.0 DEQNA to DELQA Upgrade Plan 25 replies EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sap" 8 lines 4-JUL-1988 16:48 -------------------------------------------------------------------------------- Digital's cover letter for VMS 5.0 says that sometime in the future they will withdraw support for the DEQNA and that you should replace it with a DELQA. Does anyone know if Digital has a trade-in/upgrade plan where I could get something off the cost of a new DELQA by turn- ing in my DEQNA ? ================================================================================ Note 163.1 DEQNA to DELQA Upgrade Plan 1 of 25 EISNER::HYDE "From the laboratory of Dr. Jekyll" 7 lines 5-JUL-1988 09:21 -< still too far away to worry about >- -------------------------------------------------------------------------------- Note that the cover letter does NOT say everybody run out now and dump your DEQNA's! There has been no information about any kind of trade-in or upgrade plan that I have heard of. mark ================================================================================ Note 163.2 DEQNA to DELQA Upgrade Plan 2 of 25 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 8 lines 6-JUL-1988 12:38 -< Why is DEQNA being retired? >- -------------------------------------------------------------------------------- >< Note 163.0 by EISNER::ROSEBERRY "Bert Roseberry-that Oak among Saplings" > > -< DEQNA to DELQA Upgrade Plan >- > Digital's cover letter for VMS 5.0 says that sometime in the future > they will withdraw support for the DEQNA and that you should replace > it with a DELQA. For the uninformed, could someone explain why. Is the DEQNA THAT horrendous? ================================================================================ Note 163.3 DEQNA to DELQA Upgrade Plan 3 of 25 EISNER::KILGALLEN "Larry Kilgallen" 5 lines 7-JUL-1988 10:25 -< DEQNA retirement reasons >- -------------------------------------------------------------------------------- The stated reason for DELQA superiority over DEQNA is "reliablity". This is difference from the DELUA over DEUNA superiority which is labeled as "performance". In some cases, the DELQA may actually have 5 percent *less* throughput than the DEQNA. Fewer bits, but at least they will be the right bits... :-) ================================================================================ Note 163.4 DEQNA to DELQA Upgrade Plan 4 of 25 EISNER::MOSES "Mr. Moe" 2 lines 8-JUL-1988 14:33 -------------------------------------------------------------------------------- To expand on Larry's note... "reliability" on a heavily loaded ethernet. That's what I've been told. ================================================================================ Note 163.5 DEQNA to DELQA Upgrade Plan 5 of 25 EISNER::MERRIMAN "George Merriman - CCA/NY" 5 lines 8-JUL-1988 19:36 -< More info, please -- Thanks >- -------------------------------------------------------------------------------- > To expand on Larry's note... "reliability" on a heavily loaded > ethernet. That's what I've been told. Anyone care to expand on this? How does the lack of reliabily show itself? ================================================================================ Note 163.6 DEQNA to DELQA Upgrade Plan 6 of 25 EISNER::DELARISCH "Arnold S. De Larisch" 20 lines 11-JUL-1988 11:11 -< Heavy E-net traffic causes Corrupted Data! >- -------------------------------------------------------------------------------- >> < Note 163.5 by EISNER::MERRIMAN "George Merriman - CCA/NY" > >> -< More info, please -- Thanks >- >> >> > To expand on Larry's note... "reliability" on a heavily loaded >> > ethernet. That's what I've been told. >> >> Anyone care to expand on this? How does the lack of reliabily show >> itself? As I understand it, the 'problem' only shows up on HEAVILY loaded LAVc with LOTS of I/O traffic. The DEQNA starts getting corrupted data (w/o any error indication). This problem has been fixed in the past by forcing CRC checking on the packets. This is STILL the case according to the V5.0 Release Notes. If you have VAXclustering enabled, the DEQNA goes into 'Paranoid' mode ... that is 'enables the data integrity feature'. Otherwise, there is a sysgen parameter (I can't remember it off the top of my head) which will enable this 'feature'. -Arnold ================================================================================ Note 163.7 DEQNA to DELQA Upgrade Plan 7 of 25 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 5 lines 18-JUL-1988 17:10 -< DEQNA's don't like DECOM's >- -------------------------------------------------------------------------------- Another note about DEQNAs. I recall reading on the E-Store that the DEQNA does not work DEC's Broadband Ethernet adapter (DECOM). It does work just fine with ChipCom's EtherModem as we have several running at my sight !! ================================================================================ Note 163.8 DEQNA to DELQA Upgrade Plan 8 of 25 EISNER::STAMERJOHN "RW Stamerjohn" 21 lines 19-JUL-1988 00:42 -< An example of brain damage >- -------------------------------------------------------------------------------- I conducted the following experiment with a brand new DEQNA and a MVII running VMS V4.7 (and an empty ethernet) I started four file copy streams between the MVII and some other box. All was well (ran over two hours). Using a Ethernet scope, measured load on wire at 4%. The put Ethernet scope into traffic generation mode and generated 18% load addressed to the end of cable (non-existant station address). At random intervals of about 5 minutes, the DEQNA would go dead. It appear not to receive or transmit anything. After 5-6 seconds it would recover (of course timers at other box are tripping like crazy). At end of brain damaged, DECnet would report circuit-down, circuit-up. The error count doe XQA0: incremented. We got some hits back-to-back. My conclusion is DEQNA are load sensitive, even of load is going elsewhere (the technical term for this is brain-damaged). My question is...what is the policy on upgrades? ================================================================================ Note 163.9 DEQNA to DELQA Upgrade Plan 9 of 25 EISNER::WICHMAN "Jon Wichman" 4 lines 16-NOV-1988 15:20 -< DELQA upgrade - hold your breath >- -------------------------------------------------------------------------------- Upgrades are done the via the usual DEC method. They give you a song and dance and runaround and eventually (if you really want a DELQA bad enough) - the will kindly sell you a new DELQA at the usual price. ================================================================================ Note 163.10 DEQNA to DELQA Upgrade Plan 10 of 25 EISNER::NAGY "Frank J. Nagy, VAX Guru" 18 lines 17-NOV-1988 07:56 -< Yes, we are buying DELQAs to replace DEQNAs >- -------------------------------------------------------------------------------- >> Upgrades are done the via the usual DEC method. ... >> ... they will kindly sell you a new DELQA at the usual price. Yep, just ordered two DELQAs to replace the DEQNAs in the MicroVAX-II boot nodes in our LAVC. As the LAVC has grown larger (now 22 nodes) and busier, we have developed the problem of a group of several satellites crashing all at once. I think they are going with CLUEXIT bugchecks (voluntarily exiting cluster, I can't tell since (1) we use SYSDUMP.DMP in SYS$COMMON: and (2) the system dumps seem to go into the ether anyway which makes sense if the reason for the crash is lost communications with the boot node). After staring at these for awhile, we've noticed that mostly all the satellites in the crash group are being handled by the same boot node. Our best guess is that the known DEQNA problems are biting us so... If this fixes things, then we'll have to look at replacing the other 11 or 13 DEQNAs on our systems (just the one's in my group, dozens more at the lab in total). ================================================================================ Note 163.11 DEQNA to DELQA Upgrade Plan 11 of 25 EISNER::MADISON "Matt Madison" 5 lines 12-JUL-1989 19:52 -< Does DELQA cab kit = DEQNA's? >- -------------------------------------------------------------------------------- We just received a pile of DELQA's which we'll be using to replace some DEQNA's. Taking a quick look at the cabinet kit for the DELQA, it appears to be identical to the DEQNA cab kit, except it says "DELQA" on the faceplate. Has anyone used a DELQA with a DEQNA cabinet kit? Does it work? ================================================================================ Note 163.12 DEQNA to DELQA Upgrade Plan 12 of 25 EISNER::RICE "Gary Rice" 10 lines 13-JUL-1989 08:00 -------------------------------------------------------------------------------- > We just received a pile of DELQA's which we'll be using to replace some > DEQNA's. Taking a quick look at the cabinet kit for the DELQA, it appears > to be identical to the DEQNA cab kit, except it says "DELQA" on the faceplate. > Has anyone used a DELQA with a DEQNA cabinet kit? Does it work? Not to ignore your question ('cause I don't know the answer), but did you have to BUY the DELQAs or did you get them (from DEC) some other way? Gary ================================================================================ Note 163.13 DEQNA to DELQA Upgrade Plan 13 of 25 EISNER::MADISON "Matt Madison" 6 lines 13-JUL-1989 19:20 -< They were were purchased >- -------------------------------------------------------------------------------- >Not to ignore your question ('cause I don't know the answer), but did you >have to BUY the DELQAs or did you get them (from DEC) some other way? Sorry, they were not provided free as part of some new exchange program. We bought them (got a good deal though). We would have had to anyway -- the systems are self-maintained. ================================================================================ Note 163.14 DEQNA to DELQA Upgrade Plan 14 of 25 EISNER::GRANZOW 5 lines 17-JUL-1989 11:11 -< DEQNA KIT WILL WORK >- -------------------------------------------------------------------------------- <> YES, A DELQA WILL WORK WITH A DEQNA KIT! ALL YOU HAVE TO DO IS CHANGE THE BOARD. ================================================================================ Note 163.15 DEQNA to DELQA Upgrade Plan 15 of 25 EISNER::MADISON "Matt Madison" 4 lines 17-JUL-1989 19:26 -< Now what to do with all those extras? >- -------------------------------------------------------------------------------- >YES, A DELQA WILL WORK WITH A DEQNA KIT! ALL YOU HAVE TO DO IS CHANGE THE >BOARD. Thanks. ================================================================================ Note 163.16 DEQNA to DELQA Upgrade Plan 16 of 25 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 5 lines 18-JUL-1989 13:24 -< Post a note in DECUServe SWAP_MEET :-) >- -------------------------------------------------------------------------------- || < Note 163.15 by EISNER::MADISON "Matt Madison" > || -< Now what to do with all those extras? >- || || >YES, A DELQA WILL WORK WITH A DEQNA KIT! ALL YOU HAVE TO DO IS CHANGE THE || >BOARD. ================================================================================ Note 163.17 DEQNA to DELQA Upgrade Plan 17 of 25 EISNER::HAMPARIAN "Don Hamparian - IDI" 1 line 19-JUL-1989 10:35 -< Upgrade Plan Announced? >- -------------------------------------------------------------------------------- Anyone hear if DEC is going offer an upgrade plan? ================================================================================ Note 163.18 DEQNA to DELQA Upgrade Plan 18 of 25 EISNER::RICE "Gary Rice" 6 lines 20-JUL-1989 11:00 -< DEQNA >>>>>DELQA? >- -------------------------------------------------------------------------------- > Anyone hear if DEC is going offer an upgrade plan? We have applied considerable pressure on our local office to come up with something, but the results have been negative. "Pharoh's Heart" has DEFINITELY been hardened in this case. Perhaps we need to go after their first born... ================================================================================ Note 163.19 DEQNA to DELQA Upgrade Plan 19 of 25 EISNER::FERGUSON 22 lines 21-JUL-1989 08:17 -< A DEQNA replacement program? >- -------------------------------------------------------------------------------- > Anyone hear if DEC is going offer an upgrade plan? We had our 4 Dec reps together yesterday (don't ask me why we need 4) and hit them again with this question. After a lot of "we're on top of it" and "we understand the significance of the problem" we finally were told that Digital "definitely has an upgrade program in place" and they just needed to "resolve some of the details and get back to us". I almost let that go, but asked one last question: Does this program involve us having to pay for the replacement? The answer: "That's one of the details". And Dec sometimes wonders why their customers don't feel they are well informed. I certainly am much better informed now to know that they have a program in place. Aren't you? And hearing such double talk in quadraphonic stereo lends credence. perhaps that's why we have 4 reps; one would have been too embarrassed by him/herself to say something so silly. ================================================================================ Note 163.20 DEQNA to DELQA Upgrade Plan 20 of 25 EISNER::FERGUSON "Linwood Ferguson, ARA Services" 24 lines 6-SEP-1989 07:31 -< A Plan Appears - and Disappears >- -------------------------------------------------------------------------------- Here's the latest we've heard on an upgrade plan. About a month ago our Dec reps told us Dec would replace our DEQNA's free of charge with DELQA's. Sales was "taking care of" ordering the replacements. Several of us heard this. Last week they got together and denied saying that. They say there is an upgrade program that is ONLY for systems purchased very recently where the options catalog said it came with a DELQA and they shipped a DEQNA instead. Otherwise, no deal. They also say that "Dec isn't going to drop support for it" but that they are encouraging users to switch to a DELQA so they can "make use of all the new features". We can "rest assured that the DEQNA can be used just the way it always could". This latter seems in direct contridiction of everything Dec had been saying, but then again Dec was never really clear. I've been assuming this was a phase V compatibility issue, but I never saw that in writing. I guess it could also be a LAVC problem or just a LAT problem of some sort. Anyone know WHY Dec announced de-support? Anyone heard anything more on upgrades? ================================================================================ Note 163.21 DEQNA to DELQA Upgrade Plan 21 of 25 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 9 lines 6-SEP-1989 12:06 -< Rumors only >- -------------------------------------------------------------------------------- I **SUSPECT** (nothing really hard) the reason that the DEQNA is being dropped has to do with a through put problem. I have heard that on heavily loaded DECnet's it "looses" packet. It also has a problem when downline loading (remote boot). I also heard there is some "magic" in the cluster connection manager (when running a NI cluster) that has to keep "kicking" the DEQNA or else it will hang. Supposedly all these are fixed with a DELQA. ================================================================================ Note 163.22 DEQNA to DELQA Upgrade Plan 22 of 25 EISNER::HASSINGER "Bob Hassinger" 4 lines 6-SEP-1989 16:00 -< Show them the letters packed with the VMS V5.2 update >- -------------------------------------------------------------------------------- >They also say that "Dec isn't going to drop support for it" but that they >are encouraging users to switch to a DELQA so they can "make use of all the >new features". We can "rest assured that the DEQNA can be used just the way >it always could". ================================================================================ Note 163.23 DEQNA to DELQA Upgrade Plan 23 of 25 EISNER::SAUNDERS "John Saunders, Pacer Software" 16 lines 8-SEP-1989 05:52 -< What's In A Name? >- -------------------------------------------------------------------------------- > We can "rest assured that the DEQNA can be used just the way >it always could". Keep in mind that DEQNAs have never been able to run DECnet Phase V before... They may want to stick you with Phase IV. Any DEQNA owners out there who are expecting DEC to "do the right thing" may be mistaken. The last I heard from DEC was last symposium (in the Hardware Q&A, I think), from Lou Phillipon (sp). He said DEC was in the process of figuring out what to do about the problem. Let's not permit them to think about it too long, ok? John Saunders Pacer Software, Inc. (Proud Owner of Four Pieces of Obsolete Hardware) ================================================================================ Note 163.24 DEQNA to DELQA Upgrade Plan 24 of 25 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 14 lines 12-SEP-1989 22:46 -< new TURBO-DELQA microcode exists >- -------------------------------------------------------------------------------- Heard recently from a VERY reputable source within DEC: There is new micro code for DELQAs that provides a SIGNIFICANT performance upgrade. The new proms are apparently popular within DEC. I assume DESQAs also can use them. It was suggested that the local F/S types be asked - could it be that easy? Pray that this is what has been stalling the DEQNA->DELQA upgrade (but don't count on it). ================================================================================ Note 163.25 DEQNA to DELQA Upgrade Plan 25 of 25 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 0 lines 15-SEP-1989 03:59 -< PLEASE NOTE TOPIC 20 IN BUSINESS_PRACTICES >- --------------------------------------------------------------------------------  ================================================================================ Note 164.0 Need help with DMV for 3600 5 replies EISNER::MERRIMAN "George Merriman - CCA/NY" 10 lines 4-JUL-1988 18:58 -------------------------------------------------------------------------------- A client has a number of MicroVAX 3600's that need to be connected to the rest of his network using point-to-point DDCMP links. The trouble is DEC seems to be unable to supply a DMV or equivalent that is compatable with the new style Q-bus mechanical setup. 1. Has anybody seen a DMV with the new front plate? If so, can you give me an order number or something? 2. Has anyone tried using old-style Q-bus boards in the new card cage? What happened? ================================================================================ Note 164.1 Need help with DMV for 3600 1 of 5 EISNER::RICE "Gary Rice" 18 lines 5-JUL-1988 10:45 -------------------------------------------------------------------------------- >< Note 164.0 by EISNER::MERRIMAN "George Merriman - CCA/NY" > > -< Need help with DMV for 3600 >- > 1. Has anybody seen a DMV with the new front plate? If so, can you > give me an order number or something? We asked our sales rep the same question and s/he (there were 2 answers) said that there would be a 3600 based version available in the August/September time frame. I believe the new part number will be DMV11-FA or some such. > 2. Has anyone tried using old-style Q-bus boards in the new card > cage? What happened? Don't. It won't fit. Gary ================================================================================ Note 164.2 Need help with DMV for 3600 2 of 5 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 6 lines 5-JUL-1988 14:09 -< DMV to be 'SKUNKED' >- -------------------------------------------------------------------------------- In Cincinatti Dec said they initially had not 'SKUNKED' (made to fit the 'SKUNK' - BA213 class - box) the DMV because there was some newer card coming. Since that card is not coming as fast as was expected (at all?), they would now 'SKUNK' the DMV. ================================================================================ Note 164.3 Need help with DMV for 3600 3 of 5 EISNER::MERRIMAN "George Merriman - CCA/NY" 11 lines 6-JUL-1988 17:29 -< DMVs will fit in 3600 machines >- -------------------------------------------------------------------------------- >> 2. Has anyone tried using old-style Q-bus boards in the new card >> cage? What happened? > Don't. It won't fit. This message is coming to you through DMV removed from a MV-II and installed by DEC field service in Atlanta -- works fine. DEC claims to have shipped us DMV boards that are supposed to fit, but they never seem to catch up with us. ================================================================================ Note 164.4 Need help with DMV for 3600 4 of 5 EISNER::MERRIMAN "George Merriman - CCA/NY" 12 lines 8-JUL-1988 16:10 -< mounting kit w ªãºå >- -------------------------------------------------------------------------------- > DEC claims to have shipped us DMV boards that are supposed to fit, > but they never seem to catch up with us. The boards and mounting kits finally caught up with us. Trouble is the mounting kits do not contain any connectors or ribbon cable. Does anyone know what I need to get the connectors? Is it called a cabinet kit or what? Our DEC salesman was under the impression that the mounting kit he ordered for us is the cabinet kit but it turned out to be some kind of universal adapter for mounting traditional Q-bus boards in the new BA200-style cabinets -- it contains enough hardware to start a small junk store. ================================================================================ Note 164.5 Need help with DMV for 3600 5 of 5 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 5 lines 11-JUL-1988 11:32 -< I *like* small junk stores :-) >- -------------------------------------------------------------------------------- >< Note 164.4 by EISNER::MERRIMAN "George Merriman - CCA/NY" > > traditional Q-bus boards in the new BA200-style cabinets -- it contains > enough hardware to start a small junk store.  ================================================================================ Note 165.0 DEQNA Problem that will make your day 2 replies EISNER::ROCHELEAU "Rod Rocheleau" 34 lines 5-JUL-1988 20:30 -------------------------------------------------------------------------------- I recently got BURNED by a DEQNA that had decided that it was going to service any and all requests that it could. The result was that the network cratered. A little info on the network will give you an idea of the magnitude of the problem I had, 50 plus systems 1000 terminals on DECServers Extended LAN that covers 3 states The problems were: Session disconnecting for no reason, looked like the system was kicking the user off. Periodic unexplained login failures. The problem was tracked to a system with a DEQNA in it that was answering to all ethernet address...good or invalid. Removing power from the system and then rebooting will cleared the problem ... IT IS BAD ... REPLACE IT. Now this appeared to be a very isolated case but just last week we had a power failure and there were four systems that had this same failure, fortunately we were aware of the cause and new what to look for. There is one test that can be done to check this, on a DECServer become priviledged and issue a TEST LOOP AA-00-00-00-00-00, if you get a reply, it's bad news. This is a invalid address!. Hope this helps someone fix a problem, someday. Rod Rocheleau ================================================================================ Note 165.1 DEQNA Problem that will make your day 1 of 2 EISNER::DELARISCH "Arnold S. De Larisch" 10 lines 6-JUL-1988 13:57 -< New DEQNA Rev Level for LAVc V5.0 >- -------------------------------------------------------------------------------- >> < Note 165.0 by EISNER::ROCHELEAU "Rod Rocheleau" > >> -< DEQNA Problem that will make your day >- I don't know if this will solve your problem, but ... I attended a session (Friday Afternoon) at the DECUS Symposium on LAVc's ... they indicated that V5.0 VMS/VAXCLUSTERs REQUIRE a DEQNA with a MINIMUM rev level of K3. You might want to beat on your DEC Field Circus manager to get you this rev. -Arnold ================================================================================ Note 165.2 DEQNA Problem that will make your day 2 of 2 EISNER::WICHMAN "Jon Wichman" 12 lines 16-NOV-1988 15:29 -< DEQNA accepting all calls >- -------------------------------------------------------------------------------- >> A big 10-4 on that!!! I had the same problem here at McDonell Douglas recently. After a power fail/reboot, my MVII DEQNA became a slut and accepted any and all terminal servers. I had hundreds of lost terminal server users trying to log into my VAX for a hour or so until accidently I ran across someone from CONTEL networking who was trying to hook up a new terminal server and noticed that someone's VAX had gone crazy again. I mentioned the 1000's of bad login attempts on my system. He indicated it was a known DEQNA problem solved by powering off and on. (See subject 163 in this conference re: DEQNA phase-out)  ================================================================================ Note 166.0 CDC (Control Data Corporation) VAX Service 7 replies EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 2 lines 7-JUL-1988 12:55 -------------------------------------------------------------------------------- This topic may be used for discussion of VAX systems service by Control Data Corporation (CDC). ================================================================================ Note 166.1 CDC (Control Data Corporation) VAX Service 1 of 7 EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 23 lines 7-JUL-1988 13:04 -< Report from the Swamps >- -------------------------------------------------------------------------------- We have been using CDC to service our VAXcluster (2x8650, 2x785, 6xRA81, 4xRA82, 2xHSC50, ethernet, microvax, TXA78 (TU78 box, TU79 electronics) since December 1987, seven months so far. They do a good job overall, far superior to DEC in PMs and quality of service engineer. The big problems seem to be depth of spares, which is annoying, but almost understandable since we are the first (and until this month only) 8650 sight they admit to having in the Wash DC area. CDC has been very open about any problems, and has worked very hard to resolve any problems they do have. They recently chartered a plane and flew in a complete replacement HSC50 just to get a single board fixed. (I would have rather had a single good board, but the commitment is nice to see...) CDC has also announced a product/service called ProAct, a "service enhancement" to their VAX service support. That product/service is described as topic 43 in the 3RD_Party software conference, since it is primarily a software based tool. ProAct is designed to be a tool to compete with VAXSIMplus, but is designed to allow the individual site access to the rules and actions, see 3RD_Party software topic 43 for more information. Seton ================================================================================ Note 166.2 CDC (Control Data Corporation) VAX Service 2 of 7 EISNER::CHADWICK "Keith Chadwick, Fermilab" 72 lines 14-JUL-1988 12:52 -< You get what you pay ($$$) for... >- -------------------------------------------------------------------------------- At Fermilab, we have CDC maintaining our Cybers (three 175's and an 875), and the non-BI machines (DEC does the BI machines) in our central facility VAX cluster (one 780, two 785's, one 8600, two 8650's, one 8800, two HSC50's, two HSC70's, fourty RA81's, twelve RA82's, and fourteen TA/TU78's). In addition they also serve a number of other VAXes (not microVAXes or VAXstations - that is done internally) on site. CDC has had the maintenence contract since October 1985. Fermilab is a national research lab devoted to High Energy Physics research. This research typically records data on 6250 BPI tape at the fastest rate the transport will tollerate. The data is collected at multiple data acquisition sites (typically VAXes or microVAXes) and is then analyzed using the central facility resources (CPU's, tape vault, operator's, etc.). As a result, the TA/TU78's typically see on order of 6000-7000 tape mounts per month. We have had a long go around for over a year with DEC, CDC, and our operators regarding very severe tape problems and tape drive performance issues. DEC claims that the tapes meet the published product description, CDC does not seem to be able to maintain the tape drives up to our (or DECs) standards, and the operators must be repeatedly reminded to make sure that they clean the tape transports often (I'd like to see them cleaned on evey single tape mount). My (personal) opinion is that the CDC FE's are not as competent on the internals of the transport (adjustments etc.) as the DEC FE's we had working on the transports prior to CDC winning the contract. Another problem is related to the depth of spares. We have occasionally had two disks (RA81's) fail in a short time interval, and CDC seems to only keep one spare in the office - if a second one is required, then it usually must be flown in. Just last week, we had to copy a failing RA81 onto another of our RA81's - and it took until yesterday for CDC to repair/replace the failed RA81. On the subject of RA81 repairs - CDC seems to want to solve any problems with the minimum investment. On numerous occasions they have give us back a fixed drive only to have it fail again in an hour or so. We now refuse to accept the closing of the service call until after we have had an (in house) disk exerciser run over the disk for twenty four hours. Frequently, when the FE says that the problem is "fixed", the execiser is able to cause the "fixed" problem to reappear within an hour (sometimes we reproduce the "fixed" problem in under ten seconds)! We've had a nightmare with CDC installing FCO's in another 785 on site - it was unstable (crashes every couple of hours) for nearly two months after CDC installed the FCO, and attempted to solve the problem - which finally turned out to be an *old* floppy disk interface being incompatible with the FCO. I have also heard from another systems analyst on site that, when he received a new HSC tape channel card (K.sti) with new microcode ROM's, that the CDC FE's asked if they could copy the ROM's so that they would be able to upgrade the other K.sti's on site. His answer of course was NO! And as a result, we waited for a couple of months while CDC was unable to upgrade the remaining K.sti's. My summary optinion is that (at least in the western suburbs of Chicago) DEC Field service beats the pants off of CDC in terms of quality, response, and overall knowledge of the system in question, which in turn has an immediate impact on system integrity and availability. Fermilab's hardware maintenence contract is comming up for renewal in October - I certainly hope that DEC wins the bidding it this time! Fortunately (?), I have absolutely NO say in this matter - otherwise I probably could not have made the comments above. -Keith. ================================================================================ Note 166.3 CDC (Control Data Corporation) VAX Service 3 of 7 EISNER::CONROY "Alan Conroy" 11 lines 18-JUL-1988 14:13 -< CDC Spares >- -------------------------------------------------------------------------------- We too had problems with CDC spares (or lack thereof). After a LOT of trouble on some systems, we swore threatened and otherwise made our displeasure known, and they finally have a fairly complete set of hot spares in the area. What probably helped more than anything else is that we threatened to move to another maintenance company in the area (not DEC :-}). Of course, it helped that we almost DID go to the other company. As far as I'm concerened, they're still skating on thin ice, but the fact that we have retained them thus far says something about their responsiveness. In a city as big as Washington DC, I'm sure that there are loads of third- party maintenance companies. I'd suggest you get some quotes from them and start making waves with CDC if you aren't happy with their service. ================================================================================ Note 166.4 CDC (Control Data Corporation) VAX Service 4 of 7 EISNER::OSUDAR "John Osudar" 18 lines 20-JUL-1988 15:46 -< another western Chicago opinion >- -------------------------------------------------------------------------------- > My summary optinion is that (at least in the western suburbs of > Chicago) DEC Field service beats the pants off of CDC in terms > of quality, response, and overall knowledge of the system in > question, which in turn has an immediate impact on system integrity > and availability. Down here at Argonne National Laboratory (another DOE site, a few miles south of Fermilab) we've had CDC service since our current contract was awarded (earlier this year). I am not aware of any significant problems with CDC's service -- and they've been quite responsive and competent in dealing with our problems thus far. So I don't think I'd agree with Keith's statement -- but then, even though we're both in the western suburbs of Chicago, Fermilab deals with a different DEC Field Service office than we do. (And I don't know whether the CDC service people are the same or not, either.) p.s. CDC is just now taking over service on a BI system here, and our 6220 won't come under the CDC contract until it comes off warranty service next year. I may have more to say later... ================================================================================ Note 166.5 CDC (Control Data Corporation) VAX Service 5 of 7 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 17 lines 20-JUL-1988 16:37 -< DEALING WITH PEOPLE NOT SYSTEMS >- -------------------------------------------------------------------------------- We have 85 large PDP-11's installed that support 2500+ terminals. They are located all over New England so we get to deal with 3 CDC offices and 4 DEC offices. The systems are spilt 50/50 DEC/CDC. Anyone who tells me in general the DEC people and better trained and have better logistics I agree with them - yet all of my new customers are being strongly steered to CDC. CDC is signing 8 out of the 10 systems we sell each year. WHY? Is it price? Nope - it is not price - DEC will match the CDC pricing. It is the hassle factor. With DEC we deal with what we feel is a very rigid system where the people have been trained to ignore input from the customer. With CDC I know all the techs - they know my sites - and they call me before they doing anything major. I am sure this has to do size. CDC is large enough that they always can get spares but not so large the organization has lost touch with the customer. They way I describe this to people is dealing with CDC today is like dealing with DEC field service used to be back around 1974-1975. Back then DEC's view of what a major account was is the same as CDC's today. ================================================================================ Note 166.6 CDC (Control Data Corporation) VAX Service 6 of 7 EISNER::CHADWICK "Keith Chadwick, Fermilab" 19 lines 30-SEP-1988 17:30 -< Regarding Note 166.2 - A word from Fermilab Management >- -------------------------------------------------------------------------------- As the July 15, 1988 bulletin board comments entitled "You get what you pay ($$$) for" indicate, they were solely the personal opinion of Keith Chadwick of Fermilab. Fermilab wishes to make it clear that as a matter of policy, it does not use computer bulletin boards or other third party communications to voice perceived or actual problems about any of our suppliers (sub- contractors). Any such contractual problems are resolved through the cognizant Fermilab Contract Administrator directly with the supplier or subcontractor. With regard to the CDC service agreement to maintain DEC equipment, CDC has met all requirements of its current subcontract and there are no outstanding problems or deficiencies in performance. Universities Research Association, Inc., operator of Fermilab for the U.S. Department of Energy regrets any misunderstanding that the comments may have created. Sincerely, K.C. Stanfield Head, Research Division ================================================================================ Note 166.7 CDC (Control Data Corporation) VAX Service 7 of 7 EISNER::MERRIMAN "George Merriman - CCA/NY" 7 lines 2-OCT-1988 17:41 -< See SITE_MANAGEMENT for further discussion >- -------------------------------------------------------------------------------- See topic 24 in SITE_MANAGEMENT for a discussion of third party maintenance. It seems to me that SITE_MANAGEMENT would be a better place to discuss service contract issues in the future. George Merriman -- HARDWARE_HELP co-moderator  ================================================================================ Note 167.0 good 8650s suddenly turn bad 1 reply EISNER::CLAPP 24 lines 7-JUL-1988 23:06 -------------------------------------------------------------------------------- Up until recently, our 8600/8650 machines have been very reliable. Three weeks ago, DEC installed an ECO that upgraded the L207 & L214 modules in the CPU and gave us a new console release with later microcode. We also upgraded our HSC from V3.5 to V3.7 that same week. We ran with no problems for two more weeks, but this week has been a disaster. We've suffered several crashes with "unknown machine hang" with error snaps that - at least in some cases - suggest an E-box ucode stall. The interesting part is that a 785/8650/8650 cluster exhibited several of these "machine hangs" on both 8650s one right after the other. As soon as one comes up the other goes down. The 785 runs fine throughout. The 8650s would run fine with limtied user loads, but turning the crowds loose resulted in another hang within 30 minutes. DEC eventually suggested that we upgrade the L211 module from rev E to F. Since then, we've seen several other problems and several other CPU modules have been considered suspect. The point is, two solid machines seem to have been turned into marginal ones. Several other sites in town have also had 86xx problems of questionable nature. Is anybody else seeing this sort of things. ================================================================================ Note 167.1 good 8650s suddenly turn bad 1 of 1 EISNER::CLAPP 16 lines 15-JUL-1988 19:43 -< What DEC said the problems were... >- -------------------------------------------------------------------------------- ANd here is the story after DEC packed up and went home: 8650 #1 was diagnosed as needing the L211 module upgraded to rev F. This is an ECO - not automatically installed in the field unless trouble seen. 8650 #2 was diagnosed as having a bad memory array. Both machines exhibited CPU problems at different times when various modules where mechanically vibrated. The disturbing part here - one that seemed solid after cleaning/reseating modules one night did another hang when vibrated just 24 hours later. As for the apparent coupling whereby the second machine would crash a few minutes after the first one - dec's response: "coincidence".  ================================================================================ Note 168.0 Printer Cable Pinouts? 2 replies EISNER::BANE "Doak Bane" 11 lines 8-JUL-1988 12:52 -------------------------------------------------------------------------------- Does anyone know the pinouts of the BC27B printer cable. So far I have contacted BLACK BOX who for some reason do not know that one. This cable should be a DB37 to 50 pin "M" block connector. Any help would be appreciated. Doak Bane Colsa, Inc. P.O.Box 1068 Huntsville, AL 35807-3301 Phone: (205) 895-4101 ================================================================================ Note 168.1 Printer Cable Pinouts? 1 of 2 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 12 lines 11-JUL-1988 11:25 -< Suggest you put personal info in WHO_AM_I >- -------------------------------------------------------------------------------- > Doak Bane > Colsa, Inc. > P.O.Box 1068 > Huntsville, AL 35807-3301 > Phone: (205) 895-4101 Doak, Suggest you put this information in WHO_AM_I for the convenience of yourself and your readers. Welcome! - Charlie Byrne (BYRNE_C) Co-Moderator HARDWARE_HELP ================================================================================ Note 168.2 Printer Cable Pinouts? 2 of 2 EISNER::CLAPP 34 lines 29-JUL-1988 11:12 -< Try this >- -------------------------------------------------------------------------------- > Does anyone know the pinouts of the BC27B printer cable. Here is what we used while waiting for DEC to ship us something that would work. Note the error that was present on the original print set. 30-NOV-1983 10:04 ,677;000000000060 From: CLAPP@ROWLF To: CLAPP@SNUFFL,CLAPP Subject: DMF to X2700 cable This is the wiring of the working cable: 37 pin male connector signal dataproducts at DMF name connector @ 2700 ------------ ------------------------------ ---------------- 26 data (1) B 20 data (2) F 22 data (3) L 1 data (4) R 24 data (5) V 23 data (6) Z 5 data (7) n 6 data (8) d 8 (not 37 as per print set) strobe to printer j 12 printer ONLINE y 14 printer CONN VFY V 18 printer DEMAND E 17 printer DAVFU RDY CC 30,33,31,34,35,29,28,32,27 various returns/grounds D,J,N,t/m,x,X T,R/AA,C/EE -------  ================================================================================ Note 169.0 I can't even turn the system on! 4 replies EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 19 lines 11-JUL-1988 20:40 -------------------------------------------------------------------------------- I am trying to locate a replacement rotary switch for the front panel of a PDP-11/60. The only part that is actually needed is the switch wafer itself (which got crushed in shipping). The DEC part number is 12-13284-00 which is no longer listed in the parts book. It will be a few weeks before they can get me a Product Demand Quote (at their usual outrageous prices). I have looked through all my electronic parts catalogs and can find nothing useful. The big problem is that all the switches I can find use either 20 or 30-degree indexing. The 11/60 switch uses 45-degree indexing. I would appreciate any and all help you might be able to give me. Alan ================================================================================ Note 169.1 I can't even turn the system on! 1 of 4 EISNER::HAHN "DECUServe Vice Chair" 3 lines 11-JUL-1988 21:33 -< Centralab and Oak >- -------------------------------------------------------------------------------- Centralab PA-4000 series are 24 position phenolic switches they stop at 45 degrees - (and more) - OAK also makes such a switch. Pierre ================================================================================ Note 169.2 I can't even turn the system on! 2 of 4 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 7 lines 22-JUL-1988 18:16 -< 24*45 = 3 full turns? >- -------------------------------------------------------------------------------- > Centralab PA-4000 series are 24 position phenolic switches > they stop at 45 degrees - (and more) - OAK also makes such > a switch. Pierre I think Alan wants a switch that stops ONLY at 45-degree increments. The most it could be would be an 8-position switch, no? ================================================================================ Note 169.3 I can't even turn the system on! 3 of 4 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 16 lines 25-JUL-1988 01:33 -< Problem solved >- -------------------------------------------------------------------------------- >> I think Alan wants a switch that stops ONLY at 45-degree increments. >> The most it could be would be an 8-position switch, no? Correct. The PA-4000 series (according to the catalog) won't do the job. However, the problem has been solved. A trip to a local surplus store yielded a couple of switches that were almost similar to the original. After a bit of cutting (some of it on my fingers -- ouch!) and filing, I now have a switch that is a reasonable replacement. (By reasonable, I mean one that the original engineer would assume was simply a manufacturing substitution.) Oh, the quotation from DEC Unlisted Parts came back: Unavailable at any price. Alan ================================================================================ Note 169.4 I can't even turn the system on! 4 of 4 EISNER::KENNEDY "Terry Kennedy" 5 lines 25-JUL-1988 02:50 -< Next higher assembly! >- -------------------------------------------------------------------------------- > Oh, the quotation from DEC Unlisted Parts came back: > Unavailable at any price. Let me guess: Next higher assembly = 11/60, no longer produced; thus: Next higher assembly = VAX 8800!  ================================================================================ Note 170.0 DEC printer==>IBM PC 2 replies EISNER::HAHN "DECUServe Vice Chair" 8 lines 13-JUL-1988 09:07 -------------------------------------------------------------------------------- How to connect DEC printers to IBM PC (clone et al) serial port? Has any one connected LA50, LA100, LQP02, LN03 to PCs (IBM type not VAXMate) and have them work properly. What is the 9 Pin (IBM side) to 25 Pin (printer side) schematic? ================================================================================ Note 170.1 DEC printer==>IBM PC 1 of 2 EISNER::TERRAZAS "Mike Terrazas" 13 lines 13-JUL-1988 09:40 -< Not Impossible, Good Luck >- -------------------------------------------------------------------------------- / Has any one connected LA50, LA100, LQP02, LN03 to PCs (IBM type / not VAXMate) and have them work properly. Yes, though it's not that straightforward. Many PC software packages don't recognize the escape sequences. / What is the 9 Pin (IBM side) to 25 Pin (printer side) schematic? This is actually documented in the "Technical Reference Manual, Personal Computer AT" book, available wherever PCs are sold. You'll need this book eventually anyway, so you might as well buy it now. MikeT ================================================================================ Note 170.2 DEC printer==>IBM PC 2 of 2 EISNER::MAYHEW "Bill Mayhew" 5 lines 13-JUL-1988 10:27 -< resources >- -------------------------------------------------------------------------------- There are several articles on this subject in recent issues of Digital's DESKTOPICS magazine (available in hardcopy and electronically through the E-store). There are also some files describing how users have done it in the DECPC forum on CompuServe.  ================================================================================ Note 171.0 Dec's Thinwire .ne. RG58-U 14 replies EISNER::BRUCE "Barton F. Bruce - CCA/MA" 119 lines 13-JUL-1988 13:21 -------------------------------------------------------------------------------- DEC has gone to great length to give us all good quality Thinwire Ethernet and has, in the process, created a lot of confusion that undoubtedly has resulted in MANY unreliable installation. Dec's wire is NOT RG58-U. It is NOT RG58-anything. It is a similar cable but the differences have serious implications. To install reliable networks using it, you must be aware of these differences. The vanilla RG58 was not good enough so Dec went to Belden and they went through 25 or 26 different versions of the wire to arrive at what they have now. There were many problems that were fixed. Some were as simple as changing the amount and pattern on the adhesive (that sticks the foil in place) to allow easier stripping. The braid over the foil was electrically adequate at 75% coverage, but they restored it to 92-93% simply to provide better mechanical clamping for connectors. The center conductor was made larger to cut attenuation, and now won't easily fit in vanilla BNC connectors sized for RG58 cable. To have a connector that will work, Dec went to AMP and the resulting connector crimps with a typical AMP tool, and will NOT crimp reliably with the Mil Spec hex tool most too suppliers sell as the standard RG58 tool! DEC must assume that everyone will buy ALL parts and tooling from them! If you do, the pieces should all work. BUT many companies have armed their technicians with RG58 tools when they were putting in other LANS from any of the many other vendors, and DEC never never ever spelled out loud and clear that this new wire MUST use different connectors that MUST be crimped with special tools. Proving this point is the fact that a contractor Dec had doing the wiring for DECWORLD had great trouble making the connectors stay on DEC's thinwire stuff. They were using the industry standard HEX tooling. This contractor is a division of THE large name-brand modular data connection system manufacturer here in Mass, and so, if anyone should know what to do, they should. BUT Dec never told anyone. Many RG58 strippers must be readjusted to fit Dec's wire, and I have yet to find one I like (I'm not sure DEC has either). Dec had a large (60ish?) man team with orders from on high to produce DecConnect-i fast. Obviously some things got overlooked. DecConnect-II seems to be the follow-on additions that have been coming out recently. There are more pieces to come, but the one that seems most significant has not been announced yet by DEC, but has been by the actual manufacturer they are buying from. I was at a trade show (data and telephone hardware) that had an AMP booth. AMP was showing what is going to make Thinwire ethernet viable for many more users. It was a connector that looks as though it taps into the cable like a tee, and mounts at a wall plate, and allows many such taps per cable segment. The trick is thet it is NOT a tee, but is like a data patch panel jack that has a 'normal through' connection when nothing is plugged in, but that breaks this 'normal' when a special twin coax cord is plugged in. This extends both sides of the broken through connection out onto this twin coax cable that has at the far end a single BNC connector wired to electrically look like a tee where this twin coax does a U turn and one side connects to the other. This lets users plug into their wallplate when they please, and they don't bother other users on the same cable segment that suddenly got longer by their little U shaped loop being inserted. The switching action in the jack is positive. There is no halfway position a user can mess others up with. A single packet may get trashed, and some error counter somewhere may get bumped, but there is no adverse impact on other users. This is neat. DecConnect-I allowed 1 user per DEMPR port. This allows about 8. The real limits are something like 185 meters and 60 connections in the cable. A tee used to count as 2 meaning you really got 30 taps, but the DEMPR counted as 1 so you were allowed 29 more. Few people went this far, as someone always was messing up a long cable, and you certainly couldn't go back into the DecConnect-I faceplate your connection came out of. This is why many cables simply loop from cubicle to cubicle over and under walls from user installed tee to tee. The new connector and its mating u-turn twin coax cable eat more of the connection budget, so DEC will probably suggest only 8 such taps per DEMPR segment (I think AMP claims more are ok), but that means you can do 64 DecConnect-II faceplates per DEMPR. This is a BIG savings. This new connector has been engineered to be easy to install, too. The center wire in not crimped, but is stuffed into an insulation displacment type self terminating connection. I think the vanilla AMP version's face is open, but that the DEC version may have a little trap door, but AMP didn't have samples of what they were making for DEC. DEC's thinwire cable is Belden 9907, or 89907 for the teflon version. Many other wire makers are now qualified vendors of this type cable and also sell to DEC. Other connector manufacturers make BNC connectors with the slightly larger center pin opening for this wire, and you can get connectors that will crimp reliably with your standard hex tools. The AMP tool has a round die that some feel gives a securer crimp. The center pin is square crimped in the AMP tool. I could list all sorts of part numbers, but I will instead suggest contacting the following folks. DEC's official distributor for DecConnect is Anixter, and they can easily give you the right AMP tool whether you order by a DEC number (try to find the tool number by itself, and not in a kit - good luck) or an AMP number or better yet by their own Anixter number. They also can give you 9907 cable by various vendors. The BNC connectors came in an AMP box, with DEC part numbers on the individual plastic baggies, but the pricing was a in line with commercial connectors bought from any connector distributor. They ship the same day and seem to have hugh stocks. If their local warehouse can't do it, they ship from one of their MANY other warehouses. Dec's distributors seem better trained than Dec's own sales force. Nice folks to deal with. By now DEC may have other distributors for DecConnect, but Anixer was it when I last checked. I know others want to be. ================================================================================ Note 171.1 Dec's Thinwire .ne. RG58-U 1 of 14 EISNER::MERRIMAN "George Merriman - CCA/NY" 1 line 13-JUL-1988 21:35 -< Why use funny cable? >- -------------------------------------------------------------------------------- What's the penalty for going true RG58 all the way? ================================================================================ Note 171.2 Dec's Thinwire .ne. RG58-U 2 of 14 EISNER::NAGY "Frank J. Nagy, VAX Guru" 5 lines 14-JUL-1988 08:20 -< We use standard RG58 >- -------------------------------------------------------------------------------- >> What's the penalty for going true RG58 all the way? Seemingly none since we have installed all our Thinwire using standard RG58 cable and connectors. After all, we have lots and lots of experience with RG58 - having 100's of miles of it installed in this place (Fermilab). ================================================================================ Note 171.3 Dec's Thinwire .ne. RG58-U 3 of 14 EISNER::TINKELMAN "Bob Tinkelman - CCA" 10 lines 14-JUL-1988 11:12 -< Can you do max distance and drops with RG58? >- -------------------------------------------------------------------------------- | >> What's the penalty for going true RG58 all the way? | | Seemingly none since we have installed all our Thinwire using standard | RG58 cable and connectors. I'd be interested to hear how many drops people run using real RG58 and what distances they run. I overheard someone from DEC say that "yea, most things work for a few drops, but let's see them run 29 drops off a DEMPR." Of course, there is then the question of how many drops you really want to run. ================================================================================ Note 171.4 Dec's Thinwire .ne. RG58-U 4 of 14 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 5 lines 14-JUL-1988 14:13 -< I think this establishes a minimum >- -------------------------------------------------------------------------------- >> I'd be interested to hear how many drops people run using real >> RG58 and what distances they run. 2 drops ~50 feet ================================================================================ Note 171.5 Dec's Thinwire .ne. RG58-U 5 of 14 EISNER::CLAPP 7 lines 15-JUL-1988 19:36 -------------------------------------------------------------------------------- We're using whatever follow-on to RG-58 popped out of the system - we didn't ask for anything special... Our worst-case thin-wire segment is about 100 meters and has 11 drops. We try to keep each segment as simple as possible. So far, the only problem we've seen was a cable cut by a room divider edge - can't blame the connector for that! ================================================================================ Note 171.6 Dec's Thinwire .ne. RG58-U 6 of 14 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 40 lines 15-JUL-1988 23:17 -< Use the Real Thing! >- -------------------------------------------------------------------------------- > -< Why use funny cable? >- > > What's the penalty for going true RG58 all the way? Belden has told our in-house signal cabling specialist that the Thin Ethernet cable is optimized for constant impedance and for <10MHz characteristics, while RG58 is primarily an HF and VHF product. The Thin Ethernet cable from Belden is comparably priced with RG58-C/U, at least in large quantities, and there is no reason to cut corners for cost. We have looked at a 1000' roll of Belden 9907 and see that is has MUCH smaller impedance variations that a new 1000' roll of RG58-C/U. The big bugaboo in Ethernet wiring is addition of reflections from small discontinuities in the cable (that's the reason for the 2.5M spacing of vampire taps on Etherhose; supposedly dis- continuities at that spacing will not add in-phase). On a long Thin Ethernet run, with connectors spaced pretty much at random, there is the possiblilty of seeing strange effects caused by reflections from these minor discontinuities (the connectors) adding and setting up standing waves. Elimination of impedance discontinuities in the cable itself can only help. This is unlikely (not impossible) to be a problem in shorter runs, or runs with only a few connectors. We now have some Thin Ethernet runs up to about 400' (about 130M) with eight connectors and three stations; all run without any problems. If you do use regular RG58, it would be the first thing to check if you start getting strangeness upon adding another station. A TDR is really the only instrument that will tell you anything about the cable run. Mark S. (We, too, have a lot of coax in our facility -- about 20 miles of 75 and 50 ohm. Our signal wiring specialist is extremely hard-nosed about the quality of the cable he will use, and takes pains with details. As a result, he gets to troubleshoot a cable-related problem about once a year...) ================================================================================ Note 171.7 Dec's Thinwire .ne. RG58-U 7 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 8 lines 16-JUL-1988 10:07 -< RG58 works, but.. >- -------------------------------------------------------------------------------- One of the catches about note .-1 is that a TDR (which is the best instrument to find problems on this kind of wiring) usually costs so much that it would be cheaper to buy the right kind of wire in the first place! The comments DEC made at symposia is that RG58 will usually work, but the total overall length of the cable, and the number of drops, must be reduced. ================================================================================ Note 171.8 Dec's Thinwire .ne. RG58-U 8 of 14 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 26 lines 2-AUG-1988 12:34 -< DREADFUL, but it works. >- -------------------------------------------------------------------------------- If you are really in a bind and need to connect an Ethernet node RIGHT NOW, there are some alternatives to Thin Ethernet cable that will work, but which will send your DEC Network Consultant into apoplexy. If you have some already-installed shielded cable intended for Async terminals, you may be able to use that cable. Odd though it may seem, dual pair shielded cable (we use Belden 9302 in large quantities) has a characteristic impedance of about 43 to 44 ohms between any one conductor and all other conductors tied to the shield. It's not particularly good about constant impedance, and it has much higher loss than the real thing, but we find that for temporary runs it is always usable up to about 150 foot runs, PROVIDED that it is used only to connect from a DEMPR (or equivalent) port to one thin-wire node only -- with no multidropping. Since we have reflections anyway, a short piece (2-3") of RG58 with a BNC connector on one end and spliced to the 9302 doesn't make matters much worse. The next thing is to schedule your Facilities department to get the proper cable in as soon as possible. If DEC -- or any other Ethernet equipment supplier -- finds out you are using this stuff, they may cheerfully refuse to acknowledge the existance of any network problems, even after you have replaced it. Remember, it works but you didn't read it here. Mark S. ================================================================================ Note 171.9 Dec's Thinwire .ne. RG58-U 9 of 14 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 15 lines 16-AUG-1988 01:18 -< A part #, (DecConnect II??) >- -------------------------------------------------------------------------------- >> I was at a trade show (data and telephone hardware) that had an >> AMP booth. AMP was showing what is going to make Thinwire >> ethernet viable for many more users. I found a part number for the tap assembly in AMP's 'Thinnet Tap System' that should make getting further info easier. Part No. 222503-1 The single sheet of paper I'm getting this off has a part # 83284-18M-STR-5/88. Amoung other claims: "Mounts in standard wallbox or allows free-hanging use in ceiling, trench, or over the floor." - I doubt TRENCH means to them what it means to me... ================================================================================ Note 171.10 Dec's Thinwire .ne. RG58-U 10 of 14 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 4 lines 16-AUG-1988 09:12 -< Our trenches >- -------------------------------------------------------------------------------- Our administration building has trenches. It is a U-shaped groove in the floor, about four inches square in cross section. It is covered by a flat, bolted down piece of sheet metal and then covered with the carpet. ================================================================================ Note 171.11 Dec's Thinwire .ne. RG58-U 11 of 14 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 20 lines 26-OCT-1988 23:21 -< DecConnect-II lives - some order numbers: >- -------------------------------------------------------------------------------- >>>< Note 171.9 by EISNER::BRUCE "Barton F. Bruce - CCA/MA" > >>> -< A part #, (DecConnect II??) >- In Anaheim the new DecConnect-II hardware was THERE! None was really installed, but in the network booth someone had the pieces in in their hands. The connector does have a trap door (see .0), and is H3114-AA. The special twin coax cable is BNE24-2E, and you need the new Dec faceplate H3111-C that starts totally unequipped with any sort of jack. These numbers are from my notes and I havn't yet tried them on the electronic store. There also was an AMP (the actual manufacturer) salesman at DEXPO showing them off in one of their distributor's booth. I guess they are really here at long last. ================================================================================ Note 171.12 Dec's Thinwire .ne. RG58-U 12 of 14 EISNER::MAYHEW "Bill Mayhew" 4 lines 27-OCT-1988 13:31 -< Warning: AMP version <> H3114 >- -------------------------------------------------------------------------------- From the Digital announcement text of the H3114-AA, I note that the (daisy-chain thinwire faceplate) connector is *not* the same as the AMP 'similar product', and that Digital Field Service will not support the AMP version. ================================================================================ Note 171.13 Dec's Thinwire .ne. RG58-U 13 of 14 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 14 lines 28-OCT-1988 01:21 -< need I care about F/S? >- -------------------------------------------------------------------------------- >>>... H3114-AA ... connector is *not* the same as the AMP 'similar product', >>>and that Digital Field Service will not support the AMP version. As a guess, I think all that is different is that DEC has a trap door to keep crud out, and mounts in their faceplate rather than via 2 screws in AMPs more conventional faceplate. I would be suprised if any of the important guts differ. As for F/S's position, it would seem with DEC's relatively small percentage of the office PC lan market that their position can only HURT DEC's credibility. What do they say to something like a Synoptics Twisted-Pair lan connecting someones many PCs and a few VAXes? If their NON-SUPPORT is that they won't install it for me, that is fine. ================================================================================ Note 171.14 Dec's Thinwire .ne. RG58-U 14 of 14 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 60 lines 4-NOV-1988 22:22 -< More Daisy_chain info >- -------------------------------------------------------------------------------- >>>< Note 171.12 by EISNER::MAYHEW "Bill Mayhew" > >>> -< Warning: AMP version <> H3114 >- >>> From the Digital announcement text of the H3114-AA, I note that >>> the (daisy-chain thinwire faceplate) connector is *not* the same >>> as the AMP 'similar product', and that Digital Field Service will >>> not support the AMP version. It seems many vendors (except DEC) were at the LocalNet 88 trade show that just ended in Boston today. The new AMP 'similar product' was there, too, and clearly THE difference is the way the connector mounts in the faceplate. AMP's has two screws et either end of the opening (like a wall SWITCH plate) while DEC's snaps into their faceplate. The daisy-chain cable that plugs into these outlets is IDENTICAL for the DEC or AMP versions, AND the AMP ones are listed is 4 sizes from 8' to 25', while DEC's are just 8' + 12'. The outlets are relatively inexpensive, but the cables will make you choke. The AMP ones at any of their distributors will be about 1/2 DEC's price. This is a NEW product, and most distributors probably don't have it or for some even know of it yet. Anixter (nationwide) is DEC's DecConnect distributor (they do AMP too...) with their Wilmingtom MA. site's manager being the interface to DEC, guess who probably knows most. Another MA. distributor, SAGER in Hingham, has AMP stock and has been shipping. Soon you should be able to get the stuff anywhere, but for now if you are stuck try these. DEC has new GREY, now, DecConnect manuals which include these Daisy-chain products. There is even a 'test' cord that, rather than looping the twin coax at the tee'd BNC, is cut off into 2 cables each with a BNC. They recommend each site having one, and I would add to that 'under lock and key'. I don't promise these all have daisy chain in them (they just came), but here are the new pubs that came: Office Comm. Cabinet Inst Guide EK-DECSY-OC-002 FEB 88 Planning + Config Guide EK-DECSY-CG-002 AUG 88 has daisy-chain*** Sat Equip Room Inst Guide EK-DECSY-SR-001 AUG 88 Remote Wall Enc Inst Guide EK-DECSY-WE-001 Mar 88 Facilities Cabling Inst Guide EK-DECSY-FC-001 Sep 88 Backordered were: Decconnect planning + inst guide ...-TG Decconnect Sys General Desc. ...-GD Decconnect SYS IPB Manual (Note NOT EK-DECSY...) EK-DECON-IP There are more referenced in 'associated doc' sections that I didn't get. One potentially of interest is: ThinWire Daisy-Chain Faceplate Connector Installation Instructions EK-DECSY-DC This may be what comes with the connectors. The pictures also show a new cable stripper, hopefully better than what DEC provided before...  ================================================================================ Note 172.0 Big Ben Increases Paper Capacity of LN03 No replies EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sa" 10 lines 14-JUL-1988 06:11 -------------------------------------------------------------------------------- I just bought a laser printer that doesn't do what I was told it could do. But out of this nasty little incident, I also found out something. We use LN03 printers and the users are constantly refilling the paper tray. The Laser Connection out of Mobile Alabama makes Big Ben which is a 1000 page sheet feeder. This sheet feeder (actually I would call it a paper elevator) fits in place of the normal paper tray. It fits on LN03 lookalikes that use the Ricoh 4080/4081 engine such as AST TurboLaser, Decision Data 6408, Quadram Quadlaser.  ================================================================================ Note 173.0 DF242 "fix" 4 replies EISNER::PRIGOT "Jonathan M. Prigot" 13 lines 14-JUL-1988 13:32 -------------------------------------------------------------------------------- I just ran into an interesting phenomenon with our DF242 "Scholar Plus" modems. We typically have them set for MNP error control, no error control fallback, and dial in security on. Our users have 2400 MNP modems of various manufacture at home. Things were fine until we got a user who had a 212 non-MNP modem of their own. They could dial in, we would see the modem drop down to 1200, but the user would never be prompted for password (it would simply hang up after a time). We finally fixed things by setting error control fallback to ON. According to the manuals, this should only affect calls made from the DF242, but it appears that this also affects calls coming into the modem. ================================================================================ Note 173.1 DF242 "fix" 1 of 4 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 95 lines 14-JUL-1988 14:32 -< ANY COMMENTS ON THE CURRENT DECUServe DF242 SETTINGS? >- -------------------------------------------------------------------------------- Operational : 1 Modem protocol : (1200) Bell 212 =0 (300 ) Bell 103 =1 (2400) V.22bis =2 (1200) V.22 =3 [2] 2 Character echo : Off=0, On=1 [1] 3 Answer mode : Auto=0, Man=1 [0] 4 Attached msg on incoming call : Off=0, On=1 [0] 5 Delete key effect : CRT=0, Hardcopy=1 [0] 6 Enable Rem Dig Loopback : Off=0, On=1 [1] 7 Blind Dialing : Off=0, On=1 [0] 8 Reserved For Digital Use 9 Reserved For Digital Use 10 Reserved For Digital Use 11 Automatic Retrain : Off=0, On=1 [0] 12 Speaker enable : Off=0, On=1 , On if no CD=2 [0] 13 Speaker volume : Low=1, Med=2, Hi=3 [1] 14 Ring to Answer on Ring = 1 to 10 [1] 15 Callback Timer Time = 1 to 6 [1] Press RETURN for more ... Communication : 1 Data Bits : 7 bits=3 8 bits=4 [4] 2 Parity : None=1, Even=2, Odd=3 [1] 3 Stop bits : One=1, Two=2 [1] 4 Off-line XON/XOFF recognition Off=0, On=1 [1] 5 Communication Protocol : Async=0, Sync=1 [0] 6 Sync mode timing : Int=0, Ext=1, Slave=2 [0] 7 DSR forced : On=1, Normal=2 [2] 8 EIA CD/CTS Signals Separate = 0 Common = 1 [0] Press RETURN for more ... Disconnect : 1 Send Space Disconnect Off=0, On=1 [1] 2 Receive Space Disconnect Off=0, On=1 [1] 3 Carrier Loss Disconnect Off=0, On=1 [1] 4 Current Loss Disconnect Off=0, On=1 [1] Press RETURN for more ... Error Correction : 1 Error Correction : Off=0, MNP=1, X.PC=2 [0] 2 Auto Fallback : Off=0, On=1 [0] 3 Transmit Break Reaction : Disregard=0 Destructive=1 Non-Destructive Immediate=2 Non-Destructive Sequenced=3 [0] 4 Receive Break Reaction : Disregard=0 Destructive=1 Non-Destructive Immediate=2 Non-Destructive Sequenced=3 [0] 5 Reserved For Digital Use 6 MNP Disconnect Type : Non-disruptive=0 Disruptive=1 [0] Press RETURN for more ... Speed Buffering : 1 Speed Buffering State : Off=0, On=1 [0] 2 DTE Flow Control : None=0 X-ON/X-OFF=1 RTS/CTS=2 Both=3 [1] 3 Autobaud state : Off=0, On=1 [1] 4 Local baud rate : 1200 bps=0 2400 bps=1 4800 bps=2 9600 bps=3 [1] Press RETURN for more ... Access Security : 1 Callback Memory Access : Off=0, On=1 [0] 2 Dial Memory Access : Off=0, On=1 [0] 3 Parameter Access : Off=0, On=1 [1] Press RETURN for more ... Statistical : 1 Bit Error Rate : 10E3 2 Line Level : 37 ================================================================================ Note 173.2 DF242 "fix" 2 of 4 EISNER::PRIGOT "Jonathan M. Prigot" 4 lines 14-JUL-1988 16:09 -< Once BITten... >- -------------------------------------------------------------------------------- Looks substantially like ours, except that you have error correction disabled (for obvious "historical" reasons). You may wish to try turning MNP back on AND enabling Auto Fallback. I'd be interested in the outcome. ================================================================================ Note 173.3 DF242 "fix" 3 of 4 EISNER::ERSKINE "Chris Erskine" 12 lines 15-JUL-1988 08:11 -< Some other guesses >- -------------------------------------------------------------------------------- Not having a modem or manual but based upon the Vadic setup. Try changing the following: Speed buffer = On Local Baud = 2400 Autobaud = Off. If this does what I think it does, it will say that the modem will always talk to the VAX at 2400 buad but talk to the remote modem at the needed speed. This would allow the VAX to be set with autobaud disabled and allow the MNP type of modems to connect without worring about MNP not working on your side. This might also fix the problem that you have with MNP and allow you to turn it back on. ================================================================================ Note 173.4 DF242 "fix" 4 of 4 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 6 lines 15-JUL-1988 10:13 -< MAJOR EGG ON FACE >- -------------------------------------------------------------------------------- > Speed buffer = On > Local Baud = 2400 > Autobaud = Off. YUP - this is what you think it is.  ================================================================================ Note 174.0 VAXStation 2000 and high serial line activity 13 replies EISNER::KOZAM 10 lines 14-JUL-1988 23:08 -------------------------------------------------------------------------------- My VAXstation seems to get hopelessly confused if it is hit with a large number of interrupts through the serial lines (such as if I am dong a KERMIT transfer at 9600 baud PLUS some mouse activity). Has anyone else seen this? A reboot (without powering down) fixes it. Is this a hardware problem or a software (I'm using uVMS 4.6). Marc Kozam ================================================================================ Note 174.1 VAXStation 2000 and high serial line activity 1 of 13 EISNER::MAYHEW "Bill Mayhew" 19 lines 15-JUL-1988 10:03 -< Sounds familiar, no solution here... >- -------------------------------------------------------------------------------- One of my VS2K's serial lines is a 9600-baud async DECnet link to an 11/73. The other is connected to a 2400-baud modem. I occasionally see situations where things get totally whacked out such that keystroke echoes are as many as 3 characters behind (this makes Cycle-ing from one window to the next "interesting"). Usually the mouse also becomes virtually useless at such moments. A reboot solves this. Sometimes the keystrokes are garbled as well so it becomes impossible to type "reboot", so I end up having to hit the Halt button. I've noticed in these cases that even the responses to console-mode (>>>) commands are out of sync. I have not established any correlation between these incidents and, e.g., activity on my DECnet link, but I'll look for that, now. I am also running 4.6 and never had these problems under 4.5. Because of the way it attacks console-mode commands too, it feels to me like this is actually a hardware or firmware problem, causing the CPU to miss an interrupt or two along the line and then getting out of whack (and seeing "old data" on each "new" interrupt), but then again, it didn't happen under 4.5... ================================================================================ Note 174.2 VAXStation 2000 and high serial line activity 2 of 13 EISNER::KOZAM 8 lines 15-JUL-1988 20:13 -< You're symptoms are like mine >- -------------------------------------------------------------------------------- > well so it becomes impossible to type "reboot", so I end up having > to hit the Halt button. I've noticed in these cases that even the > responses to console-mode (>>>) commands are out of sync. Yup, me too. Only a reboot fixes things. At least I know it's not a hardware failure (but that doesn't really FIX things). Marc Kozam ================================================================================ Note 174.3 VAXStation 2000 and high serial line activity 3 of 13 EISNER::ROECKEL 12 lines 19-JUL-1988 12:35 -< < NODRIVER has a bug > >- -------------------------------------------------------------------------------- Gentleman, I have no experience with async DECnet, but read a message that NODRIVER under V4.6 has a BUG. This bug causes async DECnet links to mess up. Copy a new NODRIVER from a V4.7 system, and it MAY fix your problem. In case you didn't know, NODRIVER is the driver used on your async port for DECnet (at least thats what the other note said.) Bruce ================================================================================ Note 174.4 VAXStation 2000 and high serial line activity 4 of 13 EISNER::MAYHEW "Bill Mayhew" 4 lines 19-JUL-1988 14:59 -< NODRIVER unrelated to this problem >- -------------------------------------------------------------------------------- Sorry, that's not the problem. I'm the one who reported the NODRIVER problem (at least here) and fixed it on my system; it has no bearing on the problem discussed here, unfortunately. Thanks for the try, though. ================================================================================ Note 174.5 VAXStation 2000 and high serial line activity 5 of 13 EISNER::MAYHEW "Bill Mayhew" 23 lines 21-JUL-1988 15:56 -< Another incident with a little more data >- -------------------------------------------------------------------------------- My VS2000 went south again today. There was some 2400-baud modem activity, along with a moderate amount of async DECnet activity at the time (a user was SET HOSTed to my VS, running a program which provides "status reports" of about 20 char every few seconds). The modem activity was in one of two on-screen windows, and I had the BANNER running. All of a sudden a variety of unsolicited windows popped up (e.g. Workstation Options), and my keyboard interface was badly gronked (pressing any key generated random data), along with the mouse. The SET HOST session timed out. The mouse stopped working altogether. After a few sec the network link came back up and I went to the other user's terminal and re-established the SET HOST session. That worked fine. I killed off (via STOP/ID=) all the procs that were running on the main screen, and also STOPped UIS$DISPLAYMGR (sp?). But I couldn't seem to get it to restart (I tried @STARTVWS in SYS$MANAGER). Lacking time to fool with it, and since the "damage" had already been done, I rebooted by typing REBOOT on the SET HOSTed terminal, and it's been OK since. I dunno if this provides any insight into what's going on, or into an "elegant" recovery. Just once it would be nice if it failed when I had time to "contend" with it... then I'd call Atlanta and let them have at it for a while. Murphy prevails, though. ================================================================================ Note 174.6 VAXStation 2000 and high serial line activity 6 of 13 EISNER::KOZAM 9 lines 23-JUL-1988 11:26 -< SET HOSTed via Ethernet? >- -------------------------------------------------------------------------------- > ...a user was SET HOSTed to my VS... Was this via ether net or serial line? I would be surprised if you had been able to revive a serial line. Marc Kozam ================================================================================ Note 174.7 VAXStation 2000 and high serial line activity 7 of 13 EISNER::MAYHEW "Bill Mayhew" 5 lines 23-JUL-1988 17:25 -< It was netted over a serial line... >- -------------------------------------------------------------------------------- < Note 174.6 by EISNER::KOZAM > >>...a user was SET HOSTed to my VS... > Was this via ether net or serial line? ...my Ethernet goes nowhere. Yeah, it is a little interesting, isn't it... ================================================================================ Note 174.8 VAXStation 2000 and high serial line activity 8 of 13 EISNER::MAYHEW "Bill Mayhew" 8 lines 23-JUL-1988 17:28 -< DECnet vs. "normal" use of serial port >- -------------------------------------------------------------------------------- P.S. to the preceding... It may be less remarkable. Doesn't use of the serial line by DECnet mean that the normal VMS serial line device driver is not employed? This would tend to point to that component of the VMS software as the culprit. BTW... have you installed the security update on your system? I have not (it's isolated and I saw no need)... but it has some "fixes" to TTDRIVER... Hmmm... ================================================================================ Note 174.9 VAXStation 2000 and high serial line activity 9 of 13 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 16 lines 23-JUL-1988 17:39 -< terminal class vs. port drivers >- -------------------------------------------------------------------------------- > Doesn't use of the serial line by > DECnet mean that the normal VMS serial line device driver is not > employed? Yes and no. The terminal driver is actually a class driver that handles all the stuff that's specific to terminals, and common to all serial interfaces, plus a port driver for the serial interface (DZ, DHV, DMF, etc.). Async DDCMP is implemented as a class driver which replaces the terminal class driver for the specific serial line(s) being used that way, but the same serial line port driver is still used. > [comments re. security update] The security update changed some things in the terminal class driver. As far as I know it made no changes to port drivers nor to the DDCMP class driver. ================================================================================ Note 174.10 VAXStation 2000 and high serial line activity 10 of 13 EISNER::KENNEDY "Terry Kennedy" 5 lines 24-JUL-1988 01:20 -------------------------------------------------------------------------------- > The security update changed some things in the terminal class driver. As far > as I know it made no changes to port drivers nor to the DDCMP class driver. Right - that's why it may fix the problem. The 'fixed' version seems to do a better cleanup between owners, as well. ================================================================================ Note 174.11 VAXStation 2000 and high serial line activity 11 of 13 EISNER::KOZAM 9 lines 24-JUL-1988 19:41 -< Software problem, not hardware? >- -------------------------------------------------------------------------------- > BTW... have you installed the security update on your system? Yes, it's installed. Seeing that you have the problem WITHOUT the update, I suppose that the update doesn't affect this problem. I'm encouraged that async DECnet continued to work. It makes me think that software may lie at the heart of the problem. Marc Kozam ================================================================================ Note 174.12 VAXStation 2000 and high serial line activity 12 of 13 EISNER::MERRIMAN "George Merriman - CCA/NY" 9 lines 24-JUL-1988 23:48 -< time to move to MICROVAX_SYSTEMS >- -------------------------------------------------------------------------------- Now that most everyone seems to think this is likely to be a software problem I think it would be better to take the discussion to MICROVAX_SYSTEMS where it will no doubt reach more people familiar with the idiosyncracies of MV-II hardware/software interaction. I've set this topic NOWRITE. George Merriman HARDWARE_HELP moderator ================================================================================ Note 174.13 VAXStation 2000 and high serial line activity 13 of 13 EISNER::BYRNE_C "Charlie Byrne" 1 line 1-AUG-1988 14:19 -< This thread moved to MICROVAX_SYSTEMS 61.* >- -------------------------------------------------------------------------------- Charlie Byrne - Co-Moderator - HARDWARE_HELP  ================================================================================ Note 175.0 MEDIA CONVERSION NEEDS 4 replies EISNER::PROVOST 4 lines 15-JUL-1988 11:01 -------------------------------------------------------------------------------- This TOPIC is for posting your media conversion needs. Media conversion is one of the services some LUGs provide, but some media is rare and some users are out of range of a LUG. ================================================================================ Note 175.1 MEDIA CONVERSION NEEDS 1 of 4 EISNER::PROVOST 15 lines 15-JUL-1988 11:05 -< 7-track magtape? >- -------------------------------------------------------------------------------- Clearly I had an ulterior motive for starting this TOPIC. We used to have IBM729 tape drives on uor PDP-11's. They wrote 7-track tapes with a strange bit configuration. We replaced them years ago with 9-track drives. Now we have a need to read a data tape written then, whose 9-track conversion failed. Anyone able to read 7-track anymore? If you can help, call me at Bates Linac (617)245-6600. Thanks, Tom Provost ================================================================================ Note 175.2 MEDIA CONVERSION NEEDS 2 of 4 EISNER::HAHN "DECUServe Vice Chair" 8 lines 15-JUL-1988 11:51 -< DECmate and RAINBOW >- -------------------------------------------------------------------------------- Available here : DECMATE WPS-8 (RX50) to IBM compatible ASCII files need WPS-PC (vendor Exceptional Business Solution) and a 1.2Mb disk drive RAINBOW DOS->CPM or vice versa to IBM 5.25 disk ================================================================================ Note 175.3 MEDIA CONVERSION NEEDS 3 of 4 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 14 lines 15-JUL-1988 11:51 -< FOR $ >- -------------------------------------------------------------------------------- > -< 7-track magtape? >- We can do..... TK25 TK50 7-TRACK (200,556,800) 9-TRACK (800,1600,3200,6250) But we charge for it.... ================================================================================ Note 175.4 MEDIA CONVERSION NEEDS 4 of 4 EISNER::KENNEDY "Terry Kennedy" 17 lines 15-JUL-1988 17:47 -< Our list >- -------------------------------------------------------------------------------- We do: RX01/02/33/50 IBM PC 360/720/1.2M/1.4M 800/1600/3200/6250 9-track RL02/RK07/RA60 We can format/load for you: RD3x/RD5x RAxx Most SMD drive/controller combos We can also read/write about 300 CP/M diskette formats and about 120 1/4" cartridge tape formats. For more info or additional reqirements, send mail (SEND/AUTHOR) here.  ================================================================================ Note 176.0 Need UPS Specs 4 replies EISNER::LOBMEYER "Fred Lobmeyer City of NY" 21 lines 15-JUL-1988 12:30 -------------------------------------------------------------------------------- We have been having many problems with power and have just been given the task to write specifications for a UPS system. What should I ask for? What do I need? Current configuration 780, 8200, 2 HSC50's, TA78, 9 RA disks in 3 - 3 high cabinets, various com gear, IBM Series I. I know I need a unit that provides: 3 phase power 20 minutes of reserve power What do I need in terms of SAG, Surge, Transients, etc. protection ? Power need is currently at 25 KVA No I can't have a generator. :-} Thanks up front to all those that reply ! ================================================================================ Note 176.1 Need UPS Specs 1 of 4 EISNER::COY "Dale E. Coy" 6 lines 15-JUL-1988 15:18 -< Can you specify a Rotary UPS? >- -------------------------------------------------------------------------------- A very simplistic answer, IF you can get away with it, is to specify a Rotary UPS. Any Rotary UPS that I can imagine will do OK in terms of sags, surges, transients, etc. - but be sure to specify it to work normally under your minimum voltage (brownout) conditions. Otherwise, you'll have to specify a lot of things. ================================================================================ Note 176.2 Need UPS Specs 2 of 4 EISNER::KENNEDY "Terry Kennedy" 9 lines 15-JUL-1988 17:52 -< Mixed-mode solution >- -------------------------------------------------------------------------------- Switchover time is very important for things like modems and protocol converters. They have swtich-mode power supplies and anything over 2 milliseconds may kill them. However, if you go around asking for 2 Ms. switchover time, you will probably be sold a 'continuous on-line' type of UPS. For your load, that's about a zillion dollars. I would suggest Something like a Liebert UPS (worked fine with VAX 780/8600 systems, as well as with IBM Series/1 and 43xx gear). Then get several, smaller con- tinuous on-line units (I like the 3.5 Kwatter from Superior Electric) to handle your modems and small stuff that can't take the switch time. ================================================================================ Note 176.3 Need UPS Specs 3 of 4 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 19 lines 15-JUL-1988 23:35 -< Do you NEED 20 min? >- -------------------------------------------------------------------------------- The joker here is your requirement for 20 minutes backup power. You're gonna find this extremely hard, because you're gonna also have to cool the equipment as well. Your air conditioners tale a LOT of power, and it probably is 277/480 V instead of 120/208. If you can keep the backup time down to five or ten minutes, the UPS will be less expensive. Flywheel motor-generators are great for conditioning power on a regular basis, and for riding through *short* power outages which are the most common (two or three seconds while they switch things around in the substation), but are useless beyond that. I don't know what your power outages are like; ours (Chicago northern suburbs) are very seldom longer than two minutes, unless disaster has happened and it will be off for hours. Good luck. Mark S. ================================================================================ Note 176.4 Need UPS Specs 4 of 4 EISNER::LEDERMAN "Bart Z. Lederman" 19 lines 16-JUL-1988 10:04 -< I reccomend battery float systems. >- -------------------------------------------------------------------------------- If you want a 20 minute hold-up time, you MUST have batteries. In which case, you might as well have the type of UPS that rectifies the power line, floats accross bateries, then inverts DC to power your equipment. This takes care of sags, missing cycles, spikes, most noise (depends on the inverter), low input voltage, and so on. This is what we use, and what most of the other Telecommunications companies use (though apparently not the telephone company in Framingham, judging by other notes on the system!): but part of the reason is that the 48 volt battery supply is needed to run the telecom. equipment. If you want the 20 minute hold-up time for the purpose of giving you the opportunity of shutting down in an orderly manner, or riding out short power outages in the hope the power company will come back on, then perhaps you should consider NOT putting the air conditioner on the UPS. You will save by not having to build a UPS large enough to support the air conditioner, and most computers can run for 10 to 20 minutes without the air conditioner if the room temperature is low enough when the system first goes out.  ================================================================================ Note 177.0 Buying memory in Korea 1 reply EISNER::BRYANT "Geoff Bryant" 10 lines 16-JUL-1988 19:15 -------------------------------------------------------------------------------- This may be an unusual question, and I hope this is the right place to ask: A friend has a PC and would like to add more memory. It occurred to him that since he will be on a trip and will be in Hong Kong and in Seoul, Korea, maybe he can get some cheap. The problem is who/where to buy them from - who sells retail and who is reputable? Any ideas? ================================================================================ Note 177.1 Buying memory in Korea 1 of 1 EISNER::KENNEDY "Terry Kennedy" 10 lines 17-JUL-1988 00:28 -< Arrrgghhhh! Not Korean RAMS! >- -------------------------------------------------------------------------------- > A friend has a PC and would like to add more memory. It occurred... I would *stongly* advise against *any* Korean RAM chips - here at SPC we specify Japanese only and refuse any units with others. We also drop the company from the supplier list. The QC on Korean RAMS is *abysmal* - I think they sneeze in the clean rooms or something! I have nothing against other Korean products, but not chips! I would also consider whether you'll be hassled by US Customs when you decide whether the (small) savings is worth it.  ================================================================================ Note 178.0 Peripherals and uVAX/VS-3xxx Compatibility 14 replies EISNER::NAGY "Frank J. Nagy, VAX Guru" 15 lines 18-JUL-1988 07:57 -------------------------------------------------------------------------------- This topic is for comments on compatibility between peripheral controllers and other Q-Bus cards and the MicroVAX-3xxx/VAXStation-3xxx. The issue here is not the funny handles needed for the BA200 box used by the MicroVAX-3xxx and VAXStation-3500 systems, but system compatibility - that is the equipment works as it would on a MicroVAX-II system. So far in this conference, Note 90.* had a short discussion about incompatibility of the TK50/TQK50 with the VAXStation-3200. Since the 3200 comes in a standard BA23 with a standard Q-Bus, what else does not work with it? (Does a TK50 really NOT work?) Given how Digital ships the machines, we can assume that TK70/TQK70, RQDX3/RDxx disks, KDA50 disk controllers and DELQAs all work in the 3xxx series systems. Anyone out there have other information (either positive or negative) to report? ================================================================================ Note 178.1 Peripherals and uVAX/VS-3xxx Compatibility 1 of 14 EISNER::LYMAN "Rich Lyman" 7 lines 18-JUL-1988 19:51 -< DEC supplied hardware works fine... >- -------------------------------------------------------------------------------- We have a 3600 with KDA50, DELQU, RA82, TK70, etc. The DEC hardware shipped as part of the system works fine. We also have a database coprocessor that we manufactured originally for the MicroVAX II QBUS which we put in the 3600 and with several small driver modifications, it popped right up. ================================================================================ Note 178.2 Peripherals and uVAX/VS-3xxx Compatibility 2 of 14 EISNER::CLAPP 13 lines 20-JUL-1988 18:34 -< TK50 still ?, Aviv working >- -------------------------------------------------------------------------------- > (Does a TK50 really NOT work?) I just tried to plug a TK50 in to 3200 - it did not boot the diag cart. An Emulex QD21 controller is working. A KDA50 is not, but haven't verified the hardware by swapping it back into another machine. The PU device is showing as "offline" having logged an error message about failing init. I haven't gone back and tried the TK50 again now that I'm running VMS. Interestingly, the KDA did start to boot VMS - it got past the first VMS banner before hanging. In a 3600, AVIV tape coupler seems to be working fine. ================================================================================ Note 178.3 Peripherals and uVAX/VS-3xxx Compatibility 3 of 14 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 4 lines 20-JUL-1988 18:44 -< DEC WOULD AGREE >- -------------------------------------------------------------------------------- > I just tried to plug a TK50 in to 3200 - it did not boot the diag cart. The TK50 is *NOT* supported on the MV3200 ================================================================================ Note 178.4 Peripherals and uVAX/VS-3xxx Compatibility 4 of 14 EISNER::MERRIMAN "George Merriman - CCA/NY" 11 lines 21-JUL-1988 03:28 -< DMV on 3600 >- -------------------------------------------------------------------------------- See topic 164 in this conference for a description of my adventures with a DMV in a MV 3600. We are still trying to buy cabinet kits -- DEC keeps hasling us about how the DMV isn't supported in the 3000 class machines. How does DEC expect us to use their machines if they can't supply the peripherals we need to integrate them with our networks? In any case the DMV (two of them, in fact) has worked fine for three weeks now sending news stories from the DNC in Atlanta back to New York. We used a standard MV II cabinet kit -- wrapped it in duct tape and let it rattle around in the back of the cabinet. As I pointed out before, the whole mess was installed by DEC FS in Atlanta with no complaints on their part. ================================================================================ Note 178.5 Peripherals and uVAX/VS-3xxx Compatibility 5 of 14 EISNER::NAGY "Frank J. Nagy, VAX Guru" 8 lines 21-JUL-1988 08:00 -< Situation about TK50 still confusing >- -------------------------------------------------------------------------------- From < Note 178.3 by EISNER::KILLEEN "Jeff Killeen DECUServe Chair" > >> The TK50 is *NOT* supported on the MV3200 Hmmm. Then why is the TK50 listed as a *supported* option for the VS-3200 and VS-3500 on the Electronic Store? Anywho, it looks like we will use a Summus-442 controller with CDC Wren-IV and Summus Gigatape on the first 3200s. ================================================================================ Note 178.6 Peripherals and uVAX/VS-3xxx Compatibility 6 of 14 EISNER::PROVOST 28 lines 21-JUL-1988 09:41 -< Works for us! >- -------------------------------------------------------------------------------- We obtained a VS315-AA, which includes a VS3200, VMS 19" Cab, RD54. Since we do not run LAVC or RSM, we had no way to implement the VMS license. We moved the RD54 to a VAXstation II using the RQDX3 which had been controlling the RD53. This failed. We swapped RQDX3's with the VS3200. Then the RD54 worked. We built a VMS V4.5 system on the RD54. This worked fine on the VAXstation II, but failed when moved back to the VS3200. This was due to the fact that the VS3200 is not supported under VMS until V4.6A. This failure may indicate why old diagnostics which work fine on older systems may not boot on VS3200. We moved a TK50 to the VS3200 from the VAXstation II. We used this to build a VMS 4.7 system. Now everything works, including the TK50. We decided to leave the TK50 in the VS3200 for ease of updates, since the VAXstation II's mostly have REAL tape drives. CONCLUSIONS: TK50 works on 3200's. RQDX3 needs some higher rev level than we have, to run RD54's. VMS V4.5 fails at boot time on 3200's. Tom Provost ================================================================================ Note 178.7 Peripherals and uVAX/VS-3xxx Compatibility 7 of 14 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 28 lines 21-JUL-1988 10:17 -< HERES WHAT DEC SAID IN JANUARY >- -------------------------------------------------------------------------------- > Hmmm. Then why is the TK50 listed as a *supported* option for the VS-3200 > and VS-3500 on the Electronic Store? Unless things have changed in the last 6 months - one DEC source is in error -< HARDWARE_HELP >- ================================================================================ Note 90.4 TK50, TK70 & THE VAXstation 3200 4 of 6 EISNER::KILLEEN "Jeff Killeen" 14 lines 19-JAN-1988 11:08 -< FYI >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | N e w s !___!___!___!___!___!___!___! TK50 NOT SUPPORTED AS OPTION ON VAXSTATION 3200 - 07-January-1988 The TK50 tape drive cannot be supported as an option on the VAXstation 3200 at this time. Customers requiring immediate shipment of a VAXstation 3200 with tape media must purchase the new TK70-AA tape drive and the TQK70-AA controller module. ================================================================================ Note 178.8 Peripherals and uVAX/VS-3xxx Compatibility 8 of 14 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 9 lines 21-JUL-1988 12:33 -< Max number of DELQA's on MV3x00? >- -------------------------------------------------------------------------------- Anyone know how many DELQA's one can put on an MV3600. Especially while keeping everything in a single cabinet. The only other non-standard peripheral in the 40" H9644 w/ BA123 would be a KMV1A. Yes I know it sounds ridiculous, but I may need three. No, I don't particularly care to explain why, but briefly, #1 needs to connnect to a live system, #2 needs to connect to a development system, and #3 needs to be used as a crosslink to the B side of an A/B pair. None run DECnet. ================================================================================ Note 178.9 Peripherals and uVAX/VS-3xxx Compatibility 9 of 14 EISNER::KENNEDY "Terry Kennedy" 1 line 21-JUL-1988 16:04 -< "Two", he said, without knowing why he said it. >- -------------------------------------------------------------------------------- -< Max number of DELQA's on MV3x00? >- ================================================================================ Note 178.10 Peripherals and uVAX/VS-3xxx Compatibility 10 of 14 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 6 lines 21-JUL-1988 21:31 -< Ummm, could you possibly elaborate on that a bit >- -------------------------------------------------------------------------------- > < Note 178.9 by EISNER::KENNEDY "Terry Kennedy" > > -< "Two", he said, without knowing why he said it. >- > -< Max number of DELQA's on MV3x00? >- ================================================================================ Note 178.11 Peripherals and uVAX/VS-3xxx Compatibility 11 of 14 EISNER::KENNEDY "Terry Kennedy" 16 lines 21-JUL-1988 21:43 -< Don't remember - sorry >- -------------------------------------------------------------------------------- > -< Ummm, could you possibly elaborate on that a bit >- Well, no, actually. It is just something that popped into my mind after I read your request. It's from something I read somewhere recently, tho. To narrow it down a bit, recent reading which might have been relevant was: a) DELQA User's guide b) VMS V5 full docset c) MicroVax 3000-serires literature Sorry I can't cite chapter and verse. Perhaps someone else has seen the same reference? Anyway, what is 'supported' may be a smaller number than what actually works - especially if you're not using DECnet on the boards. ================================================================================ Note 178.12 Peripherals and uVAX/VS-3xxx Compatibility 12 of 14 EISNER::NAGY "Frank J. Nagy, VAX Guru" 9 lines 24-JUL-1988 10:58 -< TK50 seems to be official now >- -------------------------------------------------------------------------------- The latest VAX Systems and Options Catalog (1988 July-Sept) just popped onto my desk. The TK50 is listed as supported on the VS3200 (page III.24). Looks like another case of Digitalitis (left hand not know what right hand do and neither care about brain). But it sounds like we have heard from first-hand experience that the TK50 works (great). Also, Summus has verified that their Summus-442 disk/tape controller and Summus GigaTape (8mm helical scan cartridge) work on the 3xxx systems. ================================================================================ Note 178.13 Peripherals and uVAX/VS-3xxx Compatibility 13 of 14 EISNER::ERSKINE "Chris Erskine" 8 lines 25-JUL-1988 06:58 -< Who claims not supported >- -------------------------------------------------------------------------------- The real trick is not what the 2 hands are doing but that an item is not supported until it has run the full set of tests. This can include the development of new diagnostics. This is not just a hardware limitation but also a software one. DEC will not always take a layered product and release it for a new VAX until it is tested. The other limitation is the marketing folks at DEC do not understand the real world so what is sold and supported vs. what the customer wants do not match. ================================================================================ Note 178.14 Peripherals and uVAX/VS-3xxx Compatibility 14 of 14 EISNER::BYRNE_C "Charlie Byrne" 18 lines 25-JUL-1988 13:50 -< Two or MORE DELQA's are *supported* >- -------------------------------------------------------------------------------- > < Note 178.8 by EISNER::BYRNE_C "Charlie Byrne, Freeport NY" > > -< Max number of DELQA's on MV3x00? >- > > Anyone know how many DELQA's one can put on an MV3600. After research, From the Vax-DECNET V5.0 SPD: >(1) When two or more Ethernet controllers are used on a > single CPU, each controller must be connected to a > different Ethernet cable. These cables must not be joined > by a bridge or a repeater, though they may be joined > by another DECnet router. >May 1988 >AE-D624Y-TE This seems to clearly indicate that "two OR MORE" Ethernet controllers are *supported*. As Terry pointed out though, it's another story of if/how well they will work.  ================================================================================ Note 179.0 Needed: TK50 switch 14 replies EISNER::CONROY "Alan Conroy" 9 lines 18-JUL-1988 14:46 -------------------------------------------------------------------------------- I am in need of an A-B switch for a TK50. We have two MicroVAX 2000's that currently share a single TK50. I don't like the operators having to unplug the cable and re-plug it every night for backup (no LAVc here!) A switch would be easier for them and easier on my mind. Inmac has some switches, but none with the kind of connector we need (they did have one that worked for our rack-mount TK50, but it has a different kind of connector). The kind of connecter needed is 50-pin centronics-type. If anyone knows of a switch (or even converter cables), I'd like to know. I'd prefer not to have to spend the extra money for custom cables or switches. Thanks in advance! ================================================================================ Note 179.1 Needed: TK50 switch 1 of 14 EISNER::KENNEDY "Terry Kennedy" 22 lines 18-JUL-1988 15:45 -< Are you sure? >- -------------------------------------------------------------------------------- > I am in need of an A-B switch for a TK50. Do you have to reboot/reset to make the TK50 appear? I'm not sure what the controller microcode thinks of the drive going away... > We have two MicroVAX 2000's that currently share a single TK50. You could just pick up another TK50 mechanism (as you already have the second controller, right?) That's easier yet. Shop around - prices vary widely. > ...connecter needed is 50-pin centronics-type. Beg pardon? Centronics is a 36-pin connector with two rows of contacts. That connector style in 50-pin is used for telephone equipment. If that is really what you need, try asking for a phone cable switch. It still may be custom, though. When I last looked at a TK50, it used a small (20 to 34, don't remember) header for a connector on both the drive and controller ends. You should be able to just cut the ribbon cable and attach whatever kinds of squash- on (IDC) connectors you can get a switch for... ================================================================================ Note 179.2 Needed: TK50 switch 2 of 14 EISNER::MAYHEW "Bill Mayhew" 8 lines 18-JUL-1988 18:33 -< A VAX is a VAX, but a TK50 isn't a TK50 >- -------------------------------------------------------------------------------- A TK50Z-FA, which is what a MicroVAX 2000 uses, is not the same as a garden-variety TK50. How much "not the same" is not altogether clear... but it definitely uses different cabling (the 50-pin connector), and it is widely reported to use a SCSI interface (does the generic TK50?), and it requires its own self-contained box with power supply. DEC makes scary statements to the effect of "do NOT plug anything except a MicroVAX 2000-specific device into this port; you will damage the device and your computer" in the docs. ================================================================================ Note 179.3 Needed: TK50 switch 3 of 14 EISNER::DOHERTY "Bob Doherty" 8 lines 18-JUL-1988 19:21 -< SCSI's use 50 pin Telcos >- -------------------------------------------------------------------------------- That 50 pin Telco connector is the standard one for SCSI devices and switches may, or may not be available for them( I have never seen them advertised in the MAC mags). However, I have seen adapters to take the 50 pin scsi to a DB25( which is what the Macintoshes use on the computer end). If you can dig these up you could use a DB25 switch box, which are abundantly available. Bob. ================================================================================ Note 179.4 Needed: TK50 switch 4 of 14 EISNER::KENNEDY "Terry Kennedy" 7 lines 18-JUL-1988 23:06 -< Eh? >- -------------------------------------------------------------------------------- > That 50 pin Telco connector is the standard one for SCSI devices... Maybe we have a difference of terminology? The SCSI standard uses .100x .100 receptacles for the cable (such as AMP P/N 1-499496-2) and headers for the device. The dimensions of the connector are about 2 1/2" x 1/4". A telephone connector is about 3" x 5/8", with a screw to hold it to the mating connector. ================================================================================ Note 179.5 Needed: TK50 switch 5 of 14 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 5 lines 19-JUL-1988 02:06 -< WHAT IS IT TERRY? >- -------------------------------------------------------------------------------- The connector has the construction of a Centronics connector - latch "wires" on the side / D shaped / center projection with all the pins round it. It appears to be 50 pin. It looks like a teleco but I have not lined them up to check. ================================================================================ Note 179.6 Needed: TK50 switch 6 of 14 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 3 lines 19-JUL-1988 02:09 -< P.S. >- -------------------------------------------------------------------------------- Be careful all the Black Box 50 pin switches only switch 48 pins. I don't know if this is true for anyone elses switch. ================================================================================ Note 179.7 Needed: TK50 switch 7 of 14 EISNER::KENNEDY "Terry Kennedy" 12 lines 19-JUL-1988 02:12 -< Scuzzy (not SCSI) >- -------------------------------------------------------------------------------- > The connector has the construction of a Centronics connector - latch > "wires" on the side / D shaped / center projection with all the pins > round it. It appears to be 50 pin. It looks like a teleco but I have > not lined them up to check. Not the standard SCSI connector, but easily pervertible to same. All you need is some 50-conductor ribbon cable and the appropriate squash- on connectors. Your best bet is probably to convert to DB-50's (as used on the LP25, for example). Then just get an LP25 A/B switch, making sure that *all 50* pins are switched. Be sure to use twist 'n flat cable, as every other line in SCSI is ground. Also best to not go over 15' or so in total length between controller and drive. ================================================================================ Note 179.8 Needed: TK50 switch 8 of 14 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 8 lines 19-JUL-1988 02:19 -< FOR THE RECORD >- -------------------------------------------------------------------------------- > -< Scuzzy (not SCSI) >- For the record (and off the subject). The TK50Z is suppose to be a legit SCSI tape drive - whatever that means. However the port it plugs into on the MV2000 (by intent) is not a full SCSI port - it is just enough SCSI to make the TK50Z work. Someday I will love to fine out why DEC did this. TK's on MAC? ================================================================================ Note 179.9 Needed: TK50 switch 9 of 14 EISNER::CONROY "Alan Conroy" 10 lines 19-JUL-1988 12:56 -< Are you sure? >- -------------------------------------------------------------------------------- Are you sure that ALL 50 pins are needed? I know that on the other (non- 2000) TK50 we have, only the low 25 pins are connected. If, in fact, only the lower 25 are needed then a switch the only switched 48 lines would work. Anyone know where I could get the appropriate converter cable? Or do I have to have someone make one? P.S. Sorry about the poor terminology in the first message. I guess the connecter is a 25-pair phone type. For some reason only Centronics came to mind at the time... ================================================================================ Note 179.10 Needed: TK50 switch 10 of 14 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 8 lines 19-JUL-1988 13:04 -< NOT THE SAME INTERFACE >- -------------------------------------------------------------------------------- > Are you sure that ALL 50 pins are needed? I know that on the other (non- > 2000) TK50 we have, only the low 25 pins are connected. If, in fact, only > the lower 25 are needed then a switch the only switched 48 lines would work. TK50Z and the TK50 *DO NOT* use the same tape interface. While the TK50 in your MVII uses a 26 pin ribbon cable who knows what the TK50Z tape interface uses. ================================================================================ Note 179.11 Needed: TK50 switch 11 of 14 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 9 lines 19-JUL-1988 13:26 -< OK - I GOT MY SCREW DRIVER OUT >- -------------------------------------------------------------------------------- Ok - the TK50Z contains a TK50Z interface card. The 50 pin cable goes into the card and the standard 26 pin cable for the TK50 comes out of the card. Just like the DEBNA/DEBET card is left over from a DEC project to put a tape interface on the E-NET controller - my guess is they were thinking of sending the MV2000 out the door with a SCSI bus. Knowing DEC I bet they got cold feet at the thought of an open bus. ================================================================================ Note 179.12 Needed: TK50 switch 12 of 14 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 11 lines 20-JUL-1988 15:16 -< My $.02 >- -------------------------------------------------------------------------------- We actually talked to DEC about usuing the SCSI inteface in the 200 and were told that the only thing not SCSI is the software !(ie. it does not use the SCSI Common Command Set). A special command interface to work only with the TK50 is used. Sounds like they didn't want to chang~re the TK50 end ! Okay all you driver people, get to work and write a CCS interface for this adapter and the doors to cheaper, bigger disks could be opened up ! ================================================================================ Note 179.13 Needed: TK50 switch 13 of 14 EISNER::KENNEDY "Terry Kennedy" 9 lines 20-JUL-1988 17:04 -< Hardware issue, too... >- -------------------------------------------------------------------------------- > We actually talked to DEC about usuing the SCSI inteface in the 200 and > were told that the only thing not SCSI is the software !(ie. it does > not use the SCSI Common Command Set). I think you will also discover that the required hardware support for disconnect/reconnect/arbitrate is either not present or nonfunctional on the MV side. This is not a problem with a single device, but represents a major loss of throughput with multiple devices. Thus, the interface performs more like SASI than SCSI. ================================================================================ Note 179.14 Needed: TK50 switch 14 of 14 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 5 lines 20-JUL-1988 18:38 -< FROM AN OFFLINE DISCUSSION WITH A DEC HARWARE ENG. >- -------------------------------------------------------------------------------- > I think you will also discover that the required hardware support for > disconnect/reconnect/arbitrate is either not present or nonfunctional on This is correct. The interface will not arbitrate.  ================================================================================ Note 180.0 DEC negotiates F/S contracts! 3 replies EISNER::MAYHEW "Bill Mayhew" 15 lines 18-JUL-1988 18:43 -------------------------------------------------------------------------------- I'm not altogether sure if this is the right conference for this... I have gotten a report from a client site that has a couple of 7xx's that indicates DEC Field Service is getting hungry and interested in "dealing" to maintain its customer base. The client in question was somewhat perturbed when they figured out that, despite the fact that service of memory on VAXen under a field service contract is now "free", the cost of maintenance on their machines went _up_ substantially over the past two years. After the client attempted several times to bring this to the attention of the proper manager (without success), I mentioned the problem to a friendly FS manager (who works in another unit) and all of a sudden the client is getting very interesting responses. DEC is offering to roll back the quote to last year's prices AND to provide a *3-year* agreement at those prices, rather than lose the business to one of the competing service vendors. ================================================================================ Note 180.1 DEC negotiates F/S contracts! 1 of 3 EISNER::KENNEDY "Terry Kennedy" 10 lines 18-JUL-1988 23:11 -< Are they? >- -------------------------------------------------------------------------------- > ...DEC Field Service is getting hungry and interested in "dealing"... DEC calls this a 'Comptetitive Discount', and it can go to 14% or so of the annual contract cost. However, up front they will only tell you they can do 'much better' and it takes a while to get the full 14%. We wound up going elsewhere due to the 'other guys' being something like 35% of the cost of DEC (yes, 1/3) as well as from deep dissatisfaction with the level of support we got from DEC. We perceived the discount as a delaying tactic by DEC - it took about 2 months to convert the 'much better' into a concrete 14%. ================================================================================ Note 180.2 DEC negotiates F/S contracts! 2 of 3 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 3 lines 19-JUL-1988 02:12 -< MAKES YOU WONDER ABOUT THE MARGINS ON F.S. >- -------------------------------------------------------------------------------- The standard practice around when DEC knows you are talking to CDC is to give you DECservice for the price of basic service. ================================================================================ Note 180.3 DEC negotiates F/S contracts! 3 of 3 EISNER::KENNEDY "Terry Kennedy" 4 lines 19-JUL-1988 02:13 -< Aha! >- -------------------------------------------------------------------------------- > The standard practice around when DEC knows you are talking to CDC is to > give you DECservice for the price of basic service. Which happens to be a 14-15% difference! Aha!  ================================================================================ Note 181.0 Ghost in the TK50 9 replies EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 15 lines 19-JUL-1988 11:13 -------------------------------------------------------------------------------- OK guys, here is a strange one. My VAXstation-II has a TK50 which is behaving in a most unusual manner. It is just sitting here idle, with no tape inserted and no activity on the system. Every five or ten minutes it goes "click", the green light goes off and the red (load button) light goes on. After about five or ten seconds, it goes "click" again and the lights return to normal status. From my installation notes, the controller revision is F1 and the diagnostics reported that the firmware revision is 4. I don't know where to look on the drive for its revision status. Any ideas? Alan ================================================================================ Note 181.1 Ghost in the TK50 1 of 9 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 7 lines 19-JUL-1988 13:08 -< MAYBE A DUMB QUESTION >- -------------------------------------------------------------------------------- > From my installation notes, the controller revision is F1 > and the diagnostics reported that the firmware revision is > 4. I don't know where to look on the drive for its revision I thought the REV level is reported by a switch setting on the controller. Which means shouldn't a REV F be set to 6? ================================================================================ Note 181.2 Ghost in the TK50 2 of 9 EISNER::KOZAM 7 lines 20-JUL-1988 19:53 -< Is the cabling right? >- -------------------------------------------------------------------------------- About TK50 clicking: I've seen this if the cabling is installed wrong (upside down), but I'd be surprised that a pro like you would make a mistake like that. Marc Kozam ================================================================================ Note 181.3 Ghost in the TK50 3 of 9 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 9 lines 21-JUL-1988 16:14 -< The ghost lets us get work done >- -------------------------------------------------------------------------------- >> I've seen this if the cabling is installed wrong (upside down), >> but I'd be surprised that a pro like you would make a mistake like that. Also, the drive still works (at least I could do a directory of VMS v5.0 yesterday). It seems to click more in the late afternoon when the sun shines on the side of the cabinet, but will do it at any time. Alan ================================================================================ Note 181.4 Ghost in the TK50 4 of 9 EISNER::KENNEDY "Terry Kennedy" 3 lines 21-JUL-1988 16:19 -< Have you tried keeping it in the dark? >- -------------------------------------------------------------------------------- > ...It seems to click more in the > late afternoon when the sun shines on the side of the cabinet, >but will do it at any time. ================================================================================ Note 181.5 Ghost in the TK50 5 of 9 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 7 lines 21-JUL-1988 16:22 -< I'm the one in the dark >- -------------------------------------------------------------------------------- >> -< Have you tried keeping it in the dark? >- Yes, and it still happens. Actually, it wasn't totally dark, but quite dim. Aren't ghosts supposed to be more active in the dark? :-) Alan ================================================================================ Note 181.6 Ghost in the TK50 6 of 9 EISNER::HAHN "DECUServe Vice Chair" 1 line 21-JUL-1988 21:29 -< keep on target >- -------------------------------------------------------------------------------- Light or dark lets keep it technical, or I may fall asleep. ================================================================================ Note 181.7 Ghost in the TK50 7 of 9 EISNER::PROVOST 22 lines 22-JUL-1988 09:26 -< Possible clue >- -------------------------------------------------------------------------------- This is probably unrelated, but I am continually being surprised by the cross correlations in weird phenomena. We use VAXstation II's in a real-time environment. Because of some strange equipment we had connected to the VAXstations, we took out the AUTOCONFIGURE and configured only relevant devices. Since TK50 is not relevant to our operations (we have real tape drives), we didn't configure it. STARTVWS.COM looks for existence of _TKA0. If it doesn't find it does a "sysgen connect TKA0/noadapter". This produces repetitive awakenings of the controller to tickle the TK50 and write to the errorlog. The clicking you are seeing sounds like what we witnessed in this situation. It took us a while to track down, since we could not conceive of the workstation software being interested in such an irrelevant device. Tom ================================================================================ Note 181.8 Ghost in the TK50 8 of 9 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 10 lines 22-JUL-1988 13:17 -< The ghost is on vacation(?) >- -------------------------------------------------------------------------------- >> This produces repetitive awakenings of the controller >> to tickle the TK50 and write to the errorlog. That is what is strange about this: there are NO TK50-related entries in the errorlog (other than the usual boot-time one). It seems to have stopped for the time being, so I will just keep an eye (ear?) on it. Very strange. As long as it keeps working I guess I shouldn't worry. Alan ================================================================================ Note 181.9 Ghost in the TK50 9 of 9 EISNER::PROVOST 8 lines 25-JUL-1988 13:00 -< SHOW ERROR? >- -------------------------------------------------------------------------------- > < Note 181.8 by EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" > > -< The ghost is on vacation(?) >- >That is what is strange about this: there are NO TK50-related >entries in the errorlog (other than the usual boot-time >one). Does SHOW ERROR show any errors on MU or PT?  ================================================================================ Note 182.0 Religious wars, Mag Tape division 31 replies EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 27 lines 19-JUL-1988 18:27 -------------------------------------------------------------------------------- Now for a topic that is sure to generate lots of heat. No, not politics, religion or text editors. Worse than that: Magnetic Tapes. I have heard/seen a number of comments about how 3M 700 "Blackwatch" tape is not good, for a number of reasons. These include low coefficient of friction between it and the capstan, thinner tape and shedding of the backing. I have found myself that they don't work very well with the TU81. What I would like to discuss are the various other tapes that abound in the marketplace. What problems have YOU had with them? Which ones are best for 6250, 1600 and (yes) 800 bpi? Do any appear to be especially good/bad for long-term storage? Do any seem to work well on one drive but not on another? Some of the tapes I would especially like to see comments on are: 3M 777 BASF Endura Memorex Mark V Your favorite Any comments? Alan ================================================================================ Note 182.1 Religious wars, Mag Tape division 1 of 31 EISNER::KENNEDY "Terry Kennedy" 36 lines 19-JUL-1988 18:45 -< First comments >- -------------------------------------------------------------------------------- > Some of the tapes I would especially like to see comments > on are: > 3M 777 Specifically *UN*recommended by DEC and IBM. See my rather long analysis on the Pageswapper system detailinmg the exact areas of noncompliance and failure... > BASF Endura Works OK, tends to 'grow' hard errors in mid-reel (5% or so) > Memorex Mark V I like this one - it is what I always request when buying new tapes. Why? I don't know - just gut feelings, I guess... > Others CDC? Alpha. Ok, beware of the 3600' flavor though. Has problems autoloading in TU78's. > Any comments? It seems that of all my drives, the TU78 is the hardest to please - it never reports less than 2 soft errors during BACKUP, even with a brand new tape. Old, really grubby tapes report 100's of errors, but ren clean on IBM 3420 drives. Yes, the TU78 has all the current FCO's. I noticed at Cinci that the TU79 (also made by Pertec) in the VAX booth had TU79 stick on in press-on letters. The door did, however have 3 3/8" square indentations. This is the IBM style for attaching unit numbers (3 hex digits worth of little stickers). The TU79 seems to be *much* better with marginal tapes. So much for the promise in the TU79 announcement that TU78's would be reworked to have equal reliability characteristics... ================================================================================ Note 182.2 Religious wars, Mag Tape division 2 of 31 EISNER::PROVOST 24 lines 20-JUL-1988 10:39 -< Possibly obsolete chart >- -------------------------------------------------------------------------------- Standard Tape Premium Tape* Super Premium Tape** BASF Endura Qualimetric Gold 2000 AD 3M 777 700-Blackwatch - Memorex MRX V Quantum Cubic HD WABASH Quadronix Mira 1000 - GRAHAM Summitt Epoch 480 Ultimag CDC Alpha Omega - Magnetic Media Magflux Magnum - *Meet the State of West Virginia's specifications for Items 1 and 2 including full width certified to be of a "0" Permanent Error Factor (at System Level:18-22% clipping level) **Meet the State of West Virginia's specifications for Items 3 and 4 including Zero Permanent Errors at 35% clipping level. ================================================================================ Note 182.3 Religious wars, Mag Tape division 3 of 31 EISNER::PROVOST 51 lines 20-JUL-1988 11:03 -< Some experience >- -------------------------------------------------------------------------------- 1. 3400 foot reels quite often have problems on most drives. The tape is too thin, and slips. 2. Blackwatch is backcoated. This is supposed to eliminate some static problems. We have humidity to eliminate static. We have abandoned Blackwatch because of deposits left in drive from flakey coating. Other vendors now produce back-coated tape. I have no information on them. 3. We use about 2000 2400' reels per year. We write each reel once. We may read each one 3 or 4 times. We do not recycle them. The data stored on these tapes represents millions of dollars worth of experiments. If a tape is unreadable, the users have no sense of humor. This usage of tape puts particular demands on the tapes which may differ from your application. 4. We used to write tapes at 1600 BPI on PERTEC T1640 (similar to TU77) drives under RT-11. Error checking and recovery was poor. We developed our own "BURN" program to check for unacceptable tape. 5. All recertified tape which we tested flunked BURN. 6. We used BASF ENDURA for many years without problems. Two years ago the ENDURA began to flunk BURN. We switched to BASF GOLD. Problems vanished. 7. We now use CDC ALPHA, which passes BURN. In spite of the importance of the data, cost is a factor in our decisions. Since we buy in quantity 2000, low cost vendor to us may not be low cost vendor to you. 8. We will accept CDC ALPHA or BASF GOLD. Anything else must pass our BURN tests before we will accept it. 9. We write most tapes to end of tape marker. 10. We clean drives for each tape. 11. We now use only 6250 BPI for data storage. Data is written to STC 1953 or STC 2925 drives via AVIV controllers on VAXstation II's. The GCR format is far less demanding on the media, because it's error recovery is so good. 12. We used to run VMS BACKUP to a TU77 on a VAX-11/780 at 1600 BPI. BACKUPs on BASF ENDURA were later found unreadable, although they had passed VERIFY when written. CDC ALPHA always worked. Now that we BACKUP at 6250 (on STC 2925 drives), everything works. 13. We sometimes need to read tapes written more than 5 years earlier. This introduces an entire new subject which I'd rather not go into in detail in this topic. However, a government study showed that life expectancy of a data tape seems unrelated to the quality of the tape. That is, tapes which have a low clipping level do not necessarily age slower. ================================================================================ Note 182.4 Religious wars, Mag Tape division 4 of 31 EISNER::PROVOST 17 lines 20-JUL-1988 11:10 -< Chart disclaimer >- -------------------------------------------------------------------------------- > < Note 182.2 by EISNER::PROVOST > > -< Possibly obsolete chart >- This chart was given to me by some vendor to provide some objective distinction between the various vendors and their high/low quality tapes. I can't vouch for its accuracy or currency. I don't know how the State of West Virginia sets or tests its specs, but it does seem to provide a cross correlation between vendors which I have found useful in justifying purchase decisions. Clipping level is clearly not the only measure of "goodness", since it would not pick up 3400' reel problems, backcoating flake or the static problems backcoating is supposed to eliminate. I wouldn't mind seeing some more objective criteria for comparing tapes. ================================================================================ Note 182.5 Religious wars, Mag Tape division 5 of 31 EISNER::DELARISCH "Arnold S. De Larisch" 17 lines 20-JUL-1988 13:40 -< WE LOVE BLACKWATCH TAPE! >- -------------------------------------------------------------------------------- Sorry folks, I don't agree with all the BAD press that BLACKWATCH tape gets. We use HUNDREDS of reels of 700 and 701 (Yes thats the 3600' reels) and NOT ONCE have we had a tape we couldn't read! Yes we get LOTS of grief from Field Service about using them ... but we like them! Further, we have had NOTHING but PROBLEMS with DIGITAL's own tape. The stuff is ROTTEN! We keep around just enough for FS to use to 'Test' the tape drive out with. We have trained our OPERATORS to CLEAN the tape drives ... a lot of sites who have problems with BLACKWATCH tapes DON'T keep their drives clean! All right ... I'm ready for the MUD to get slung my way! 8-{) -Arnold ================================================================================ Note 182.6 Religious wars, Mag Tape division 6 of 31 EISNER::STERN "Seth Stern, (216) 266-6721" 8 lines 20-JUL-1988 14:47 -< BLACKWATCH didn't work but Digital did >- -------------------------------------------------------------------------------- >> < Note 182.5 by EISNER::DELARISCH "Arnold S. De Larisch" > >> -< WE LOVE BLACKWATCH TAPE! >- Funny, but we fought with 3M and DEC about BLACKWATCH, and then tried several other tapes with mixed results. We finally gave up and started using Digital magtape and haven't had much trouble since. We use the tapes for backups and consume less than 100 reels a year. It just wasn't worth the expense to try to find the cheapest tape that really works for our volume. ================================================================================ Note 182.7 Religious wars, Mag Tape division 7 of 31 EISNER::HASSINGER "Bob Hassinger" 23 lines 20-JUL-1988 15:41 -< DEC will not stand behind their back coated claims... >- -------------------------------------------------------------------------------- Because our EDP division buys Blackwatch tape in large numbers we have a choice here - Blackwatch for little or nothing or anything else at market price. QED we use Blackwatch. DEC FS people frequently blaim everything in sight on the tape, *verbally*. However, when we have asked them to put it in writing so we could pursue the issue, we have never been able to get anything from them. The best they ever did was to give us a photocopy of some kind of a memo or note saying bad things about the tape but it had NO attribution. No letter head, no names, nothing; totally anonymous (and useless). Until DEC is willing to stand behind their claims that Blackwatch (or any back coated) tape is bad, we will have no choice except to continue using it. Generally the Blackwatch works OK here - almost always as very low number of passes backup tapes BTW. Oddly, the tapes we write are generally quite readable but tapes that come from outside are frequently hard to read on our TU77. We have complained many times and DEC always claims the drive is "fine". Bob H ================================================================================ Note 182.8 Religious wars, Mag Tape division 8 of 31 EISNER::OSUDAR "John Osudar" 10 lines 20-JUL-1988 16:18 -< miscellaneous tape comments >- -------------------------------------------------------------------------------- Only a few comments -- (a) Wabash Quadronix tapes have mucked up our TU77 and TU78 drives more than once; we no longer allow them on our system except when it can't be avoided (and then, we give the drive a doubly-thorough cleaning afterwards; (b) we've used Memorex Cubic HD tape regularly (including some tapes that are reused about once a month) and have had excellent results; (c) our purchasing people attempted to switch to OPUS "CriticalFile" tapes a while back; most of these wouldn't autoload, and they had other problems as well. (We went back to Cubic HD after yelling at purchasing for a while.) ================================================================================ Note 182.9 Religious wars, Mag Tape division 9 of 31 EISNER::KENNEDY "Terry Kennedy" 31 lines 20-JUL-1988 17:38 -< DEC document # disclaiming back-coated tapes... >- -------------------------------------------------------------------------------- > Until DEC is willing to stand behind their claims that Blackwatch (or > any back coated) tape is bad, we will have no choice except to continue > using it. See the following extract from the Pageswapper system: <<< NODE::DISK$PAGESWAPPER:[000000]IO.NOTE;3 >>> -< Submissions to the Pageswapper INPUT/OUTPUT Column >- ================================================================================ Note 971.29 Master versus slave! 29 of 32 NODE::US192903 "Terry Kennedy" 26 lines 30-APR-1988 07:26 -< I hope you have a large dumpster! >- -------------------------------------------------------------------------------- [material not related to back-coated discussion deleted] > We *always* use back-coated tapes. Two references: o "Tape Requirements for IBM 1/2" Tape Units at 800, 1600 and 6250 BPI", IBM Pub. # GA32-0006-5. Page 6, Paragraph 9 - "Fractional Drag - Magnetic Coating to Back Surface" - BW tape fails this specification. Page 7, Paragraph 12 - "Coating In- tegrity Failure" - 3 of 25 reels of BW exhibited failure at or immediately before load point after 100 hours of use o "TU80 Subsystem User's Guide", DEC Pub. # EK-0TU80-UG-002. Page 1-2, "Tape Media Requirements" - "...To date, there are no back-coated magnetic tapes that meet DIGITAL's specification..." ================================================================================ Note 182.10 Religious wars, Mag Tape division 10 of 31 EISNER::KENNEDY "Terry Kennedy" 8 lines 20-JUL-1988 17:42 -< DEC tapes usually = BASF >- -------------------------------------------------------------------------------- If your mind is made up that DEC tape is bad, *don't* read this reply! DEC tape is manufactured for them, not by DEC itself. The actual vendor varies, but is usually BASF. It isn't the world's best grade of tape, but it isn't bad. In fact, this has helped us in the past - we've given the DEC service rep a new reel of DEC tape and said "fix the drive so it will read and write this with no errors". His job would have been a lot easier if DEC's tape was better! ================================================================================ Note 182.11 Religious wars, Mag Tape division 11 of 31 EISNER::MERRIMAN "George Merriman - CCA/NY" 11 lines 20-JUL-1988 17:57 -< check your head alignment >- -------------------------------------------------------------------------------- > Oddly, the tapes we write are generally quite readable but tapes that > come from outside are frequently hard to read on our TU77. We have > complained many times and DEC always claims the drive is "fine". I had a similar problem with a system that would not read DEC software distribution tapes. The FS people would put up a scratch tape and run the exerciser all night with no errors and say the drive was OK. Trouble is the head alignment was bad and this test would not pick it up. I had a lot of trouble convincing FS to run a test against a standard alignment tape, but when they did they found the problem and fixed it. ================================================================================ Note 182.12 Religious wars, Mag Tape division 12 of 31 EISNER::HASSINGER "Bob Hassinger" 23 lines 20-JUL-1988 17:58 -< Yes, but it is still not quite the smoking gun I would need... >- -------------------------------------------------------------------------------- > Note 971.29 Master versus slave! Yes, I recall noticing this at the time. > o "TU80 Subsystem User's Guide", DEC Pub. # EK-0TU80-UG-002. > Page 1-2, "Tape Media Requirements" - "...To date, there are no > back-coated magnetic tapes that meet DIGITAL's specification..." Note we have a TU77, not a TU80, but that aside, no one has ever made "DIGITAL's specification" available to us. Has any one else ever seen it? How does one get a copy? I am unclear about the IBM reference. It is unclear if the writer of the note or the cited reference is saying Blackwatch fails to meet the spec. The most obvious reading is that the spec says you can't have coating integrity failures based on some specified measurement and that the writer of the note is reporting that he observed 3 failures out of 25 reels in his testing and therefore he concludes that BW fails the spec. If, instead, it is actually the IBM spec itself that says that BW fails the spec, then that is more useful when you go starting a fight with city hall... Bob H ================================================================================ Note 182.13 Religious wars, Mag Tape division 13 of 31 EISNER::HUNT_AL "Alan B. Hunt (Ford Motor Co.)" 8 lines 20-JUL-1988 18:04 -< WABASH? Well??? >- -------------------------------------------------------------------------------- Due to the usual purchasing reasons we have been "stuck" with Wabash tapes here. As stated earlier DO NOT USE the Quadronix tapes unless you plan to have a permanent maid per drive to cleanup. The MIRRA 1000 tapes are better. However, we have noticed that new ones tend to shed a bit, too. We tried a box or two of the DEC tapes but the operators complained that they wouldn't load very well. ================================================================================ Note 182.14 Religious wars, Mag Tape division 14 of 31 EISNER::HASSINGER "Bob Hassinger" 20 lines 20-JUL-1988 18:07 -< Yes, head alignment is an 'interesting' subject... >- -------------------------------------------------------------------------------- > -< check your head alignment >- Yes, of course. That was the point of mentioning the many go-rounds with FS about it. We have understood about tape head alignment issues since the TU55 DECtape days and I have explicitly made that part of my discussions with DEC. They keep insisting there is no problem. My guess is that maybe it is not that easy to check and adjust and that may have something to do with them not getting into it. It may be that the FS people depend on having been told the heads "can't" or "only very rarely" go out of alignment on this type drive or something to that effect. BTW, is there any easy test short of getting some kind of a "standard" alignment tape. I can recall seeing some kind of schemes involving writing tapes then flipping them or somehow turning them around but I think that goes back to DECtape days. It is not quite obvious how to make a simple check with only one drive available and no standard. Bob H ================================================================================ Note 182.15 Religious wars, Mag Tape division 15 of 31 EISNER::CLAPP 16 lines 20-JUL-1988 18:19 -< "bad tapes" or bad drives? >- -------------------------------------------------------------------------------- >4. We used to write tapes at 1600 BPI on PERTEC T1640 (similar to TU77) > drives under RT-11. Error checking and recovery was poor. We developed > our own "BURN" program to check for unacceptable tape. > >5. All recertified tape which we tested flunked BURN. So, lot's of other folks have trouble with DEC FS about "bad tape" vs. "bad - under contract - drive". Is there any sort of a standalone box available - say under $10K - that can determine if a given tape is suitable for 6250 use? Something more than a cleaner/counter... Or, I once heard that some part of DEC offered a service to "certify" tapes, leading to the dream of handing the claimed "bad tape" back to local FS a few weeks later with a DEC stamp of approval. Is there an easy answer here? ================================================================================ Note 182.16 Religious wars, Mag Tape division 16 of 31 EISNER::LEDERMAN "Bart Z. Lederman" 16 lines 20-JUL-1988 18:36 -< Reliability depends a lot on you >- -------------------------------------------------------------------------------- I'd have to check, but I believe a lot of systems here use 3M 700 tape (and a number of other types as well). What is important is that we have our own tape cleaning and certifying equipment, and all the important tapes go through it before being re-used. What is even more important is that the critical data systems have tape drives which are maintained in-house. They get regular re-alignment and read and write level adjustments by well-trained personnel. If this doesn't should too impressive I should point out we have two sets of tape drives (in large quantities, we have lots of "one each" drives): Kennedy 9000 series, which the techs like, and DEC TU10 drives (some of which are the old negative bus drives). These drives have horrible reputations in the industry and user community, yet ours work very well indeed. They just get competent and regular maintenance (and by the way, most of them are on line and and have tape loaded in them 24 hours a day, 7 days a week). ================================================================================ Note 182.17 Religious wars, Mag Tape division 17 of 31 EISNER::KENNEDY "Terry Kennedy" 33 lines 20-JUL-1988 22:09 -< Clarification >- -------------------------------------------------------------------------------- > Note we have a TU77, not a TU80, but that aside, no one has ever > made "DIGITAL's specification" available to us. Has any one else > ever seen it? How does one get a copy? It's a DEC STD something-or-other. Some of these are available through Documentation Services - I'll try to post the address if I can find it. However, some DEC STD items are hopelessly out of date by the time you get a copy (such as the DCL specs), and others are not available outside of DEC without a major hassle (MSCP protocol, LAT protocol, BI specs). > I am unclear about the IBM reference. It is unclear if the writer of > the note or the cited reference is saying Blackwatch fails to meet the > spec. As the original author of the note on the Pageswapper, let me clear this up. The IBM standard quotes are verbatim from the referenced IBM document. The tests were done by an independent testing orgranization for a major computer company, which at the time was going for bids on about a quarter million reels of tape annually. I'm sorry, but I can't tell you when or who, but I can say that the tests still accurately reflect the 3M product as it is being marketed today. Regarding head alignment, and other adjustment tapes you need for about $100 to $350 total. You need: a) alignment tape (sometimes called Master Skew Tape) b) Reference level tape With these you can determine both the head alignment *and* the signal level being recovered by the head. Both are important. You may also need specialized test fixtures to do the adjustment - that depends on the drive in question. 'Smart' formatters like the TU78's automate most of the testing for you. ================================================================================ Note 182.18 Religious wars, Mag Tape division 18 of 31 EISNER::MERRIMAN "George Merriman - CCA/NY" 11 lines 21-JUL-1988 03:17 -< head alignment is a pain >- -------------------------------------------------------------------------------- > My guess is that maybe it is not that easy to check and adjust and that > may have something to do with them not getting into it. That's the reason I had so much trouble getting them to do it -- they considered it a real pain in the you-know-what. I remember that the adjustment was electronic rather than mechanical as on an audio tape drive -- something to do with adjustable delay lines or one-shots in the read and write electronics. A good dual-channel 'scope with delayed sweep, etc. was needed -- the biggest problem the techs has was figuring out how to set up the 'scope. had was figuring out how to ================================================================================ Note 182.19 Religious wars, Mag Tape division 19 of 31 EISNER::KENNEDY "Terry Kennedy" 10 lines 21-JUL-1988 04:40 -< Some mechanical adjustments still required >- -------------------------------------------------------------------------------- > adjustment was electronic rather than mechanical as on an audio tape > drive -- something to do with adjustable delay lines or one-shots > in the read and write electronics. That was the PE deskew adjustment, which is only one part of a complete alignment. Different transports and densities have different requirements for alignment, but there are *always* at least two mechanical adjustments which need to be made initially. They should not drift unless there has been some shock to the transport. Fiddling the electronics can compensate for them to some extent, but isn't the same... ================================================================================ Note 182.20 Religious wars, Mag Tape division 20 of 31 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 19 lines 21-JUL-1988 08:45 -< TU78 tape cleaner >- -------------------------------------------------------------------------------- Re: 182.13 I work at a different site at FoMoCo than Alan Hunt and we also use Wabash tapes. The funny thing is that we used Wabash Quadronix tapes for YEARS without any serious problems (my daily tapes are about five years old and log almost no errors). Being a one man show, I was not always religious about cleaning the heads either ! Now suddenly, brand new Quadronix tape are logging lots of errors and the heads need to be cleaned after every tape (if it can make it through the tape!). This, coincidentally happened just after or FE upgraded my TU78. Alan is right, the Wabash Mirra 100 tapes seem better, but the Quadronix tapes seem okay after 4 or 5 passes ! ================================================================================ Note 182.21 Religious wars, Mag Tape division 21 of 31 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 25 lines 22-JUL-1988 20:04 -< The (slipping) REEL real reason >- -------------------------------------------------------------------------------- Aside from the high levels of crud some people complain of when using backcoated tapes, there is a very simple and valid reason NOT to use them on SOME drives. The CDC "Keystone" streamers (TU80, TU81 to those who buy them the expensive way) don't have any capstan at all, but rely on clever microcode figuring out just how to crank the reel motors to keep a constant tape velocity past the heads even though the circumference of the tape on the reel keeps changing. This is all happening while the tape tension is monitored (I think) by the back pressure on the air supply to the air bearings the tape floats on as it turns the corner heading for the head which is the ONLY solid thing the tape touches. Now try doing all that when the layers of tape start slipping and the torqued reel can turn tightening the pack of tape on it but not moving the expected length of tape off the other reel and past the head! Then remember how often these drives reverse and reposition even with /buf=5. The backcoated tapes slip against themselves much more easily than regular tape. Drives that have capstans and vacuum column buffers simply don't have this kind of problem. ================================================================================ Note 182.22 Religious wars, Mag Tape division 22 of 31 EISNER::FULLER_S "Stuart Fuller" 24 lines 29-JUL-1988 22:13 -< Skew adjustment is mechanical! >- -------------------------------------------------------------------------------- > That's the reason I had so much trouble getting them to do it -- they > considered it a real pain in the you-know-what. I remember that the > adjustment was electronic rather than mechanical as on an audio tape > drive -- something to do with adjustable delay lines or one-shots > in the read and write electronics. A good dual-channel 'scope with > delayed sweep, etc. was needed -- the biggest problem the techs > has was figuring out how to set up the 'scope. Head alignment on the TU77/78/79 is a strictly mechanical adjustment. There are 2 adjustments. The first is to make sure the head is square to the tape path, such that the signal amplitude between for and reverse is the same. If not, the head must be "twisted" back or forth to get it right. The second adjustment involves making sure the head is square to the tape path, such that all 9 bits of a tape character hit the head at the same time. This adjustment requires the master skew tape and both require a scope. AS for the elctronic adjustment you're thinking of, I remember that the TU77 has a write deskew buffer that you adjust with one-shots. The TU78/79 has no such adjustment (but it does have the GCR write current adjustment, which must be right or you'll have GCR problems. Stu ================================================================================ Note 182.23 Religious wars, Mag Tape division 23 of 31 EISNER::MCMICHAEL "Chuck McMichael" 13 lines 10-OCT-1989 14:32 -< 3600' REELS >- -------------------------------------------------------------------------------- Let's see if I can jump start this topic again. Our installation has been using Black Watch tape successfully for a number of years. Lately, the amount of data we've been pushing through the tape drive has increased dramatically and so has the number of times Digital has had to replace our read/write head. Assuming for the sake of argument that backcoated tapes are a problem, let's go back to first principles. The reason we went with Black Watch is that the tapes are available in 3600' reels. DEC's are a maximum of 2400'. When you only have one drive, the fewer tapes you have to load, the better. Is there anything I can do to convince DEC to sell longer tapes (and save money on replacing heads under a service contract)? Is there anyone who sells non-backcoated tapes of good quality in 3600' lengths? ================================================================================ Note 182.24 Religious wars, Mag Tape division 24 of 31 EISNER::MERRIMAN "George Merriman - CCA/NY" 4 lines 10-OCT-1989 16:03 -< I don't see how this follows >- -------------------------------------------------------------------------------- > Is there anything I can do to convince DEC to sell longer tapes (and > save money on replacing heads under a service contract)? ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ================================================================================ Note 182.25 Religious wars, Mag Tape division 25 of 31 EISNER::KENNEDY "Terry Kennedy" 44 lines 10-OCT-1989 20:26 -< Some answers >- -------------------------------------------------------------------------------- > Let's see if I can jump start this topic again. Our installation has been using > Black Watch tape successfully for a number of years. Lately, the amount of data > we've been pushing through the tape drive has increased dramatically and so has > the number of times Digital has had to replace our read/write head. What model drive(s) do you have? What sort of 'dramatic' increas are we talk- ing about here (10x, 100x, ?). Be aware that heads *do* wear out eventually, but can age prematurely due to a number of things: 1) excessive abrasion 2) incorrect installation/adjustment 3) defective drive logic 4) improper cleaning procedures or chemicals What is the type of failure? If it can *read* but not *write*, I'd suspect #3. If you have a TU77/78, be aware that many service techs think the head change is a plug-and-play. It isn't. Lomg-term correct operation requires careful adjustment. > Assuming for the sake of argument that backcoated tapes are a problem, ... Backcoated tapes are *a* problem. They may not be *your* problem, however. Backcoated tapes are exactly that - coated on the *back*. They should *not* shed enough to contaminate the data surface (and thus the head) under normal operating and storage conditions. However, out-of-tolerance storage (see the tape box for the vendor's specs) *or* a misadjusted tape drive can cause large quantities of the coating to come off in the drive. Is there one area of the drive which seems to collect a lot of coating dust? There may be a problem with mechanical alignment there. > The reason we went with Black Watch is that the tapes > are available in 3600' reels. See note 213.23 for more info on 3600' tapes by vendors other than 3M. Please remember, however, that longer tapes have reduced storage lifetimes, as well as being more easily stretched during processing. > Is there anything I can do to convince DEC to sell longer tapes (and save money > on replacing heads under a service contract)? I doubt it. Service pricing is set based on cost-to-repair averages plus over- head and profit. They may raise the service cost for your model tape drive, but I doubt they'd offer longer tapes. DEC *does* disclaim the use of 3600' tapes on software grounds, by the way. ================================================================================ Note 182.26 Religious wars, Mag Tape division 26 of 31 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 3 lines 10-OCT-1989 23:54 -< Watch that humidity! >- -------------------------------------------------------------------------------- A number of years back I attended a DECUS session in which the speaker (a DECie) mentioned that they had found excessive tape head wear in computer rooms which had humidity BELOW 45%. ================================================================================ Note 182.27 Religious wars, Mag Tape division 27 of 31 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 53 lines 11-OCT-1989 11:52 -< HIGH humidity INCREASES tape head wear >- -------------------------------------------------------------------------------- > A number of years back I attended a DECUS session in which the > speaker (a DECie) mentioned that they had found excessive tape > head wear in computer rooms which had humidity BELOW 45%. I have been accused of many things, but never of being a DECie! :-) If the session you attended was given by DEC, then once again DEC and I disagree. :-) At the Spring 1983 Symposium in St. Louis, I discussed a study done by NASA. It is reported in NASA Reference Publication #1075 (April 1982) entitled "Magnetic Tape Recording for the Eighties". They aimed most of their research and the report at high-speed instrumentation tape, but most of it is applicable to the tape we use. I strongly suggest that anyone seriously interested in understanding mag tape get a copy of this 176 page report. Head wear is a complex function of many factors: Abrasivity, Contamination, Humidity, the "Brown Stain" phenomenon, and Lubrication. This subject alone takes 16 pages of the book. I will quote just a portion of it: "Empirical results have shown that head wear is a function of the relative humidity (RH) of the environment in which the head operates. When RH is GREATER [my emphasis] than 40%, head wear dramatically increases. It is unresolved as to whether increasing RH causes the abrasiveness of the tape to increase, or increases in RH accelerate breakdown in the molecular structure of the head and tape surface." The curve they show of head wear vs RH shows a relatively flat curve from 2% to about 40%, then a *dramatic* increase beyond that. It looks almost like two straight-line segments joining at about 42%. If you looked only at this curve, you might think that the lower the RH, the better. However, several pages later, the book goes on to say: "The converse of the humidity consideration -- that is, the lower the humidity, the more optimum the operating environment -- does not hold true either. In conditions of very low humidity, the phenomenon known as brown stain occurs. (With some head material the stain appears to be more of a blue color, but the problem remains the same.) ... Brown stain on tape heads is a phenomenon that has been observed for 20 years. It should not be confused with ordinary debris that is easily removed with a cotton-tipped applicator and head cleaning solvent. Brown stain can only be removed by mechanical means. ... There is an inverse relationship between brown stain and head wear." Please remember that tape and head wear is a complex function of several factors that are difficult to summarize, so do not leap to any conclusions based on this short note. ================================================================================ Note 182.28 Religious wars, Mag Tape division 28 of 31 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 4 lines 11-OCT-1989 15:00 -< Apologies! >- -------------------------------------------------------------------------------- Woops.... my mistake! That sounds like the session that I attended. I may have meant to call Alan many things, but never a DECie ;-} Ahem.... It's been a long time and my mind is fading fast. Guess I was distracted by all the DEC 10/20 system engineer bodies ;-} ================================================================================ Note 182.29 Religious wars, Mag Tape division 29 of 31 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 7 lines 11-OCT-1989 20:53 -< I think it was the Hardware Hints & Kinks session >- -------------------------------------------------------------------------------- > It's been a long time and my mind is fading fast. Guess I was > distracted by all the DEC 10/20 system engineer bodies ;-} My mind is fading fast too, but it is difficult to forget everyone laughing as I tried to describe the problem of having a brown stain on your head. You hardware folks must have dirty minds. :-) ================================================================================ Note 182.30 Religious wars, Mag Tape division 30 of 31 EISNER::MCMICHAEL "Chuck McMichael" 11 lines 13-OCT-1989 16:36 -< Details >- -------------------------------------------------------------------------------- > What model drive(s) do you have? What sort of 'dramatic' increas are we talk- >ing about here (10x, 100x, ?). Be aware that heads *do* wear out eventually TU81-Plus. We've gone from monthly backups of 3 RA60s to daily incremental backups of 2 RA81s and 6 RA60s. (The information explosion, don't cha know?) > See note 213.23 for more info on 3600' tapes by vendors other than 3M. Please >remember, however, that longer tapes have reduced storage lifetimes, as well >as being more easily stretched during processing. But are they backcoated? ================================================================================ Note 182.31 Religious wars, Mag Tape division 31 of 31 EISNER::KENNEDY "Terry Kennedy" 12 lines 14-OCT-1989 02:58 -< More comments >- -------------------------------------------------------------------------------- > TU81-Plus. We've gone from monthly backups of 3 RA60s to daily incremental > backups of 2 RA81s and 6 RA60s. (The information explosion, don't cha know?) 6 tapes/month to maybe 240. I can see you getting a bit more head wear, but not enough to destroy the heads, under normal conditions (as noted in my pre- vious reply). Back-coated tapes are a definite no-no on TU8x drives. > But are they backcoated? Nope. As far as I know, backcoating is a 3m affliction only.  ================================================================================ Note 183.0 Digital has it now ! 15 replies EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 22 lines 21-JUL-1988 16:04 -------------------------------------------------------------------------------- From a press release on the Digital Electronic Store dated July 19,1988. Digital's storage capabilities also were enhanced by the introduction of Digital's new high-performance SA600 Storage Array, which has a storage capacity of 9.7 billion bytes. Its thin-film technology, used in the system's heads and disks, permits large data capacity, fast access times, and high reliability, all important in transaction processing applications. . . . The SA600 is available in two configurations. The fully configured, 9.7 billion-byte system is priced at $215,000. A partially configured system with half that capacity is priced at $115,000. The SA600 is available now. I'm guessing that this is made up of 8 RA90's. The cost per Mbyte is $$22.16 versus $26.40 for a SA482 !! It should be also noted that the electroinc store lists the prices as $225,00 and $125,000 so who is correct ? ================================================================================ Note 183.1 Digital has it now ! 1 of 15 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 8 lines 21-JUL-1988 16:19 -< How fast does it get from there to here? >- -------------------------------------------------------------------------------- >> ...which has a storage capacity of 9.7 billion bytes. >> Its thin-film technology, used in the system's heads and >> disks, permits large data capacity, fast access times, and >> high reliability I don't see anything about transfer rates. Anybody know? Alan ================================================================================ Note 183.2 Digital has it now ! 2 of 15 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 10 lines 21-JUL-1988 21:29 -< Hopefully E-store will have info soon >- -------------------------------------------------------------------------------- > < Note 183.1 by EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" > > -< How fast does it get from there to here? >- >I don't see anything about transfer rates. Anybody know? As of this afternoon, the only info on the E-Store about DECtp and this new mini-disk-farm (and VIDA) was in the form of the press releases dated 7/19. Each one said something to the effect of "see attached brochures for additional information". I'm assuming (hoping) that the E-store just hasn't had time to put them up yet; the releases contained very little technical information. ================================================================================ Note 183.3 Digital has it now ! 3 of 15 EISNER::NAGY "Frank J. Nagy, VAX Guru" 10 lines 22-JUL-1988 08:33 -< Still a long way to go >- -------------------------------------------------------------------------------- > < Note 183.1 by EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" > > -< How fast does it get from there to here? >- >I don't see anything about transfer rates. Anybody know? Same as the RA82 I believe. However, DEC is still too-little-too-late: checkout prices on CDC's new Wren-V with 638 MB in 5.25" drive. Glad that my group is using QBus systems and so can buy controllers and massive disks at other than DEC's ripoff prices... ================================================================================ Note 183.4 Digital has it now ! 4 of 15 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 40 lines 22-JUL-1988 09:57 -< SA600 / RA90 SPECS >- -------------------------------------------------------------------------------- SPECIFICATIONS SA600-JA SA600-HA Performance: (1) Number of RA90 actuators Eight Four Transfer rate 2.8 MBytes/second 2.8 MBytes/second Disk subsystem response time (2) 23.5 milliseconds 23.5 milliseconds Average seek time (3) 17.5 milliseconds 17.5 milliseconds Media Characteristics: Formatted capacity 9.728 GB 4.864 GB Unformatted capacity (4) 12.863 GB 6.431 GB Tracks per inch 1750 1750 Bits per inch 22,839 22,839 Rotational speed 3600 rpm 3600 rpm Number of data surfaces 104 (13/drive) 52 (13/drive) Number of heads per surface 1 1 Sectors per track 69 + 1 spare 69 + 1 spare Tracks per logical cylinder 13 13 Power Requirements: (5) Power receptacle type option -xA L21-30R L21-30R option -xD IEC 309 IEC 309 Heat dissipation 7900 Btu/h 3950 Btu/h 2314 watts 1157 watts (1)Performance specifications (and others as indicated) are provided on a per-drive basis. All others apply to a fully configured array. (2)For a typical VAX I/O environment. (3)With specification margin and subsystem overhead--19.0 milliseconds. (4)Unfomatted capacity provided for comparison purposes; only formatted capacity is user accessible in any disk drive. (5)Consult the Storage Array Family Information Sheet for additional data. ================================================================================ Note 183.5 Digital has it now ! 5 of 15 EISNER::BYRNE_C "Charlie Byrne, Freeport NY" 7 lines 22-JUL-1988 10:10 -< I like it! >- -------------------------------------------------------------------------------- > < Note 183.4 by EISNER::KILLEEN "Jeff Killeen DECUServe Chair" > > -< SA600 / RA90 SPECS >- >SPECIFICATIONS > SA600-JA SA600-HA > Average seek time (3) 17.5 milliseconds 17.5 milliseconds >(3)With specification margin and subsystem overhead--19.0 milliseconds. ================================================================================ Note 183.6 Digital has it now ! 6 of 15 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 7 lines 22-JUL-1988 13:13 -< Gotta watch those asterisks! >- -------------------------------------------------------------------------------- > Average seek time (3) 17.5 milliseconds 17.5 milliseconds >(3)With specification margin and subsystem overhead--19.0 milliseconds. It's beginning to seem like stuff you see on the inside back cover of Consumer Reports, the "Selling It" column. Alan ================================================================================ Note 183.7 Digital has it now ! 7 of 15 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 9 lines 22-JUL-1988 20:19 -< Yes! look at new CDC drives >- -------------------------------------------------------------------------------- >> However, DEC is still too-little-too-late: checkout prices on CDC's >> new Wren-V with 638 MB in 5.25" drive. Glad that my group is using >> QBus systems and so can buy controllers and massive disks at other >> than DEC's ripoff prices... I'll second that, but for more capacity look at the CDC 8" 1.23G drives and run them on a 4 port Dilog DQ256 (or DU256 for Unibus). 4 drives + one controller of this (or 8 + 2) looks remarkably similar in capacity to Dec's new offering, but the price does seem to be a bit different... ================================================================================ Note 183.8 Digital has it now ! 8 of 15 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 9 lines 22-JUL-1988 20:59 -< WHEN WREN? >- -------------------------------------------------------------------------------- > However, DEC is still too-little-too-late: checkout prices on CDC's > new Wren-V with 638 MB in 5.25" drive. Glad that my group is using As I start this topic off track.... I assume these are 5 1/4 drives with optional ESDI or SCSI interfaces? Anything funny about power that keep me from putting one in a BA23? ================================================================================ Note 183.9 Digital has it now ! 9 of 15 EISNER::KENNEDY "Terry Kennedy" 8 lines 22-JUL-1988 21:37 -< Apples and kiwi fruit >- -------------------------------------------------------------------------------- > However, DEC is still too-little-too-late: checkout prices on CDC's > new Wren-V with 638 MB in 5.25" drive. Glad that my group is using Somehow, I don't see some dozens of Wrens attached to a cluster of 88x0 processors. Compare instead the SA600/RA90 with the IBM 3380 models. The SA600/RA90 doesn't seem to be intended for the small sys- tem market, where DEC currently has nothing outstanding in price/per- formance. However, I'll bet that is the next area to be addressed. ================================================================================ Note 183.10 Digital has it now ! 10 of 15 EISNER::NAGY "Frank J. Nagy, VAX Guru" 14 lines 24-JUL-1988 10:54 -< DEC losing ground on low end disk/tape >- -------------------------------------------------------------------------------- >> I assume these are 5 1/4 drives with optional ESDI or SCSI interfaces? Yes. We are using (er, will be using) the SCSI version with a disk/tape controller (Summus 442). Only item to be resolved is mounting hardware and we believe we have that one (skid plane) in hand/on order. >> Anything funny about power that keep me from putting one in a BA23? I don't have the exact specs, but the Wren-V seems to use *less* power than an RD54! In fact we are configuring a couple of VS3200s with Summus-442, Summus Gigatape (8mm helical) and Wren-V as opposed to TQK50/TK50 plus RQDX3/RD54. Less power used, less cost, faster tape, greater disk capacity (almost exactly 4x)... Digital certainly seems to be falling behind on the low end. ================================================================================ Note 183.11 Digital has it now ! 11 of 15 EISNER::ERSKINE "Chris Erskine" 11 lines 25-JUL-1988 07:09 -< DEC never had it for the low end >- -------------------------------------------------------------------------------- >> Anything funny about power that keep me from putting one in a BA23? The question should also include controller. Will the controller handle the drive? > greater disk capacity (almost exactly 4x)... Digital certainly seems > to be falling behind on the low end. I never knew DEC had it in the low end. Their Q-Bus storage has never been what it could and should have been. From DEC I can get the power for the small systems but I can not put enough disk on it. ================================================================================ Note 183.12 Digital has it now ! 12 of 15 EISNER::COY "Dale E. Coy" 1 line 25-JUL-1988 17:39 -< How many SDI ports? >- -------------------------------------------------------------------------------- How many SDI "ports" on a controller does a SA600 take? ================================================================================ Note 183.13 Digital has it now ! 13 of 15 EISNER::NAGY "Frank J. Nagy, VAX Guru" 12 lines 26-JUL-1988 07:59 -< Lots and lots of ports >- -------------------------------------------------------------------------------- >> How many SDI "ports" on a controller does a SA600 take? The full SA600 (9.7 GB) appears to take 8 SDI ports. The half-full SA600 would then take 4 ports. At least according to my reading of some literature our Sales Rep dropped by. If I'm wrong, I'm sure someone will correct me (with glee, I'm sure). There is at least one 3rd party company (name escapes me now) which is offering an SDI-compatible disk system called a RACE. This puts 2.2 GB of disk space on a single SDI-port; this unit is made up of several small (5.25") high capacity disk drives which appear to be a single disk to the SDI port. ================================================================================ Note 183.14 Digital has it now ! 14 of 15 EISNER::BLOOM "Bob Bloom, Raytheon Company" 9 lines 2-AUG-1988 14:36 -< Name of vendor with 2.2GB per port >- -------------------------------------------------------------------------------- > There is at least one 3rd party company (name escapes me now) which > is offering an SDI-compatible disk system called a RACE. This puts > 2.2 GB of disk space on a single SDI-port; this unit is made up of > several small (5.25") high capacity disk drives which appear to be > a single disk to the SDI port. Micro Technololgy, Inc. has the kind of unit described. 2.2GB per SDI-port with I believe three 5.25" drives made to appear as one. ================================================================================ Note 183.15 Digital has it now ! 15 of 15 EISNER::CLAPP 8 lines 9-AUG-1988 17:34 -< pseudo-Sahdow at low end? >- -------------------------------------------------------------------------------- > Micro Technololgy, Inc. has the kind of unit described. 2.2GB per SDI-port > with I believe three 5.25" drives made to appear as one. Has anybody heard anything about products using multiple 5.25" drives in a package that includes redundancy such that a single drive failure is transparent to the user? I believe Micropolis or Maxtor had claimed they were working on this - anything real going on?  ================================================================================ Note 184.0 This is an UPgrade? No replies EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 25 lines 21-JUL-1988 16:10 -------------------------------------------------------------------------------- The following is extracted from the VMS v5.0 Cover letter (AV-EF56G-TE): Potential problems with TQK70 tape controller when loading VMS version 5.0: The TQK70 rev A3 controller may fail to boot the VMS Version 5.0 Standalone Backup TK50 tape. If you have the TQK70 rev A3 controller and this happens during the installation of VMS Version 5.0, there are two ways to solve the problem: 1. If you have the VMS v4.6 binary TK50 tape, boot that version of standalone backup. ... 2. Or, contact Digital Field Service to upgrade the TQK70 rev A3 controller to the TQK70 rev B3 controller which fixes the boot problem. Note that the TQK70 controller with rev B3 takes approximately 80 minutes to boot standalone backup. ^^ !! 80 minutes to boot standalone backup!?!?!? That is worse than TU58s on an 11/730. What is going on here? Alan  ================================================================================ Note 185.0 HSC discussion No replies EISNER::MERRIMAN "George Merriman - CCA/NY" 2 lines 21-JUL-1988 22:31 -------------------------------------------------------------------------------- Those of you with an interest in HSC70's might want to check out topic 82 in DEC_SOFTWARE.  ================================================================================ Note 186.0 LP11 Cabinet kit part number needed (CK-LP???-??) 3 replies EISNER::KENNEDY "Terry Kennedy" 12 lines 25-JUL-1988 00:55 -------------------------------------------------------------------------------- Here's a new stump-the-band item: What is the part number for an LP11 (Unibus LP25 controller) cabinet kit for an FCC-compliant cabinet? For some reason *no* DEC literature lists this item seperately - I know it is included in the LP25 product, but I don't *need* another printer and controller - just the cabinet kit! I really wish DEC would settle on either requiring a cabkit order with the product *or* always including the cabkit - right now it is a little of both. ================================================================================ Note 186.1 LP11 Cabinet kit part number needed (CK-LP???-??) 1 of 3 EISNER::NORTON "Bill Norton" 3 lines 25-JUL-1988 09:47 -< Not stumped for long! >- -------------------------------------------------------------------------------- How about #70-18953-01 from Unlisted Parts. We bought this for $795 in 1985 when we wanted to switch an LP25 between two 11/44's, and ordering an M7258 didn't get us the cabinet kit. ================================================================================ Note 186.2 LP11 Cabinet kit part number needed (CK-LP???-??) 2 of 3 EISNER::KENNEDY "Terry Kennedy" 6 lines 26-JUL-1988 00:20 -< Are we talking the same item? >- -------------------------------------------------------------------------------- > How about #70-18953-01 from Unlisted Parts. We bought this > for $795 in 1985 when we wanted to switch an LP25 between... Ugh! Let me get this straight: $795 for a 40-pin ribbon to DB-37 panel, 15' of ribbon cable, and a mounting bracket? Are we talking the same item? ================================================================================ Note 186.3 LP11 Cabinet kit part number needed (CK-LP???-??) 3 of 3 EISNER::NORTON "Bill Norton" 2 lines 26-JUL-1988 09:24 -< ...afraid so... >- -------------------------------------------------------------------------------- The packing list is long gone, so I can't be positive, but that sounds about right. There might have been a BC27 included too.  ================================================================================ Note 187.0 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 15 replies EISNER::HASSINGER "Bob Hassinger" 57 lines 25-JUL-1988 16:15 -------------------------------------------------------------------------------- We have an 11/750 (no cluster). Quite a long way back the famous ECO/FCO was installed that, among other things, installed a programmable control store. Field service provided some PCS750 software with this. You call a small command file from SYSTARTUP.COM to get a loader program to load the PCS microcode. There seem to be a few problems with this now that I am trying to install a virgin VMS V5.0 system. First, I don't believe they gave us a software kit, just a copy of the files. What I can find is files on disk from an old VMSUPDATE style kit. To simply reinstall them on a new system disk appears to require they be on a proper TU58 rather than on disk. To avoid this I just read the command files and did what they did, but, the result does not work on the V5 system so far. There are two versions of this package, one for a normal 11/750 and one for an 11/750 with battery backup. In the first case you just have a command file, loader program and .BIN file of the microcode. In the second case you add a "device driver" that gets installed in your system, as I recall to help handle warm restarts. Now, don't you know, we have the battery backup. Of course, V5 requires driver changes. The kit I have includes a .OBJ file for the driver so I tried relinking it. However, VMS V5 still refuses to install it, complaining of version mis-match. At that point I go to plan B which is to pretended I don't have battery backup. Since I am trying to do a virgin install I must put new copies of the files in the specified places. After doing that per the old command file the loader program still fails to work. Claims it can not find the binary file (I assume it means the .BIN file). This last may have something to do with the new SYS$COMMON stuff and so on and might (ought?) to be solvable but it really does not resolve my real need anyway which is to solve thebattery backup case. What I don't understand is that I can't find anything in the release notes or the processor specific installation information about this issue. It seems that nearly every 11/750 must have this FCO by now (was it *required* for a previous release of VMS perhaps?) so someone should know what is going on. I called CSC and they claimed no one knew anything and after researching the question said all I could do was call my local field service. I have called but I don't know when I may get a good answer from them. As I recall I went around this before, maybe around VMS V4.0, with less than fully satisfactory results ("what, software? no field service doesn't have anything to do with software...", "what is VMSUPDATE?", etc.) so I am not holding out much hope in that direction at this point. Someone is going to have to really figure out and document this business of field service getting involved in software updates for hardware. I see it coming up more and more and there seems to be a serious gap someplace in the system. Bob H ================================================================================ Note 187.1 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 1 of 15 EISNER::TANNENBAUM "Saul Tannenbaum" 16 lines 25-JUL-1988 17:59 -< PCS750 >- -------------------------------------------------------------------------------- As an old 750 user (clustered), let me offer the following, somewhat vague recollections. At some point during the VMS 4.n cycle, the old PCS microcode stuff was rendered obsolete with the microcode patching being handled by VMB from the TU58. The reason for this was that, in a cluster, you may well need the patches to get the CI-750 operative and, if the patches were on an HSC based disk, you were stuck. It also made the patches operative for say, standalone backup. At some point, we received updated console media that contained the patch file. I'd suggest looking at your console TU-58, seeing if there is a PCS file there, and if so, deleting all that stuff from your system startup. Saul ================================================================================ Note 187.2 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 2 of 15 EISNER::ERSKINE "Chris Erskine" 8 lines 26-JUL-1988 07:58 -< Field Service does software >- -------------------------------------------------------------------------------- > Someone is going to have to really figure out and document this > business of field service getting involved in software updates for > hardware. I see it coming up more and more and there seems to be > a serious gap someplace in the system. DEC already handled this. Field Service now does both hardware and software. This happened after 4.0 came out (I think it was about 2 years ago). ================================================================================ Note 187.3 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 3 of 15 EISNER::HASSINGER "Bob Hassinger" 19 lines 26-JUL-1988 11:52 -< Have NEVER seen a new TU58 & NEVER use them >- -------------------------------------------------------------------------------- 1) I know field service is *supposed* to do both hardware and software. I am talking about reality. 2) I never saw an upgraded console TU58. Nothing in what I have has anything to do with it. In fact, how *does* one every get a new console media. The one I am using still derives from the one that came with the system at VMS V3.0 with possible additions and changes at various points along the way since. I have never seen a new one like we get for BACKUP from time to time and no instructions for creating one from scratch, only instructions on updating the one we already have. 3) Since we do not have an HSC we NEVER use the TU58 at all. It NEVER gets involved in system booting. Even for standalone backup we use copies built in a SYSE root on both the system disk and on another drive so both saves and restores can be done without every waiting an hour or so for the TU58s. Bob H ================================================================================ Note 187.4 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 4 of 15 EISNER::HASSINGER "Bob Hassinger" 37 lines 26-JUL-1988 15:38 -< "Rev 7" vs "rev 8" - this is getting out of hand... >- -------------------------------------------------------------------------------- Now I get another story - this time from Colorado hardware support to whom I was finally referred by my local field service backup person when I finally reached him (i.e. - local knows nothing about it; as usual). It seems that the famous 750 ECO that installed the PCS was rev 7. Apparently that is still the rev level my system is at. It also seems that since then there has been another ECO/FCO that gives you rev 8. I am told you can distinguish this by the fact that now when you boot you get "@@" rather than the old "%%" that I am still getting with rev 7. It seems rev 8 introduced additional, considerable, changes including something to FORCE you to boot from a TU58 instead of direct from disk and in the deal you also get loading of the PCS microcode off the tape. The person I have been talking to is still researching all this to see if it is really true. He suspects this may be the answer to my problem - i.e. - I have to have rev 8 to work with VMS V5 because that is the way the microcode is now supposed to get loaded. It is not at all clear this is really the case because the main problem I see is with "device driver" that I believe is designed to catch a warm restart after a battery backed power outage in order to be sure the PCS code is reloaded. I don't think anything about a TU58 boot such as he described will solve the V5 driver update problem. It would seem that unless rev 8 eliminated the warm restart issue in some way that an updated device driver (PCS750RES.EXE) is still needed. Does anyone know the straight story, particularly for the non-CI, non-HSC case? I sure don't want to have to involve the TU58 in my boot, it takes quite long enough already. I wonder, was rev 8 an "optional" rev that was installed only in the HSC case by any chance? I know I don't recall ever being offered a chance to have the ECO installed and I don't think the VMS V5.0 release notes say anything about a required rev level (I think perhaps VMS V4.0 *did* document a requirement for the rev *7* change). Bob H ================================================================================ Note 187.5 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 5 of 15 EISNER::TANNENBAUM "Saul Tannenbaum" 11 lines 26-JUL-1988 18:20 -< Here a Rev, there a Rev, everywhere a... >- -------------------------------------------------------------------------------- My 750 went off DEC maintenance in February, but I'd never, ever heard of a Rev. 8. Having just spent some time plowing through old TU58s in the (vain) hope of jogging my memory, I can only say that my recollection was that at some point in my cluster installation (the 750 used to be non-clustered) Field Service provided an updated console cassette with the PCS stuff on it (CI microcode is also loaded from the console cassette, so that upgrade was necessary to get the cluster operative). Remembering how difficult it was to get a straight story on Rev. 7, you have my sympathies. ================================================================================ Note 187.6 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 6 of 15 EISNER::ERSKINE "Chris Erskine" 3 lines 27-JUL-1988 07:05 -< You have to put in all of the rev's >- -------------------------------------------------------------------------------- Your best bet would be to put in rev 12843. This rev replaces the current cabinet with a BA123 and a new CPU board. Of course this is not covered under any contract but you end up with a faster CPU. ================================================================================ Note 187.7 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 7 of 15 EISNER::HASSINGER "Bob Hassinger" 20 lines 27-JUL-1988 10:55 -< We are thinking about a "rev" like that but... >- -------------------------------------------------------------------------------- > Your best bet would be to put in rev 12843. This rev replaces the > current cabinet with a BA123 and a new CPU board. Actually I *am* trying for a slightly different "rev" that installs a 6210 but it looks like there still has to be a considerable period when both it and the 750 are running, maybe even LAVc'ed together - so - I think I still have to solve this problem. BTW - both "revs" have the same problems - no Unibus, no Massbus, so integrating them into the existing environment is non-trivial - for example an important part of my user base connects via a DCA stat-mux system that interfaces with the 750 via a Unibus board and there is no practical Qbus or BI replacement. There are several other devices we can't or don't want to give up. The Unibus box for the 6210 does not look attractive and does not solve the Massbus part anyway... What is really strange is that I can't be the only one worried about this PCS750 issue (although the main problem may be limited to non-CI 750s with battery backup which does narrow down the field a good bit). Bob H ================================================================================ Note 187.8 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 8 of 15 EISNER::HASSINGER "Bob Hassinger" 18 lines 29-SEP-1988 12:20 -< PCS750_V5 appears >- -------------------------------------------------------------------------------- Update - we "escalated" this issue a couple of levels and made a good bit of noise - that and the fact our company is now a "named account" or something to that effect finally got some attention. Tuesday our service rep and a higher level support person showed up with a mag tape that had what looked like a not quite ready for release PCS750 kit for V5 on it. We installed it and things seem fine now. Interestingly all the dates on the files were back in the Spring. It looks like somebody dropped the ball along the way after the software was developed and before it was released to the field and/or Colorado and Field Service were told about it. While they were here they confirmed that the battery back up hardware was dead however which I have been telling them for over a year - but that's another story... (does *your* Field Service rep check *your* battery backup on each PM?) Bob H ================================================================================ Note 187.9 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 9 of 15 EISNER::KENNEDY "Terry Kennedy" 5 lines 29-SEP-1988 20:07 -< Don't believe the BATT light... >- -------------------------------------------------------------------------------- > (does *your* Field Service rep check *your* battery backup on each PM?) And the BATT light is not a good indicator - failing batteries may appear good until called into action. Do you let your Field Service people pull the plug at least quarterly? ================================================================================ Note 187.10 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 10 of 15 EISNER::MCCORMICK_B "Bob McCormick" 9 lines 29-SEP-1988 21:09 -< Test under load! >- -------------------------------------------------------------------------------- Heck, not only should you put a load on the battery (ie pull the plug) but you should measure the voltage under load -- batteries that can light silly LED's and show a charge under no load measurd with a DVM may still be useless when they are really needed. Maybe the problem is that nowadays many FS engineers don't even know Ohm's Law. When in doubt, swap it out. ================================================================================ Note 187.11 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 11 of 15 EISNER::KILGALLEN "Larry Kilgallen" 3 lines 29-SEP-1988 21:41 -< Swap what out? >- -------------------------------------------------------------------------------- > When in doubt, swap it out. The field servant? ================================================================================ Note 187.12 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 12 of 15 EISNER::KENNEDY "Terry Kennedy" 15 lines 30-SEP-1988 04:59 -------------------------------------------------------------------------------- >> When in doubt, swap it out. > The field servant? Or his firm (whoever it may be). Seriously, DEC battery backup units are designed to cut off if the battery has insufficient charge - that is why the plug-pull test is important. You can actually get the BATT light to light if your have a battery backup unit, but *no* batteries in it! Silly but true... Also, some battery backup units have an option to connect to the power control cable to the CPU, to shut down in the event of an overtemperature condition. This has been incorporated into the VAX power supplies, mostly, but is a handy feature for PDP-11's. ================================================================================ Note 187.13 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 13 of 15 EISNER::RENES "Stuart Renes, DFWLug Chair" 17 lines 30-SEP-1988 12:33 -< How well I remember! >- -------------------------------------------------------------------------------- Boy, all this talk about the old battery backup units brought back tears to my eyes (and other places!).... I used to have the BBU option on an old 785 and found out the hard way about the unreliability of those units. We found the best way to test em was to go over to the circuit breaker (not the ones on the back of the machine) and powerfail the damn processor... that will tell you at least two things... condition of the battery backup unit(s) and if powerfail restart still works on YOUR version of VMS (remember when it would work on one release and then NOT work on the next?....ad infinitum!) Hmmm.. I wonder if if works on VMS 5.0-1 (my 8530 doesn't have a BBU since it wasn't available when we leased this thing.) Stuart. :-O ================================================================================ Note 187.14 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 14 of 15 EISNER::PRIGOT "Jonathan M. Prigot" 5 lines 2-OCT-1988 20:53 -< Non-BBU works >- -------------------------------------------------------------------------------- On a variant of this note, I tried the PCS750.COM file and it seems to work (= no error messages generated). The problem with the "missing" .BIN file arises because you are expected to be in the SYS$SYSTEM: directory when you execute it (the same way the system is when it gets invoked at startup time). Don't know what it buys us though. ================================================================================ Note 187.15 11/750 - battery backup - PCS750DRV - VMS V5.0 gotcha 15 of 15 EISNER::HASSINGER "Bob Hassinger" 16 lines 2-OCT-1988 21:32 -< Loads the PCS. Where is it written...? >- -------------------------------------------------------------------------------- > ... Don't know what it buys us though. PCS750 buys any 11/750 (with rev 7) loading of the Patchable Control Store (which is supposed to be a real good thing to do). > ... you are expected to be in the SYS$SYSTEM: > directory when you execute it (the same way the system is when it > gets invoked at startup time). Yes, I sorted that one out too. BTW however, I can't remember where we are told the SYS$SYSTEM: is the default directory at SYSTARTUP time. In fact, it seems as though there are a number of bits of information like this that are hard to track down. Yes,  ================================================================================ Note 188.0 installing a second KDB50 1 reply EISNER::MERRITT "Del" 6 lines 25-JUL-1988 18:15 -------------------------------------------------------------------------------- I am about to install a second KDB50-A in my 8200. Is there anything special to watch out for, or is it pretty much just plug-n-play? In addition to the CPU and memory, I have a UNIBUS adaptor, a DEBNT (yeah, ick), and a controller for my TU81+. Thanks! ================================================================================ Note 188.1 installing a second KDB50 1 of 1 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 7 lines 26-JUL-1988 07:15 -< Problem with the DEBNT >- -------------------------------------------------------------------------------- >a DEBNT >(yeah, ick), and a controller for my TU81+. The DEBNT is no longer supported under DECnet-VAX Version 5.0 See the DECnet-VAX cover letter. It should be upgraded to a DEBNA. FCO # (T1032-R002)  ================================================================================ Note 189.0 8800,8700 third party memory? 4 replies EISNER::BLOOM "Bob Bloom, Raytheon Company" 6 lines 26-JUL-1988 10:55 -------------------------------------------------------------------------------- Does anyone have experience using third party memory in an 8800, 8700 or 8500? We have two 8800s which we are planning to add memory. There are a number of vendors which offer memory for these systems, and I would like to know whether they are worth considering. EMC claims that their memory is faster than DEC's. Can anyone confirm this claim with real performance data? ================================================================================ Note 189.1 8800,8700 third party memory? 1 of 4 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 5 lines 26-JUL-1988 22:01 -< Yo\ >- -------------------------------------------------------------------------------- I have a 8700 with a 64MB Clearpoint board in it. (plus 32MB of DECs). Its been in and running for about a year with no problems at all. I didnt really look at other folks because Ive got about 80MB of Clearpoint memory in my 8600's that hasnt burbed in the 2.5 or so years since they were installed. ================================================================================ Note 189.2 8800,8700 third party memory? 2 of 4 EISNER::SHANNON "Recovering Rumourmonger" 7 lines 26-JUL-1988 23:24 -< Doctor R Sez: Take This With a Grain 'O Salt >- -------------------------------------------------------------------------------- > EMC claims that their memory is faster than DEC's. Can anyone confirm this claim with real performance data? I'm drifting back to the dark days when I wrote for an infamous tabloid which shall remain nameless, but--if my memory serves me correctly--the claim was based on NON-REAL-WORLD conditions. In other words, it wasn't any faster on this planet. Or even the planet that C. Matco hailed from. ================================================================================ Note 189.3 8800,8700 third party memory? 3 of 4 EISNER::BYRNE_C "Charlie Byrne" 8 lines 27-JUL-1988 11:41 -< Appears to be plenty of alternatives >- -------------------------------------------------------------------------------- > < Note 189.0 by EISNER::BLOOM "Bob Bloom, Raytheon Company" > > -< 8800,8700 third party memory? >- > Does anyone have experience using third party memory in an 8800, > 8700 or 8500? Not I, but speaking of Digital Review, there was an article on July 13, 1987 that listed some 320 boards from 25 vendors for DEC compatible memory. ================================================================================ Note 189.4 8800,8700 third party memory? 4 of 4 EISNER::FULLER_S "Stuart Fuller" 5 lines 30-JUL-1988 16:59 -< Faster? How? >- -------------------------------------------------------------------------------- > EMC claims that their memory is faster than DEC's. I don't know how, since the speed of the memory subsystem is controlled by the memory controller. Do they supply a new memory controller board for the cpu?  ================================================================================ Note 190.0 VAX to FAX 11 replies EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 13 lines 28-JUL-1988 17:36 -------------------------------------------------------------------------------- I'm not really certain where to drop this so I'll try here. I'm am interested in purchasing a system (that's hardware and software) that will allow my users to send FAXes directly from our VAX. RAXCO now has a product but I know almost nothing about it. If the world were perfect, the software would tie into the MAIL BUS so that my ALL-IN-1 users could send directly from there (instead of making hardcopy and then walking down the hall, dialing ...) While we are talking about paradise, receiving FAXes would be neat too, even if all we could do was view/print them (using SIXEL ???) ================================================================================ Note 190.1 VAX to FAX 1 of 11 EISNER::LEDERMAN "Bart Z. Lederman" 16 lines 28-JUL-1988 18:17 -< Everyone so far uses IBM PC clones. >- -------------------------------------------------------------------------------- There are many, many units which interface FAX to computers, and all the ones I've ever seen (and I've looked) hook up to PCs only. I have good reason to believe that the major carriers which are offering "other communications" to FAX service use PC units. There are dedicated units which have RS-232 connectors on them so you send in text and out goes FAX. In the reverse direction you can get an image back as data. I've never seen a SIXEL converter, but it probably would not be too hard. At least one vendor did their software in C and would probably be willing to licence it for other machines, if you can get one of the modem manufactuers to put a unit out for the VAX. FYI: over 90% of all FAX machines in the world now use a Rockwell modem module. Emulex has announced a plug-compatible unit. It's NOT a complete modem, just the module that does most of the work. ================================================================================ Note 190.2 VAX to FAX 2 of 11 EISNER::KENNEDY "Terry Kennedy" 10 lines 28-JUL-1988 21:48 -< Outboard Fax modem now available >- -------------------------------------------------------------------------------- The main hangup with using PC-based FAX gear on a VAX is that the boards are all internal to the PC. A company called Everex has come out with an external (RS-232) fax modem. This at least solves the hardware part of the problem... Model: EverFax EF12/48 Fax/Data Modem Price: $495.00 (Mfg. Sug. list) Phone: (415) 683-2100, Fax: (415) 651-0728 Hope this helps... ================================================================================ Note 190.3 VAX to FAX 3 of 11 EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sa" 20 lines 29-JUL-1988 23:50 -< RAXMAIL >- -------------------------------------------------------------------------------- > RAXCO now has a product but I know almost nothing about it. It is called RAXMAIL. I spoke with a sales person today. It will let you use your word processor to develop a message and then you can send it out where it waits in a queue until the machine is free. It lets you go out Telex or FAX. There is software and hardware to buy. The cost for a VAX cluster was about $6k total. An additional advantage is that the person that receives your FAX gets a better quality copy because you don't lose what you normally would in the scanning. There is no facility (or even plans) for handling incoming FAX. > While we are talking about paradise, receiving FAXes would be neat > too, even if all we could do was view/print them (using SIXEL ???) Well actually there is some hardware and software for the Mac that will let you view/printt incoming FAX messages. So maybe someone really clever could come up with the procedures to use some of the currently available software to convert from the Mac to Regis for viewing ... ================================================================================ Note 190.4 VAX to FAX 4 of 11 EISNER::SHANNON "Recovering Rumourmonger" 5 lines 30-JUL-1988 22:45 -< One Picture Is Worth . . . >- -------------------------------------------------------------------------------- > It will let you use your word processor to develop a message . . . Does it let you transmit pictures, diagrams or other graphic images (such as signatures or purchase orders)? ================================================================================ Note 190.5 VAX to FAX 5 of 11 EISNER::CONROY "Alan Conroy" 6 lines 1-AUG-1988 14:40 -< DMG >- -------------------------------------------------------------------------------- DMG, in Canada (Eh!), also has a VAX FAX product (or soon will). I know nothing about it, except that I was talking to them about DMGNET last week and they mentioned it to me. They can be reached at 416-225-7788 or 416-225-3219. Understand that what I have just told you is the sum total of my knowledge of this product, and therefore do not take this as any kind of endorsement for the software. ================================================================================ Note 190.6 VAX to FAX 6 of 11 EISNER::WAX "Andrew L. Wax" 12 lines 14-AUG-1988 11:07 -< Another option >- -------------------------------------------------------------------------------- We have implemented the sending of FAX's/telex's using Western Union's Easylink service. We purchased Alisa Softwarewhich interfaces with Western Union. We wrote an application in All-In-1 which has a central and a personal telex/fax directory. Depends upon whether the monthly cost of the Western Union box is worth it. They also do international faxing. Alisa's # is 800 99A-LISA (whatever that translates to. More info on our setup - don't hestitate to call ALW (212) 214-7163 ================================================================================ Note 190.7 VAX to FAX 7 of 11 EISNER::MAYHEW "Bill Mayhew" 19 lines 14-AUG-1988 15:07 -< Doesn't FAX mean "pictures"? >- -------------------------------------------------------------------------------- I guess I don't understand... FAX is short for "facsimile transmission", right? As in, it sends a picture -- i.e. some sort of bit-mapped graphical image? I don't understand what the connection is between this and Telex, which simply transmits text. Yet I keep hearing people using the terms cheek-by-jowl. Conceptually, sending a Telex can be done with an ASR33 Teletype (bother me not with implementation details like character sets and bit-lengths) while sending a FAX requires something to generate a visual image, e.g. (again conceptually) a VT24x/330/340 in sixel screen-dump mode. Are you saying simply that you can send a textual message with no graphics appendages in such a way that a genuine fax machine can print it at the receiver's end, or what? ================================================================================ Note 190.8 VAX to FAX 8 of 11 EISNER::LEDERMAN "Bart Z. Lederman" 15 lines 14-AUG-1988 16:27 -< Telex to FAX becomming a standard service >- -------------------------------------------------------------------------------- Telex machines are bulky and somewhat expensive, and the dedicated line can be costly. So, more people are using dial-up services. In addition, a LOT of offices and businesses have FAX machines, but don't have Telex machines: so people who want to send messages to them would like to have a way to send plain text to a FAX machine. A number of the Telex carriers offer the service of delivering a text message to a FAX machine by converting it. (Most if not all of the PC based FAX interfaces will do this as well.) To the best of my knowledge, no-one offers an automatic service to convert text within a FAX message to Telex or ASCII text (this would be something like an OCR conversion). But there is software which will convert a picture to a sort of image made up by printing characters on a printer (remember those Snoopy and pin-up pictures people used to print?) ================================================================================ Note 190.9 VAX to FAX 9 of 11 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 18 lines 15-AUG-1988 09:58 -< Re-read your sales lit. >- -------------------------------------------------------------------------------- DEC has defined a standard for the combination of text and graphics called the the Digital Document Interchange Format (DDIF). You and Bart are correct in that text must be converted to a "picture" in a bit map memory the same way a laser printer would do it (gee, display PostScript might work good here !). But this "picture" will be much better because we have eliminated the scanning process that could pick up dirt and smudges on the lens. Pictures of text would be received and stored in bit map form also but character recognition software will have a much easier time of deciphering the text without the "noise" you get from the scanning process. BTW, PostScript has the capability of printing almost any bit mapped image because you tell it width and the height of the pictures and the number of bits/sample (see image and imagemask operators in the PostScript Language Reference Manual). ================================================================================ Note 190.10 VAX to FAX 10 of 11 EISNER::MAYHEW "Bill Mayhew" 4 lines 15-AUG-1988 15:20 -< DEC only sends sales lit to Fortune companies :-) >- -------------------------------------------------------------------------------- Are you saying that a DDIF "file" is fax-compatible, or are you only making a conceptual analogy, Jack? And, what products are out there, today, which support DDIF? I had thought they were mostly still on the drawing boards. ================================================================================ Note 190.11 VAX to FAX 11 of 11 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 5 lines 17-AUG-1988 08:20 -------------------------------------------------------------------------------- DDIF has had a longing gestation period but I think it will be "born" very soon, and Apple may be the first (ahead of even DEC) in supporting this "standard" (see latest news releases on the E-Store). The reference to DDIF and FAX was "conceptual"  ================================================================================ Note 191.0 RA60 Data Check and Reformat 5 replies EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 16 lines 3-AUG-1988 10:46 -------------------------------------------------------------------------------- I have an RA60 that seems to have gotten lost on the servo surfaces and will no longer allow the drive to load it. Is it possible to reformat the disk with one of the many handy dandy diagnostic programs? We are "self-maintenance" as far as DEC is concerned (We went to CDC) and therefore have full diagnostics for all our machines, including HSC 50s. Also, I have a pack that may have data reliability problems. Can I run some sort of utility to read/write test the pack? When I attempted to run BAD on a new RA82 it said I couldn't, and the RA?? disks are not mentioned in the writeup for BAD anywhere. Seton P.S. Note 63 (this conference) mentions that RA60 packs are guaranteed for life, who at DEC to you call to take advantage of this? SRD ================================================================================ Note 191.1 RA60 Data Check and Reformat 1 of 5 EISNER::RENES "Stuart Renes, DFWLug Chair" 15 lines 3-AUG-1988 13:57 -< Try HSC Utilities! >- -------------------------------------------------------------------------------- All of the RA class of disks are members of the "DSA" family which amoung other things specifies that they display themselves as perfect devices to VMS. The controllers and/or HSCs' take care of the bad block replacement, etc. to insure the drives look perfect to VMS. That's why you cannot run BAD, etc. on DSA disks. The HSC utility cartridges/floppies contain diagnostics that allow formatting and other error testing to be performed on DSA disks. I believe the HSC manual described some of these. That's the limit of my knowledge on the subject... you should here more detail in other replies from other users... ================================================================================ Note 191.2 RA60 Data Check and Reformat 2 of 5 EISNER::KENNEDY "Terry Kennedy" 28 lines 3-AUG-1988 20:58 -< Possible help >- -------------------------------------------------------------------------------- > I have an RA60 that seems to have gotten lost on the servo surfaces and > will no longer allow the drive to load it. Is it possible to reformat > the disk with one of the many handy dandy diagnostic programs? Well, if the problem is indeed the servo surface you are in trouble - that can only be written at the factory. Also, if the drive won't load with that pack, I don't think you could run the formatter. Note that the following info is relevant to RA8X drives - I don't have any RA60's (lucky me). When you pop the top of the drive open with the generic DEC Allen wrench, you should see a DB-25 connector next to the drive motor start capacitor. *AFTER* powering on the drive and waiting a minute or so, connect a terminal set to 300 baud to the connector. Fire off ^C's until you get a prompt. Now type RUN DIAG. Wait for all tests to complete. Spin up the drive. Wait for more tests. Do another RUN DIAG. Do SET DIAG LOOP=HALT RUN DIAG. If no errors, you are ok. If errors, look at the error text and see what you got. Post here and I'll ponder it... > P.S. Note 63 (this conference) mentions that RA60 packs are guaranteed > for life, who at DEC to you call to take advantage of this? SRD I *think* the actual guarantee is 'free service forever while the system is covered by a DEC maintenance agreement'. At least that's the way memory works. By the way, the most common problem with RA60's (according to my handy- dandy failure statistic guide) is a head failure, followed by the servo board... ================================================================================ Note 191.3 RA60 Data Check and Reformat 3 of 5 EISNER::LYMAN "Rich Lyman" 9 lines 5-AUG-1988 20:54 -< Once burned, forever burned... >- -------------------------------------------------------------------------------- RA series disks can be reformatted with the diagnositics used by DEC field service. These diags can also be purchased from DEC. (However, you might have to be an OEM or whatever they are calling OEM's these days.) Unfortunately, I have not had a lot of luck with disks that have been reformatted. It kind-of seems that one a disk goes bad it never really recovers. Good luck... ================================================================================ Note 191.4 RA60 Data Check and Reformat 4 of 5 EISNER::FULLER_S "Stuart Fuller" 17 lines 7-AUG-1988 02:04 -------------------------------------------------------------------------------- If the drive will not load heads on the pack, then you can't run any controller or host based formatter. How ya gonna get the heads on the pack? The 'RUN DIAG' only works on RA81/82 drives. The RA60 is a bit retarded in that regard. If it's just one pack, it's probably not worth the bother. If you can get it swapped under some warranty (how old is it?), do so. Else, forget it. If it's multiple packs, and only on one drive, then the most common failure is probably a head (as someone else has already noted). Which head? Let CDC figure it out. Stu ================================================================================ Note 191.5 RA60 Data Check and Reformat 5 of 5 EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 3 lines 8-AUG-1988 13:03 -< Down the tubes... >- -------------------------------------------------------------------------------- We went with the "one fried disk, down the tubes..." approach. Seton  ================================================================================ Note 192.0 VT103 INFO 2 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 6 lines 3-AUG-1988 17:22 -------------------------------------------------------------------------------- Does anyone know...... 1. What model backplane is in the VT100? 2. Can it be coverted to 22-bit operation? ================================================================================ Note 192.1 VT103 INFO 1 of 2 EISNER::KOZAM 17 lines 3-AUG-1988 22:29 -< VT103s >- -------------------------------------------------------------------------------- > 1. What model backplane is in the VT100? None of the VT1XX have a backplane except for the VT103. The VT103 contains a 18-bit Q-bus backplane - I think in the style of the H9270 (4 x 2, serpentine arrangement). > 2. Can it be coverted to 22-bit operation? I've yet to find a 18-bit backplane that couldn't! (I've done it to H9270s, H9273s, and MINCs). I don't think it would be a problem. Whether this is a good idea or not is another issue. I think that the VT103 would get pretty hot, but I don't have any statistics about board failures. Marc Kozam ================================================================================ Note 192.2 VT103 INFO 2 of 2 EISNER::MCALLISTER "Brian McAllister" 27 lines 8-AUG-1988 19:22 -< No pins on the back, but it has a fan >- -------------------------------------------------------------------------------- >> 1. What model backplane is in the VT100? >VT103 contains a 18-bit Q-bus backplane - I think in the style of the >H9270 (4 x 2, serpentine arrangement). Like the H9270, it is a 4x4 serpentine. The only difference with the H9270 is that the SRUN signal is available in the last (half) slot, for use with the M8208 VT103 Maintenance Module. [ Provides a RUN/HALT switch, a LTC ON/OFF switch, and DCOK & RUN LEDs. ] >> 2. Can it be coverted to 22-bit operation? Theoretically, yes. However, unlike the H9270 & H9273, there are no exposed pins on the back of the backplane, so a 22-bit modification would require some significant hacking of the backplane itself. > Whether this is a good idea or not is another issue. I think >that the VT103 would get pretty hot, but I don't have any statistics >about board failures. The VT103 has a cooling fan specifically for the backplane/card cage. It uses the same vents a normal VT100 uses for convective cooling. We have made extensive use of VT103s over the years, and have never had any reliability problems. We used the model that includes a TU-58 tape drive. We would still be using them DEC hadn't stopped selling them.  ================================================================================ Note 193.0 So: how fast IS your VAX, anyway? 14 replies EISNER::LEDERMAN "Bart Z. Lederman" 16 lines 10-AUG-1988 10:11 -------------------------------------------------------------------------------- Some time ago, I wrote some programs which, when run, would tell you how fast a particular CPU executed instructions. I did this for the PDP-11, and also for the VAX. I put the programs on the SIG tapes, and thought people would be interested in knowing how fast various models of CPU are, but didn't get any response. I recently made some extensive additions to my VAX test programs, and am running tests on a Micro-VAX II and an 11/785. I am going to post the preliminary results for the Micro-VAX II here. If people are interested in pursuing this (I suggest responding by VAX mail so we don't clutter up the conference), I will post the results for the 785, an explanation of how the tests are done, and of course will arrange to share the programs so tests can be done on other processors. I could even post the PDP-11 results. ================================================================================ Note 193.1 So: how fast IS your VAX, anyway? 1 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 84 lines 10-AUG-1988 10:11 -< Micro-VAX II Instruction Timings (Preliminary) >- -------------------------------------------------------------------------------- Preliminary instruction timings for a Micro-VAX II, sorted by Mnemonic  Number Instruction Addressing of Execution Time, Mode Tests Nano-seconds ADAWI 5 5 427 ADDB2 5 5 431 ADDB2 6 5 1,837 ADDF2 6 5 4,265 ADDL2 6 5 1,842 ADDP4 6 3 257,472 (packed instructions are emulated) ADDW2 5 5 427 ADWC 5 5 429 ADWC 6 5 1,838 BCC 2 1,056 BCS 2 223 BICW3 1 1,688 BITW 5 5 428 BRB 2 1,053 BRW C 5 1,068 CLRB 5 2 616 CLRD 5 2 845 CLRF 5 2 616 CLRG 5 2 820 CLRL 5 2 620 CLRQ 5 2 820 CLRW 5 2 620 CMPF 6 5 3,321 CMPL 6 5 1,523 CMPP3 3 92,557 CMPP4 6 3 94,905 CMPW 5 5 427 CVTBW 5 5 832 CVTPL 5 3 114,070 CVTPL 6 3 111,183 DECB 5 2 620 DECL 5 2 621 DECW 5 2 618 DIVW2 5 5 8,949 EXTZV 1 4,312 INCB 5 2 622 INCB 9 1 1,892 INCL 5 2 640 INCL 9 1 1,748 INCW 5 2 631 INCW 9 1 1,688 JSB 6 2 3,127 (Includes the time for an JSB 9 1 3,564 RSB instruction.) MCOMW 5 5 428 MNEGL 5 5 434 MNEGL 6 5 1,268 MNEGW 5 5 427 MNEGW 6 5 1,272 MOVB 5 5 429 MOVB 6 5 1,216 MOVB 78 5 1,272 MOVL 5 5 433 MOVL 6 5 1,220 MOVL 78 5 1,271 MOVQ 5 5 836 MOVQ 6 5 2,063 MOVQ 78 5 2,114 MOVW 5 5 427 MOVW 6 5 1,223 MOVW 78 5 1,269 MOVZBL 5 5 429 MULW2 5 5 5,504 TSTB 5 2 421 TSTB 9 1 1,180 TSTD 5 2 1,032 TSTD 9 1 2,000 TSTF 5 2 823 TSTF 9 1 1,576 TSTL 5 2 418 TSTL 9 1 1,180 TSTW 5 2 416 TSTW 9 1 1,180 XORB2 5 5 427 XORB2 6 5 1,846 XORL2 5 5 426 XORL2 6 5 1,840 XORW2 5 5 430 XORW2 6 5 1,844 ================================================================================ Note 193.2 So: how fast IS your VAX, anyway? 2 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 141 lines 11-AUG-1988 18:38 -< PDP-11/70, Early 11/84 (11/84P), PRO-350 >- -------------------------------------------------------------------------------- PDP-11/70, Early 11/84 (11/84P), PRO-350  "Basic Instructions" includes the TeST, ROtate Right, SWAp Bytes, MOVe, MOVe Byte, CoMPare, ADD, and BIt Test instructions, all of which execute at the same speed when operating on registers. "addr" is a symbolic address reference: when there are two references to "addr" in the same instruction they are to two different locations. "Short" and "Long" is the length of the instruction sequence: a short sequence will all fit into cache and give the fastest possible CPU times. The long sequence does not fit into cache, and shows the effect of main memory speed. All times are converted to micro-seconds. 11/70 11/70 11/84 11/84 PRO- Instruction Short Long Short Long 350 Basic Instructions 0.31 0.91 0.3 0.7 2.38 TST (R0) 0.76 1.36 0.7 1.3 4.20 CMP (R0), (R3) 1.06 1.67 1.6 2.1 6.11 MOV -(SP), (SP)+ 1.70 2.55 2.1 2.7 6.50 MOVB -(SP), (SP)+ 2.00 2.86 2.1 2.7 6.51 MUL R1, R0 3.19 3.81 5.7 6.5 26.64 DIV R1, R0 1.98 2.59 9.3 10.2 8.96 BEQ addr 0.31 0.90 2.38 BNE addr 0.61 1.21 2.38 CCC, NOP 0.61 1.21 0.8 1.25 3.23 MOV #1, R0 0.77 1.95 0.6 1.5 4.07 ASHC #1, R0 1.38 2.56 2.1 3.1 11.92 MOV @#addr, R1 1.67 2.42 1.1 2.0 5.49 TST @#addr 1.67 2.42 1.1 2.0 5.49 SWAB @#addr 2.03 3.66 1.9 2.9 6.78 MOV addr, R1 1.43 2.27 1.4 2.3 6.07 TST 2(R0) 1.51 2.26 1.3 2.2 6.11 TST addr 1.48 2.27 1.3 2.3 6.10 MOV 2(R0), (R3) 1.87 3.52 2.3 3.4 8.12 JSR PC, (R2) 3.06 3.68 9.03 JSR PC, @#addr 3.80 4.63 10.25 MOV #1, @#addr 2.04 4.08 1.9 3.2 7.34 MOV @#addr, @#addr 2.76 4.54 2.4 3.7 8.96 MOV 4(R0), 2(R3) 2.62 4.44 3.0 4.3 9.68 ADD #1, @#addr 3.02 4.34 2.1 3.5 8.39 ADD @#addr, @#addr 2.81 4.84 2.7 4.0 10.10 ADD 4(R0), 2(R3) 2.68 4.75 3.2 4.5 10.73 SETF 1.22 1.83 1.6 2.0 6.14 MULF AC2, AC0 2.57 2.68 15.4 16.3 61.35 ADDF AC0, AC1 2.72 2.81 14.3 15.3 72.78 NEGF AC0 1.22 1.83 4.8 5.4 15.83 ABSF AC0 1.22 1.83 5.2 5.9 15.99 LDF R0, AC0 1.22 1.82 11.47 LDF (R3), AC0 1.97 2.60 14.67 LDCIF R0, AC0 2.27 2.37 29.14 LDCIF (R3), AC0 2.27 2.59 20.47 STCFI AC0, R0 2.73 3.36 36.00 STCFI AC0, (R3) 3.36 4.17 37.59 SETD 1.22 1.82 1.6 2.1 6.13 MULD AC2, AC0 3.94 4.01 44.4 46.4 215.44 ADDD AC0, AC1 2.72 2.83 19.2 20.4 98.83 NEGD AC0 1.22 1.82 5.9 6.5 19.32 ABSD AC0 1.22 1.82 6.2 7.0 19.29 LDF #1, AC0 1.81 2.88 3.4 4.4 14.29 LDF @#addr, AC2 2.72 3.49 16.06 STF AC0, @#addr 4.22 5.26 4.0 5.1 11.80 CMPF @#addr, AC0 2.89 3.49 5.7 6.9 30.40 LDCIF #1, AC0 2.61 2.89 42.01 LDCIF @#addr, AC0 2.44 3.18 25.20 STCFI AC0, @#addr 3.41 4.89 9.9 11.1 38.24 I had only a limited amount of time on the 11/84, and ran the shortest tests: it would be desirable to run them again to obtain the same precision as for the 11/70 and PRO-350 tests. Also, the longest sequence might not have hit the maximum memory access times on the 11/84. The times obtained for the 11/70 are the same using core and MOS memory, it's not certain if memory was fully interleaved: interleaving makes main memory access faster, and so the times on the 11/70 for the long loop might improve on a machine with interleaved memory. The values for the short loop, where the entire loop is in cache, do match very well the published values in the 11/70 processor handbook This was an early 11/84P which had the 15 MHz processor clock, and not the faster 18 MHz clock since announced. It also did not have the Floating Point Co-Processor, which is supposed to execute floating point instructions 5 to 8 times faster than the basic J-11 processor. The times for the PRO-350 are close to but slightly slower than the times published for an 11/23, which uses the same processor. The PRO-350 CPU has to do extra work such as updating the video screen and keeping the system clock working, which may account for the difference. Effect of Cache PDP-11/84P Length of Instruction Sequence Instruction 100 250 500 1,000 2,500 5,000 10,000 ROR R0 .3 .3 .3 .3 .3 .4 .8 TST (R0) .7 .8 .8 .8 .8 1.0 1.3 CMP (R0), (R3) 1.6 1.6 1.6 1.6 1.6 1.8 2.1 MOV -(SP), (SP)+ 2.1 2.2 2.2 2.2 2.1 2.3 2.7 MULF AC2, AC0 15.4 15.6 15.7 15.8 15.8 16.1 16.3 MULD AC2, AC0 44.4 45.2 45.5 45.7 45.9 46.1 46.4 PDP-11/70, Core Memory Length of Instruction Sequence Instruction 100 250 500 1,000 2,500 5,000 10,000 ROR R0 .3 .3 .3 .4 .8 .9 .9 TST (R0) .7 .8 .8 .7 1.2 1.3 1.3 CMP (R0), (R3) 1.0 1.1 1.1 1.4 1.6 1.7 1.7 MOV -(SP), (SP)+ 1.7 1.7 1.7 2.0 2.5 2.5 2.5 MULF AC2, AC0 2.5 2.6 2.6 2.6 2.6 2.7 2.7 MULD AC2, AC0 3.8 3.9 4.0 4.0 3.9 3.9 3.9 Though the tests showed the effects of the larger cache on the 11/84, the transition point was shifted more twords long loops than I had anticipated. If I had the opportunity to run the tests again I would use a longer loop to insure obtaining the maximum execution times, and would run a larger total number of instructions to obtain one more significant digit of precision. It is interesting to see what happens with certain instructions, especially floating point: the instructions which take the most time to execute suffer least from not being cached as the time needed to fetch the instruction from memory is only a small part of the total instruction execution time. The fastest instructions suffer most, as here the time needed to fetch the instruction can be as much as or longer than the time the CPU needs to actually execute the operation. ================================================================================ Note 193.3 So: how fast IS your VAX, anyway? 3 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 117 lines 13-AUG-1988 11:00 -< How the programs work >- -------------------------------------------------------------------------------- This project originated when we wanted to see if we could replace some PDP-11/70s, and we wanted some method of verifying the speed of the 11/84 for which there were no published instruction times. I therefore wrote some programs which will measure the speed at which a CPU will execute instructions with reasonable accuracy. I later wrote a set of similar programs for the VAX. The method I use is quite simple: a single instruction is executed many times (generally a minimum of 1,000,000 executions), and the time this takes is measured. Because the largest single program segment on a PDP-11 is about 32,000 words, it is necessary to use a loop: a long sequence of instructions is repeated as many times as needed to get the total number of executions wanted. This is also done on the VAX to avoid outrageous working set sizes, though the instruction sequence can be more than 32,000 words. There is also the effect of cache: a short loop will fit entirely into cache and will execute at the fastest possible speed; a longer loop will not all fit into cache, and the effect of the speed of main memory can be seen. On machines with no cache, long and short loops return essentially identical times, though on the VAX one sometimes sees the effect of not crossing page boundaries when using very short loops. Finally, it is necessary to test the speed of many different instructions (one at a time) to get an adequate profile of a given CPU's performance: however, not all instructions and addressing modes can be easily tested, nor is it necessary to do so to get a good relative measurement between machines. There are two global equates at the beginning of the program: $NREP is set to the number of instructions wanted in the loop, and $NPAS is the number of times the loop is executed. These are changed and the program re-assembled to create different versions with different loop lengths (to test the effect of cache, etc.). Then each version is run several times. Although there are 3 instructions in the loop to make it repeat, if the loop is reasonably long (several hundred instructions or more) the effect of the loop repeat instructions will be negligible. Don't be surprised if the program takes more time to assemble when $NREP is large, as it takes more time to fill the instruction area. In order to make life easier (for me, the programmer) there is more than one program. Instructions which are all three bytes long are tested in one program, instructions which are 6 bytes long are in another program, etc. Also, packed decimal instructions are in their own program. There are also a few special test cases: the BRANCH instructions are set to always branch to the next instruction, which allows them to be tested even though they are relocated within the program. Two conditional branches have been selected so that one will always branch and the other will never branch: on many machines there will be a difference in execution speed for these two conditions (on a "no branch" condition the processor doesn't have to calculate and load a new program address). However, different branch instructions (branch on zero, branch on carry, branch on overflow, etc.) generally all execute at the same speed under the same condition. The other special case is the Jump to Subroutine instruction: for this to be tested there must be a Return from Subroutine instruction (which is in a fixed location), and the time of execution really is for both instructions together. As most subroutines end with a return, this should be a valid test as long as one remembers that the times obtained are for two instructions. The data which is used in the instructions is either absolute references or loaded into registers: the data for the Multiply and Divide instructions tests only one set of numbers, and the speed of these instructions is often data sensitive (especially on floating point data) so these figures are only approximations of the execution time. Auto-increment and Auto-decrement addressing modes are difficult to test as the loop is at least 1,000,000 instructions long, and any continuous test would push the register out of addressing range: the only way out I could see was to have the same instruction do a decrement and increment together, so only two addressing locations are used, and this can be seen in the MOV -(SP), (SP)+ test. The system directive $GETTIM (or equivalent) is used just before the start of the loop and just after the loop to see how long it took: it is necessary to make the test loop long enough to be significantly greater than the time interval resolution of 0.01 seconds (0.1 second for RSX). This means the test loop must run into the several second range to get good results. The use of the system directive to measure time brings up several important factors in running these tests, of which the first is the amount of time it takes to execute the directive. I measured this on an 11/70 by issuing it several times in succession and measuring the time each directive took to execute with SPM-11: this package from DEC hooks into the executive and measures time in 0.00001 second intervals (100,000 per second). The GETIM$ took about 0.0005 seconds (500 micro-seconds) to execute, which is so much less than the minimum resolution of 0.1 second that it can be ignored if the loop is long enough to register at least 0.1 seconds. On the VAX, the $GETTIM is issued many times in sequence and the time before and after the sequence is compared: I measured about 90 micro-seconds on a Micro-VAX II (VMS has only to copy a value from a system data location for this directive, so it's relatively fast). A related point is that when the test program is running, it should be the only program running, and at a high enough priority that it will "grab" the CPU more or less exclusively (the CPU will still have to service the system clock if you want the test times). I have run the tests on otherwise idle systems, and at a priority of 9 on VMS (151 or 152 on RSX-11M-Plus which is high enough to be above just about everything except MCR: I want the test to be below MCR in case I want to abort it). While this might make the instructions test a little slow, it is still a valid comparison between different machines, and when I have compared the test results obtained with the published instruction times in the PDP-11 processor handbooks, I have obtained very good matches, indicating that the program is testing accurately. On VMS I'd like to run it at a higher priority, but can't completely block the machine from doing any other work. The best indication that the priority is high enough is that running the test program several times yields consistent answers: if the priority was too low the program would be interrupted and some runs would be returning longer instruction times than others. ================================================================================ Note 193.4 So: how fast IS your VAX, anyway? 4 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 92 lines 13-AUG-1988 17:12 -< VAX 11/785 >- -------------------------------------------------------------------------------- The first time is for a short sequence, in which the entire test should fit in cache memory: the long sequence should be long enough (100,000 instructions) so that it does not fit into cache. This 11/785 has a floating point accelerator, and EMC 4MB memory. Short Sequence Long Sequence Number Number Addressing of Nano- of Nano- Mode Runs Seconds Runs Seconds ADAWI 5 9 809 1 860 ADDB2 5 5 274 1 792 ADDB2 6 5 1,396 1 2,270 ADDF2 6 5 1,400 1 2,330 ADDL2 6 5 1,394 1 2,261 ADDP4 6 1 17,620 3 18,302 ADDW2 5 5 278 1 792 ADWC 5 5 270 1 793 ADWC 6 5 1,398 1 2,275 BCC 2 540 2 934 BCS 2 276 2 528 BICW3 2 1,384 1 2,935 BITW 5 5 278 1 793 BRB 2 540 2 939 BRW C 5 534 1 1,110 CLRB 5 4 417 1 529 CLRD 5 4 805 2 833 CLRF 5 2 408 2 530 CLRG 5 4 806 2 837 CLRL 5 2 408 2 530 CLRQ 5 2 800 2 834 CLRW 5 2 408 2 535 CMPF 6 4 1,070 1 1,153 CMPL 6 5 804 1 907 CMPP3 4 4,909 2 4,923 CMPP4 6 1 5,090 2 5,337 CMPW 5 5 272 1 795 CVTBW 5 5 538 1 808 CVTPL 5 1 5,010 1 5,253 CVTPL 6 1 4,860 1 5,088 DECB 5 2 278 2 528 DECL 5 2 276 2 528 DECW 5 2 278 2 528 DIVW2 5 2 4,075 2 4,249 EXTZV 2 2,110 3 2,212 INCB 5 2 278 2 529 INCB 9 2 1,388 1 3,102 INCL 5 2 276 2 533 INCL 9 1 3,104 2 1,384 INCW 5 2 276 2 528 INCW 9 1 1,388 1 3,106 JSB 6 2 1,854 2 2,319 JSB 9 2 2,120 1 3,332 MCOMW 5 5 272 1 794 MNEGL 5 5 270 1 794 MNEGL 6 5 1,394 1 2,213 MNEGW 5 5 274 1 806 MNEGW 6 5 1,394 1 2,214 MOVB 5 5 274 1 806 MOVB 6 5 1,394 1 2,214 MOVB 78 5 1,400 1 2,217 MOVL 5 5 274 1 805 MOVL 6 5 1,392 1 2,215 MOVL 78 5 1,396 1 2,225 MOVQ 5 4 933 1 981 MOVQ 6 2 2,755 1 3,647 MOVQ 78 3 2,783 1 3,650 MOVW 5 5 274 1 793 MOVW 6 4 1,393 1 2,215 MOVW 78 5 1,392 1 2,229 MOVZBL 5 5 534 1 798 MULW2 5 3 1,063 1 1,132 NOP 6 272 1 277 TSTB 5 2 280 2 528 TSTB 9 2 544 1 1,798 TSTD 5 2 276 2 528 TSTD 9 2 808 1 1,808 TSTF 5 2 280 2 533 TSTF 9 2 548 1 1,799 TSTL 5 2 276 2 528 TSTL 9 2 548 1 1,806 TSTW 5 2 276 2 529 TSTW 9 2 544 1 1,806 XORB2 5 5 274 1 800 XORB2 6 5 1,398 1 2,265 XORL2 5 5 274 1 805 XORL2 6 5 1,396 1 2,261 XORW2 5 5 270 1 797 XORW2 6 4 1,400 1 2,265 ================================================================================ Note 193.5 So: how fast IS your VAX, anyway? 5 of 14 EISNER::FULLER_S "Stuart Fuller" 6 lines 14-AUG-1988 12:30 -------------------------------------------------------------------------------- > < Note 193.4 by EISNER::LEDERMAN "Bart Z. Lederman" > > This 11/785 has a floating point accelerator, and EMC 4MB memory. --- ^ Not that the brand of memory has any effect on system performance... ================================================================================ Note 193.6 So: how fast IS your VAX, anyway? 6 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 5 lines 14-AUG-1988 16:28 -< Brand doesn't count, but speed does. >- -------------------------------------------------------------------------------- The brand of memory may (should) not affect performance, but the speed of the memory will effect performance. I don't know if anyone has figures for access time for EMC versus any other memory, but if they do it would be interesting to see if there really is any effect. ================================================================================ Note 193.7 So: how fast IS your VAX, anyway? 7 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 72 lines 14-AUG-1988 16:33 -< VAX-11/785 Compatibility Mode >- -------------------------------------------------------------------------------- The following is the result of testing a VAX-11/785 in compatibility mode (using the PDP-11 programs). Most of the instructions were run for 100,000,000 instruction loops, which may actually have been too long as the floating-point instructions took more than an hour each to test. (Instructions marked * were tested in 10,000,000 instruction loops.) Note that though the 785 has compatibility mode in hardware, the floating point instructions are emulated: and they take so long that I suspect that the hardware floating point is not used.  Mnemonic Time, Micro-seconds (replace the . with a , for nano-seconds) ABSD 35.771 ABSF 34.978 ADD .477 ADD23P 2.670 * ADD33P 2.700 * ADD66 2.700 * ADDD 31.723 ADDF 30.121 ASHC 2.047 BEQ .476 BIT .476 BNE .774 CCC .686 CMP .477 CMP 1 .712 CMPF 3 32.144 DIV .955 JSR 2 2.468 LDCIF 30.816 LDCIF1 30.026 LDCIF2 32.183 LDCIF3 32.088 LDF 28.801 LDF 1 29.603 LDF 2 33.781 LDF 3 31.711 MOV .476 MOV 2 .654 MOV 3 .956 MOV 42 1.811 MOV 6 .964 MOV 62 2.257 MOV23P 2.600 * MOV33P 2.620 * MOV66 2.600 * MOVB .501 MOVB42 1.814 MUL .819 MULD 32.950 MULF 30.133 NEGD 34.538 NEGF 32.538 NOP .702 ROR .503 SETD 22.347 SETF 21.944 STCFI 36.803 STCFI1 35.899 STCFI3 37.936 STF 3 31.699 SUBRTN 2.744 SWAB .684 SWAB 3 2.378 TST .477 TST 1 1.177 TST 3 1.258 TST 6 1.267 TST I 1.259 ================================================================================ Note 193.8 So: how fast IS your VAX, anyway? 8 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 90 lines 25-AUG-1988 15:15 -< VAX 11/750 >- -------------------------------------------------------------------------------- These are some instruction timings that Bob Hassinger send me from his 11/750, 8MB memory, VMS V5.0-1 Packed instructions, NOP and address mode 9 (absolute) were run in long loops (generally no effect of cache memory seen), others were run in short loops (cache seen).  Number of Average Mnemonic runs Time (Nano-Seconds) ADAWI 5 3 970 ADDB2 5 3 993 ADDB2 6 3 1,913 ADDF2 6 3 9,357 ADDL2 6 3 1,910 ADDP4 6 3 37,330 ADDW2 5 3 957 ADWC 5 3 953 ADWC 6 3 2,230 BCC 3 1,709 BCS 3 921 BICW3 3 4,087 BITW 5 3 957 BRB 3 1,708 BRW C 3 2,017 CLRB 5 3 1,081 CLRD 5 3 1,396 CLRF 5 3 1,079 CLRG 5 3 1,399 CLRL 5 3 1,096 CLRQ 5 3 1,396 CLRW 5 3 1,081 CMPF 6 3 3,180 CMPL 6 3 1,910 CMPP3 3 17,913 CMPP4 6 3 21,658 CMPW 5 3 953 CVTBW 5 3 1,907 CVTPL 5 3 12,020 CVTPL 6 3 12,030 DECB 5 3 1,080 DECL 5 3 1,080 DECW 5 3 1,081 DIVW2 5 3 5,437 EXTZV 3 5,889 INCB 5 3 1,091 INCB 9 3 4,075 INCL 5 3 1,077 INCL 9 3 3,619 INCW 5 3 1,080 INCW 9 3 4,053 JSB 6 3 5,053 JSB 9 3 6,748 MCOMW 5 3 953 MNEGL 5 3 950 MNEGL 6 3 1,787 MNEGW 5 3 953 MNEGW 6 3 1,787 MOVB 5 3 990 MOVB 6 3 1,793 MOVB 78 3 1,793 MOVL 5 3 953 MOVL 6 3 1,787 MOVL 78 3 1,797 MOVQ 5 3 1,320 MOVQ 6 3 2,747 MOVQ 78 3 2,760 MOVW 5 3 953 MOVW 6 3 1,787 MOVW 78 3 1,817 MOVZBL 5 3 1,913 MULW2 5 3 5,270 NOP 3 1,769 TSTB 5 3 1,080 TSTB 9 3 3,187 TSTD 5 3 1,632 TSTD 9 3 3,381 TSTF 5 3 1,317 TSTF 9 3 3,007 TSTL 5 3 1,089 TSTL 9 3 3,179 TSTW 5 3 1,091 TSTW 9 3 3,192 XORB2 5 3 953 XORB2 6 3 1,910 XORL2 5 3 957 XORL2 6 3 1,910 XORW2 5 3 957 XORW2 6 3 1,907 ================================================================================ Note 193.9 So: how fast IS your VAX, anyway? 9 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 61 lines 25-AUG-1988 15:15 -< VAX 11/750 Compatibility Mode >- -------------------------------------------------------------------------------- These are some preliminary results from Bob Hassinger. There wasn't time to run all of the tests yet, but I figure some data is better than none. The machine was a VAX 11/750, 8 MB memory, VMS 5.0-1, VAX-11 RSX V2.4  MNEMONIC Average time, Micro-Seconds ADD 1.8 ASHC 4.8 BEQ 1.6 BIT 1.8 BNE 2.1 CCC 1.4 CMP 1.8 CMP 1 3.0 DIV 6.4 JSR 2 7.0 (includes return) MOV 1.8 MOV 2 2.4 MOV 3 2.7 MOV 42 4.1 MOV 6 2.7 MOV 62 4.6 MOVB 1.8 MOVB42 4.1 MUL 7.2 NOP 1.4 ROR 2.3 SUBRTN 7.1 SWAB 2.3 SWAB 3 4.3 TST 1.6 TST 1 2.3 TST 3 2.3 TST 6 2.3 TST I 2.3 ABSD 86.4 ABSF 89.7 ADDD 86.2 ADDF 79.9 CMPF 3 78.8 LDCIF 79.4 LDCIF1 75.6 LDCIF2 83.2 LDCIF3 81.9 LDF 72.5 LDF 1 72.3 LDF 2 83.0 LDF 3 76.9 MULD 113.8 MULF 84.1 NEGD 84.2 NEGF 82.8 SETD 53.2 SETF 51.3 STCFI 91.8 STCFI1 89.6 STF 3 74.8 ================================================================================ Note 193.10 So: how fast IS your VAX, anyway? 10 of 14 EISNER::MERRIMAN "George Merriman - CCA/NY" 87 lines 22-SEP-1988 04:05 -< VAX 8550 timings >- -------------------------------------------------------------------------------- Here are the results of a run of Bart's software on a VAX 8550, 64MB, VMS V4.7:  Mnemonic Time (Nano-Seconds) ADAWI 5 130 ADDB2 5 140 ADDB2 6 360 ADDF2 6 470 ADDL2 6 180 ADDP4 6 3355 ADDW2 5 130 ADWC 5 180 ADWC 6 220 BCC 248 BCS 68 BICW3 356 BITW 5 130 BRB 248 BRW C 240 CLRB 5 92 CLRD 5 136 CLRF 5 92 CLRG 5 136 CLRL 5 88 CLRQ 5 136 CLRW 5 92 CMPF 6 350 CMPL 6 180 CMPP3 2150 CMPP4 6 2150 CMPW 5 130 CVTBW 5 180 CVTPL 5 1140 CVTPL 6 1140 DECB 5 88 DECL 5 88 DECW 5 92 DIVW2 5 400 EXTZV 656 INCB 5 92 INCB 9 292 INCL 5 88 INCL 9 256 INCW 5 88 INCW 9 296 JSB 6 696 JSB 9 776 MCOMW 5 140 MNEGL 5 130 MNEGL 6 130 MNEGW 5 130 MNEGW 6 270 MOVB 5 130 MOVB 6 270 MOVB 78 310 MOVL 5 90 MOVL 6 140 MOVL 78 180 MOVQ 5 220 MOVQ 6 530 MOVQ 78 570 MOVW 5 140 MOVW 6 270 MOVW 78 350 MOVZBL 5 90 MULW2 5 620 NOP 44 TSTB 5 92 TSTB 9 256 TSTD 5 224 TSTD 9 292 TSTF 5 180 TSTF 9 254 TSTL 5 92 TSTL 9 224 TSTW 5 92 TSTW 9 260 XORB2 5 130 XORB2 6 350 XORL2 5 140 XORL2 6 180 XORW2 5 140 XORW2 6 350 ================================================================================ Note 193.11 So: how fast IS your VAX, anyway? 11 of 14 EISNER::MERRIMAN "George Merriman - CCA/NY" 91 lines 5-OCT-1988 14:37 -< VAX 3600 timings >- -------------------------------------------------------------------------------- Here are the results of a run of Bart's software on a VAX 3600, 32MB, VMS V4.7:  Mnemonic Time (Nano-Seconds) ADAWI 5 180 ADDB2 5 180 ADDB2 6 650 ADDF2 6 1070 ADDL2 6 530 ADDP4 6 97910 ADDW2 5 180 ADWC 5 180 ADWC 6 530 BCC 444 BCS 104 BICW3 1244 BITW 5 180 BRB 440 BRW C 400 CLRB 5 268 CLRD 5 452 CLRF 5 272 CLRG 5 452 CLRL 5 272 CLRQ 5 452 CLRW 5 272 CMPF 6 880 CMPL 6 450 CMPP3 36840 CMPP4 6 37565 CMPW 5 180 CVTBW 5 350 CVTPL 5 39640 CVTPL 6 38670 DECB 5 272 DECL 5 272 DECW 5 272 DIVW2 5 2290 EXTZV 1572 INCB 5 272 INCB 9 1244 INCL 5 272 INCL 9 940 INCW 5 272 INCW 9 1240 JSB 6 1256 JSB 9 1668 MCOMW 5 180 MNEGL 5 180 MNEGL 6 450 MNEGW 5 180 MNEGW 6 650 MOVB 5 180 MOVB 6 640 MOVB 78 640 MOVL 5 180 MOVL 6 370 MOVL 78 530 MOVQ 5 350 MOVQ 6 740 MOVQ 78 820 MOVW 5 180 MOVW 6 640 MOVW 78 650 MOVZBL 5 180 MULW2 5 2370 NOP 92 TSTB 5 180 TSTB 9 744 TSTD 5 1136 TSTD 9 1620 TSTF 5 776 TSTF 9 1256 TSTL 5 180 TSTL 9 744 TSTW 5 180 TSTW 9 740 XORB2 5 180 XORB2 6 640 XORL2 5 180 XORL2 6 530 XORW2 5 180 XORW2 6 650 ================================================================================ Note 193.12 So: how fast IS your VAX, anyway? 12 of 14 EISNER::AGRAWAL 24 lines 3-APR-1989 12:19 -< Different numbers for VAX relative performance explained. >- -------------------------------------------------------------------------------- -< So: how fast IS your VAX, anyway? >- Sorry for such a late reply to this note. (I am a new comer to DECUServe.) How fast is your VAX? That is a good question because there seems to be so many different answers to the question. Many of the DEC's publications rate a VAX8700 to be 6 times as powerful as a VAX-11/780. Where as in some of the technical papers (written by DEC employees), a VAX8700 is taken to be 4.8 times as powerful as a VAX-11/780. Curious?! I have done a mathematical analysis of processor performance for multi-user environments and show that under optimal benchmark conditions, a 4.8 times as fast processor would deliver 6.0 times throughput (if the slower processor reached its response time threshold at 75% busy.) If any one needs a copy of the paper ("On relative performance of VAX processors"), write or call: Dr. Subhash C. Agrawal BGS Systems, Inc. 128 Technology Center Waltham, MA. 02254 (617)891-0000x360 ================================================================================ Note 193.13 So: how fast IS your VAX, anyway? 13 of 14 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 2 lines 4-APR-1989 12:14 -< I was there >- -------------------------------------------------------------------------------- Didn't you present this information at the Anaheim Symposium? If so, I was at your session. ================================================================================ Note 193.14 So: how fast IS your VAX, anyway? 14 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 100 lines 12-SEP-1989 15:34 -< MV3100 timings >- -------------------------------------------------------------------------------- This is the result of a quick test of a MV3100 with 8 MB of memory running VMS 5.1-1 All tests used 10,000 instruction loops, with 1000 passes through the loop. Since people sometimes discuss the fastest way to clear a register I included a test of three ways to clear a register, move a literal zero, multiply by zero, and exclusive or the register with itself, which have the instruction listed to the right of the time. The normal CLR instructions (which are fastest, by the way) are tested as part of the 'normal' instructions.  Instruction Mode Nanoseconds ADAWI 5 6,106 ADDB2 5 1,541 ADDB2 6 889 ADDF2 6 1,674 ADDL2 6 912 ADDP4 6 119,007 ADDW2 5 313 ADWC 5 327 ADWC 6 995 BCC 573 BCS 207 BICW3 1,176 BITW 5 308 BRB 583 BRW C 612 CLRB 5 307 CLRD 5 516 CLRF 5 310 CLRG 5 494 CLRL 5 304 CLRQ 5 486 CLRW 5 301 CMPF 6 1,154 CMPL 6 714 CMPP3 42,976 CMPP4 6 44,666 CMPW 5 298 CVTBW 5 392 CVTPL 5 45,314 CVTPL 6 41,095 DECB 5 312 DECL 5 309 DECW 5 304 DIVW2 5 2,434 EXTZV 1,664 INCB 5 306 INCB 9 1,142 INCL 5 305 INCL 9 1,163 INCW 5 307 INCW 9 1,184 JSB 6 1,470 JSB 9 1,813 MCOMW 5 297 MNEGL 5 302 MNEGL 6 717 MNEGW 5 298 MNEGW 6 739 MOVB 5 298 MOVB 6 682 MOVB 78 863 MOVL 5 298 MOVL 5 324 MOVL #0, R6 MOVL 6 681 MOVL 78 868 MOVQ 5 393 MOVQ 6 1,094 MOVQ 78 1,343 MOVW 5 298 MOVW 6 681 MOVW 78 859 MOVZBL 5 298 MULL2 5 1,152 MULL2 #0, R6 MULW2 5 2,568 NOP 135 TSTB 5 204 TSTB 9 828 TSTD 5 1,415 TSTD 9 1,797 TSTF 5 947 TSTF 9 1,359 TSTL 5 208 TSTL 9 818 TSTW 5 206 TSTW 9 833 XORB2 5 309 XORB2 6 957 XORL2 5 305 XORL2 6 889 XORL2 6 4,817 XORL2 R6, R6 XORW2 5 302 XORW2 6 890  ================================================================================ Note 194.0 TS11 lookalike problem under VMS V5 7 replies EISNER::ANGLEY "Kevin Angley" 7 lines 11-AUG-1988 18:37 -------------------------------------------------------------------------------- Since going to VMS version 5, we are having positioning problems and starnge end of tape behaviour on our TS11-lookalike Western Peripherals TD-III controller. One thing I have noticed is that VMS calls this a TSU05 under V5, where it used to call it a TS11. Does anyone else have a Ts11 lookalike that either works or doesn't under V5? ================================================================================ Note 194.1 TS11 lookalike problem under VMS V5 1 of 7 EISNER::DOHERTY "Bob Doherty" 10 lines 11-AUG-1988 19:55 -< Emulex TC13's seem OK >- -------------------------------------------------------------------------------- > Does anyone else have a Ts11 lookalike that either works or doesn't > under V5? We have two Kennedy 9400's on Emulex TC13's and they seem to work fine under V5. In particular they come up as TS11's and they run the Mandatory update fine. The machine is a '750. Bob. ================================================================================ Note 194.2 TS11 lookalike problem under VMS V5 2 of 7 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 20 lines 11-AUG-1988 21:33 -< THE NEWER TC13'S HAVE THE EXTEND FEATURE CODE >- -------------------------------------------------------------------------------- <<< EISNER::DUA0:[NOTES$LIBRARY]VMS.NOTE;1 >>> -< V5.0 is the current shipping version of VMS >- ================================================================================ Note 255.5 Backup / TS11 lookalive problem 5 of 6 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 12 lines 11-AUG-1988 21:29 -< CHEAP EMULATION WILL GET YOU EVERYTIME >- -------------------------------------------------------------------------------- > Version 4 called this device a TS11. Version 5 calls it a TSU05. > Is there noone else with a TS11 lookalike that either works or doesn't > work under V5? The TSV05 (& TS11 class drives) has something called extended features. Since RSX and VMS did not use them most third party lookalikes did not implement them. When RSTS went from V8 to V9 they started using them and all the third party controllers broke. I do not know what the extended features do exactly but without them activated you do not get full steaming on a TS%O5. Maybe DEC turned them on in V5 and thats your problem. ================================================================================ Note 194.3 TS11 lookalike problem under VMS V5 3 of 7 EISNER::ANGLEY "Kevin Angley" 16 lines 12-AUG-1988 13:58 -< I think we're on to something >- -------------------------------------------------------------------------------- >>> fine under V5. In particular they come up as TS11's and they run When you say they come up as TS11's ... you do a SHO DEV/FULL and it says they are TS11's ??? If so, then I think that may be the problem (with mine, not yours). Mine say TSU05, so something configure or device init does checks to see if it is a TS11 or a TSU05. Apparently mine are making it think it is a TSU05, when in fact it does not return XSTAT4. What is making this distinction, do you suppose? Is there a way I can manually configure these drives forced to be TS11's ... like SYSGEN>CONNECT MSA0/CSR=.../VECT=.../THIS_IS_A_TS11_NOT_A_A_TSU05 ?????? ================================================================================ Note 194.4 TS11 lookalike problem under VMS V5 4 of 7 EISNER::DOHERTY "Bob Doherty" 6 lines 13-AUG-1988 19:34 -< You may be onto something >- -------------------------------------------------------------------------------- Because of some initial idiocy when we were setting up our system our second TC13 is at a nonstandard vector, so we manually configure everything, i.e., SYSGEN autoconfigure all NEVER runs. That may be the difference. And yes, show dev/full ms shows our tape drives as TS11's. ================================================================================ Note 194.5 TS11 lookalike problem under VMS V5 5 of 7 EISNER::RENES "Stuart Renes, DFWLug Chair" 5 lines 15-AUG-1988 09:19 -< DILOG is ok >- -------------------------------------------------------------------------------- I have a TS11 lookalike using a DILOG board on my 11/725 that seems to work just fine under V5.0 Stuart. ================================================================================ Note 194.6 TS11 lookalike problem under VMS V5 6 of 7 EISNER::ANGLEY "Kevin Angley" 27 lines 16-AUG-1988 16:58 -< It's not nice to fool the muther, naturally >- -------------------------------------------------------------------------------- Oh, what a tangled web we weave when first we practice to deceive the TSDRIVER by stuffing DT$_TS11 in the device type field of the UCB... Based on the assumption that the problem was that VMS thought the TD-III was a TSU05 when in fact it is a TS11, I first looked at all possible GETDVI device information to see what the difference was. The ONLY different bit of information returned was the device type field. I wrote a little program to go into the device UCB and set the device type to TS11. This worked as far as I can tell - when you do a SHOW DEV MSA0 now you get: Magtape SULU$MSA0:, device type TS11, is online, allocated, record-oriented device, file-oriented device, error logging is enabled. instead of TSU05. However, this had no affect on the problem of it reporting end of tape at the end of the first tape of a multivolume backup. The error log (for errors reported during the backup pass) DID change. Instead of reporting XSTAT4 as a TSU05, in now correctly reports only through XSTAT3 like a good old TS11. But since it doesn't fix my problem, I don't think I've gotten anywhere. ================================================================================ Note 194.7 TS11 lookalike problem under VMS V5 7 of 7 EISNER::FEATHERLY 14 lines 17-AUG-1988 17:32 -< DEC hardware also has problem >- -------------------------------------------------------------------------------- > instead of TSU05. However, this had no affect on the problem of it > reporting end of tape at the end of the first tape of a multivolume > > But since it doesn't fix my problem, I don't think I've gotten > anywhere. For what it's worth, a DEC only hardware system (750 with TU80) using the TSDRIVER has the same problem. Not only that, it will fatal drive error if you don't init a new tape before you use it and try to have /REWIND init it. Also if you have a multi tape output save set and are doing verify, it dies before the verify pass cause it can't find the saveset cause it didn't rewind the tape!!! John  ================================================================================ Note 195.0 DHQ11 and TTDRIVER 7 replies EISNER::BOGART "N.A. Bogart, $SET CHAOS/TOTAL" 6 lines 17-AUG-1988 11:31 -------------------------------------------------------------------------------- We have installed a DHQ-11 on our MicroVAX II. We also have 3rd party graphics terminals that can be configured for 38.4 Kb. We would like to use them on this MicroVAX. However, it seems that the MicroVMS terminal driver does not support this speed, I checked this out with Colorado Springs. We cannot upgrade the system to a BI-bus system. Any suggestions on how to cope with this problem? ================================================================================ Note 195.1 DHQ11 and TTDRIVER 1 of 7 EISNER::ROECKEL 8 lines 24-AUG-1988 12:37 -< 19.2 O.K.? >- -------------------------------------------------------------------------------- Well, I don't have a suggestion to get you to 38.4kb, but a DECserver 200 will support 19.2 kb. We have a few folks here that run at that speed and, boy, is it fast !! Of course, I am assuming you have Ethernet installed. If not, this would be a very expensive option !! Bruce W. Roeckel ================================================================================ Note 195.2 DHQ11 and TTDRIVER 2 of 7 EISNER::BOGART "N.A. Bogart, $SET CHAOS/TOTAL" 6 lines 6-SEP-1988 09:09 -< 19.2 not OK >- -------------------------------------------------------------------------------- -< 19.2 O.K.? >- We are currently running at 19.2. The client isn't too thrilled with anything more than a 6 second screen update, which we haven't achieved yet. ================================================================================ Note 195.3 DHQ11 and TTDRIVER 3 of 7 EISNER::MAYHEW "Bill Mayhew" 22 lines 6-SEP-1988 10:23 -< 6+ sec screen refresh at 19.2Kb?? >- -------------------------------------------------------------------------------- < Note 195.2 by EISNER::BOGART "N.A. Bogart, $SET CHAOS/TOTAL" > > We are currently running at 19.2. The client isn't too thrilled with > anything more than a 6 second screen update, which we haven't > achieved yet. If you're saying that screen updates take 6 seconds or more at 19.2Kb, then I would look somewhere *other* than the transmission rate for the problem (and solution). A 24x80 screen is 1920 characters. At 19.2Kb, a full screen refresh would take 1 second. Even if you alternate every other character with, say, bold and normal-text, you're at worst transmitting 4 additional characters for each significant character, which means it would take 5 seconds. It seems like it would be hard work to get a screen update to take longer than 6 seconds, *if* you're really running at a *full-tilt* 19.2Kb, so my long-distance guess is that you're not... which suggests that the application, or the CPU, is not keeping the wire full of bits. If I'm out to lunch with this analysis, please tell me so... ================================================================================ Note 195.4 DHQ11 and TTDRIVER 4 of 7 EISNER::HAHN "DECUServe Vice Chair" 4 lines 6-SEP-1988 12:00 -< check the TERMINAL first >- -------------------------------------------------------------------------------- You should assertain that when the particular terminal that you are using is connected directly to the computer at 19.2Kb it does update in ~ 1 sec. That will check the terminal. ================================================================================ Note 195.5 DHQ11 and TTDRIVER 5 of 7 EISNER::KILGALLEN "Larry Kilgallen" 12 lines 6-SEP-1988 12:31 -< Not close, therefore no cigar >- -------------------------------------------------------------------------------- > Note 195.3 > EISNER::MAYHEW "Bill Mayhew" > A 24x80 screen is 1920 characters. At 19.2Kb, a full screen refresh > would take 1 second. Even if you alternate every other character Your calculations only work for character cell terminals, which she stated is not her situation. > Note 195.0 > EISNER::BOGART "N.A. Bogart, $SET CHAOS/TOTAL" > We have installed a DHQ-11 on our MicroVAX II. We also have 3rd > party graphics terminals that can be configured for 38.4 Kb. We would like ================================================================================ Note 195.6 DHQ11 and TTDRIVER 6 of 7 EISNER::MAYHEW "Bill Mayhew" 1 line 6-SEP-1988 16:06 -< Whoops - my face is red! >- -------------------------------------------------------------------------------- ================================================================================ Note 195.7 DHQ11 and TTDRIVER 7 of 7 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 11 lines 6-SEP-1988 18:16 -< DMA mode? + can terminal really do 19.2 >- -------------------------------------------------------------------------------- >>> We are currently running at 19.2. The client isn't too thrilled with >>> anything more than a 6 second screen update, which we haven't achieved yet. Are you set for DMA mode? If not, you are trying to simulate a DZ (yuk). Some terminals simply CAN'T accept characters as fast as their comm port settings might imply. The VT100 was a good example of this. It let you run at 9.6, but X-ON/X-OFFed to keep the flow at bay. VT52s did better! What will your terminal truly do??  ================================================================================ Note 196.0 8800 versus 8820 7 replies EISNER::VELA 12 lines 23-AUG-1988 11:43 -------------------------------------------------------------------------------- I've looked through most of the applicable conferences and haven't seen anyone ask *the* question, so I'll bite: What's the difference between an 8800 and an 8820? I imagine that there had to be an architecture change or the 8820 could run VMS v4.7. Naturally, our local DEC sales office couldn't answer this one! Debbie Vela Mobil Oil (214) 951-3537 ================================================================================ Note 196.1 8800 versus 8820 1 of 7 EISNER::BLOOM "Bob Bloom, Raytheon Company" 22 lines 23-AUG-1988 12:10 -< There is a difference >- -------------------------------------------------------------------------------- Yes, there is a difference, sort of. If you order an 8820 it is slightly larger and comes with a MicroVAX for a console. The Major architectural difference is that the NMI backplane has been split into two chunks. This is so that four processors can be plugged into one NMI and the memory into the other. The 8800 has only one NMI. There is a bridge between the two NMIs so that data flows properly, but the bridge creates enough of a difference in internal registers etc. that it is necessary to run VMS 5. From a performance point of view an 8800 and an 8820 should be identical. There is considerable room for confusion in the 88xx line. If you start with an 8810, you get something identical to an 8700 except for the numbers on the front of the box. This has a PRO console just like the 8700. I don't know whether it is supported, but an 8810 will presumably run VMS 4. If you buy an upgrade from an 8810 to an 8820, what you wind up with is something called an 8820N (for Nautilus.) This is also has a PRO console and since it is identical to an 8800 will presumably (someone out there correct me if I'm wrong) run VMS 4. The strange thing is that DEC prices the 8820N to 8830 upgrad$168K less than an 8800 to 8830 upgrade, although the hardware change is identical. ================================================================================ Note 196.2 8800 versus 8820 2 of 7 EISNER::FULLER_S "Stuart Fuller" 4 lines 23-AUG-1988 20:35 -------------------------------------------------------------------------------- Bottom line: If it has a PRO for a console, it'll run under V4. If it has a MicroVAX for a console, it requires V5. ================================================================================ Note 196.3 8800 versus 8820 3 of 7 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 23 lines 23-AUG-1988 23:39 -< 8820N eqs 8800, 8800 neqs 8820N >- -------------------------------------------------------------------------------- | If you buy an upgrade from an 8810 to an 8820, what you wind up with is | something called an 8820N (for Nautilus.) This is also has a PRO | console and since it is identical to an 8800 will presumably (someone | out there correct me if I'm wrong) run VMS 4. Since we've got an 8820N sitting on our loading dock, I should have some solid info tomorrow. As I recall, you can't order a non-CI 8820, so you've gotta order an 8810 and the upgrade to 8820N. What you apparently get is an 8800 with a different number on the front panel. As such, it should run V4. | The strange thing is that DEC prices the 8820N to 8830 upgrade $168K | less than an 8800 to 8830 upgrade, although the hardware change is | identical. Ah, what you get when you buy an 8820N is a non-expandable 8800 with a codicil (sp?) that when you upgrade, DEC roles out the 8800 and roles in a dual-NMI system with three CPUs. (Don't get in the way!) Seems that the 8800 owners don't get this benefit. Perhaps this is reflected in a price difference between the 'old' 8800 and the 8820N. Anyone got a DEC price list? Again, in the next few days I'll have definite, hands-on experience. ================================================================================ Note 196.4 8800 versus 8820 4 of 7 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 6 lines 23-AUG-1988 23:45 -< Big inventory solution? >- -------------------------------------------------------------------------------- One last comment... my guess is that DEC has lot's of 8800s in stock and would rather ship the 'big' 88X0 chassis to sites that immediately need an 8830 or 8840. The other sites get the 8700/8800 for now and can get an 8830 when they need it. Looks like a good inventory management solution to me. ================================================================================ Note 196.5 8800 versus 8820 5 of 7 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 4 lines 25-AUG-1988 00:28 -< Some more info >- -------------------------------------------------------------------------------- Well, we've got a big heavy box with the number 8810 on it and a cardboard box with a second CPU, 16 MB, and a '2'. After the upgrade, I'll have the '1' and a '0' from our 8500 to 8530 upgrade. Is there a used market for these numbers? :-} ================================================================================ Note 196.6 8800 versus 8820 6 of 7 EISNER::VELA "Debbie Vela, VAX Vixen" 5 lines 1-SEP-1988 17:02 -< Are you game? >- -------------------------------------------------------------------------------- To: SCOPELLITI "Whatsa behind is uva no importance!" Just out of curiosity, are you going to try to boot with v4.7? If you do, I'd like to know how it goes. No plans to do this myself (our 8820 won't even get here until November), just would like to hear. ================================================================================ Note 196.7 8800 versus 8820 7 of 7 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 7 lines 1-SEP-1988 23:52 -< Throw some corn on the ground - you'll see! >- -------------------------------------------------------------------------------- Sorry.. we're going right to 5.0-1 (as a matter fo fact, its up and running already!). We didn't want to come up with V4.7, run for a couple of weeks, then bring the system down and install V5.0. OTOH, V4.7 should certainly work - near as we can tell, the only difference between the 8800 and the 8820N is the '2' digit replacing the '0'.  ================================================================================ Note 197.0 56Kbps 64Kbps : The Difference? 13 replies EISNER::BYRNE_C "Charlie Byrne" 13 lines 24-AUG-1988 11:21 -------------------------------------------------------------------------------- Just what is the difference between 56Kbps and 64Kbps? (OK jokers: 8 Kbps) Sometimes these seem to be used interchangably, other times not. For example, the spec's for the following devices: KMS11 "supports 4 lines at 56 Kbps" KMS1P "64 or 19.2 Kbps" KCT32 "2 lines at 64Kbps" I could go on. I have wondered about this for sometime now. If this was the async world I could guess it might be a start/stop bit or something, but... ================================================================================ Note 197.1 56Kbps 64Kbps : The Difference? 1 of 13 EISNER::BYRNE_C "Charlie Byrne" 16 lines 24-AUG-1988 12:00 -< One Answer >- -------------------------------------------------------------------------------- Here is an answer I received locally: ------------------------------------------------------------- 1) 64 KBPS is the "standard" in Europe and 56 Kbps in the US 2) In the US, the channel operates at 64 Kbps but 8 Kbps are not available to the user since the carrier company uses this for internal timing/framing. Hence the user device should be set to operate at 56 kbps if possible. 3) Recently, the carriers in the US have started to provided 64 Kbps channels (they have alternate way to handle their overhead) if you wish to transmit at this speed. These are known as "Clear Channels"... but I am not sure if they are available at all locations. ----------------------------------------------------------------------- Any other comments? ================================================================================ Note 197.2 56Kbps 64Kbps : The Difference? 2 of 13 EISNER::BYRNE_C "Charlie Byrne" 3 lines 24-AUG-1988 15:33 -< Further help needed >- -------------------------------------------------------------------------------- Even if .1 is valid, this does not explain why DEC uses 56 and 64 seemingly interchangably in their documentation. ================================================================================ Note 197.3 56Kbps 64Kbps : The Difference? 3 of 13 EISNER::TINKELMAN "Bob Tinkelman - CCA" 42 lines 25-AUG-1988 06:09 -< 56 vs 64 on a T1 link >- -------------------------------------------------------------------------------- > 2) In the US, the channel operates at 64 Kbps but 8 Kbps are not > available to the user since the carrier company uses this for internal > timing/framing. Hence the user device should be set to operate at 56 kbps > if possible. This has a kernel of truth to it, but is somewhat mis-stated. What the person making that remark was referring to was the way that the phone companies traditionally multiplex voice and data channels onto higher speed (T1) links. This scheme breaks the total bandwidth into 64Kb pieces and then "wastes" 8Kb on internal phone company signalling. More details - if you're interested.  T1 channels were not designed for multiplexing data channels. They were designed by the phone company to solve the problem of sending multiple voice channels (with people talking on them) over two pairs of wires. The solution is to digitize the analog voice channels and then (TDM) multiplex them over a higher speed digital link. Most human voice is in the 0-4000 cps range. To digitize and avoid aliasing, you need about twice that many samples per second. So they filter the analog signal to clip it to something like 300-3700 cps and then sample it at 8000 times per second. Each sample is coded into 8 bits - 7 for the amplitude (how loud was the sound) and one for signalling (was the phone on or off hook). Take the eight bits from each of 24 channels, throw in a 193rd bit, called the framing bit, to help the receiver resynchronize if he gets confused, and call that a T1 frame. Do that 8000 times per second and you have 193*8000 bps - or 1.544 Mbps - or "T1". Now that's fine for voice channels. (Actually, it's not. But that's a different story.) If you want to use the same scheme for multiplexing data channels, you find that you have only 7 bits available 8000 times per second - 56 Kbps. You should be really upset that the designers thought it was necessary to see if the phone was on/off hook 8000 times per second! You certainly could make better use of that 8th bit than they. There are other schemes of chopping up high speed channels (T1 and above). Some of these carry the "signalling" separate from the digitized voice and make 64Kb available to data users. One of these is the standard in Europe. Hence the remark in your note about the US vs Europe. ================================================================================ Note 197.4 56Kbps 64Kbps : The Difference? 4 of 13 EISNER::RICE "Gary Rice" 11 lines 25-AUG-1988 10:16 -------------------------------------------------------------------------------- >< Note 197.2 by EISNER::BYRNE_C "Charlie Byrne" > > -< Further help needed >- > Even if .1 is valid, this does not explain why DEC uses > 56 and 64 seemingly interchangably in their documentation. The numbers are really meaningless if you consider the DMV-11 as your benchmark. It is "rated" at 56Kb but actual throughput is about 19Kb. Gary ================================================================================ Note 197.5 56Kbps 64Kbps : The Difference? 5 of 13 EISNER::MERRIMAN "George Merriman - CCA/NY" 8 lines 26-AUG-1988 11:09 -< Why 19kb limit? >- -------------------------------------------------------------------------------- > The numbers are really meaningless if you consider the DMV-11 > as your benchmark. It is "rated" at 56Kb but actual > throughput is about 19Kb. Please expand on this. Do you mean a DMV11 cannot keep up with a 56kb DDCMP circuit? If so, our DEC salesman has sold us another useless product! ================================================================================ Note 197.6 56Kbps 64Kbps : The Difference? 6 of 13 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 79 lines 26-AUG-1988 15:39 -< More details - whys+hows >- -------------------------------------------------------------------------------- This is to further amplify on what Bob said in .3 for anyone wanting more details. >>> This scheme breaks the total bandwidth into 64Kb pieces and then >>> "wastes" 8Kb on internal phone company signalling. >>> If you want to use the same scheme for multiplexing data >>> channels, you find that you have only 7 bits available 8000 times >>> per second - 56 Kbps. You should be really upset that the >>> designers thought it was necessary to see if the phone was on/off >>> hook 8000 times per second! You certainly could make better use >>> of that 8th bit than they. They have a valid problem. The wire line repeaters are spaced about 6,000' apart, and a whole string of them can be series powered over the simplex path (DC power on the center-tap of the pulse transformer's winding facing the line) on the same pairs carrying the bits. There need be NO other wires to those repeaters, especially there is NO clock source other than the data itself. The whole beauty of doing all this digitally is that you don't amplify a severely distorted pulse, you simply regenerate it. That scruffy odd shaped mole hill lumping in some thirty plus DB down is the clue to whang out the next 1 bit in your 3 volt bipolar square wave on the transmit side. (NOTE the BIPOLAR, we come back to it later). But you need to have a clock to accurately reconstruct the pulsetrain. The local clock is ok for a short while, but needs to see 1 bits regularly to stay in step. (Hence the rules about no long trains of zeros for folks using a leased t1 circuit.) If the telco only gave you 7 of the 8 bits from your time slot, they could do any signaling needed on your data ckt. (e.g. loopback, etc.) and could be sure there are 1s present. AT&T wants to change the world to use 'clear channel' signaling, or whatever it may be called, but her children have all left home and don't all want to spend money just because Mommy wants to sell 64kb rather than 56kb channels. Retrofitting existing equipment is a lot of expense, and some telcos may only do 64kb on newer equipment they phase in. The 'trick' to getting 64kb of wireline ckts. is to break the 'no bipolar violation' rule. Existing equipment expects each successive 1 bit to be the opposite polarity of the one preceeding it, reguardless of how many 0 (no pulse) bits may interveen, and treats any violation as an error, and may raise an alarm, switch to a spare line, or whatever. The 'new' way is to allow a special recognizable bipolar violation pattern to be substituted for some zeros, and then be replaced by the zeroes before delivery. But, alarm equipment has to know these particular patterns are legal, and if one jusisdiction uses the new encoding, another one at the other end of a ckt better be able to restore the zeroes, etc. >>> So they filter the analog signal to clip it to something like >>> 300-3700 cps and then sample it at 8000 times per second. Each >>> sample is coded into 8 bits - 7 for the amplitude (how loud was >>> the sound) and one for signalling (was the phone on or off hook). This is what they did in the early days, now there is a concept of 'signaling frames' where there are 'A' and 'B' signaling frames some of the time, and the rest of the time the 8th bit is simply the LSB of the digitized voice. Having more seperate signals in each direction allows for such things as 'coin-return', 'coin-collect', 'ring-back', etc. and none of these need anything like 8000 samples per second. Which frame (signaling 'A', 'B', no signaling) type is determined by watching the pattern in the framing bit mentioned: >>> Take the eight bits from each of 24 channels, throw in a 193rd >>> bit, called the framing bit, to help the receiver resynchronize >>> if he gets confused, and call that a T1 frame. Do that 8000 >>> times per second and you have 193*8000 bps - or 1.544 Mbps - or >>> "T1". ================================================================================ Note 197.7 56Kbps 64Kbps : The Difference? 7 of 13 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 42 lines 26-AUG-1988 16:05 -< Throughput is never equal to rated speed. >- -------------------------------------------------------------------------------- > Please expand on this. Do you mean a DMV11 cannot keep > up with a 56kb DDCMP circuit? If so, our DEC salesman > has sold us another useless product! No, it doesn't mean that. In fact the DMV11 can run a bit faster than 56 k, though ours began to choke somewhat before 64 k. What it means is that you won't see that much end-to-end throughput, as measured, for instance, by setting up a program that writes 1000 512-byte blocks at one end, a program that reads 1000 512-byte blocks at the other end, timing the run, and doing the arithmetic. Remember that there is a time delay from the issuance of a $QIO write to the movement of data to the board, and another delay before the data gets sent on the wire. The receiving DMV will take a little time to CRC the packet and send the ACK, and your DMV will take a little time to process the ACK and tell the host that it can complete the $QIO write. Then, and ONLY then, your application sends the next packet... and we haven't even talked about packet transmission time or propagation delay. If you're not doing $QIO writes to DDCMP, but are using DECnet instead, there is, of course, more overhead. Now, some of you are jumping up and down and waving your hands, anxious to say: "But DDCMP doesn't have to wait for the ACK to send the next packet!" Ah, but that only helps you if you're stacking up lots of write $QIOs. The completion of a $QIO write to a DDCMP device (*any* DDCMP device) means that the packet has arrived safe and sound in a receive buffer in the far end's memory and that the far end has correspondingly sent the ACK (which might be imbedded in another data packet), and that your end has received the ACK. (Note that I said "receive buffer". The receiving application doesn't necessarily have the data yet.) So if you are doing your write $QIOs with a wait-for-completion after each one, you have effectively reduced DDCMP's transmit window size to one. The same comments apply to writes to a DECnet logical link. If your application waits for each write to complete (which means end-to-end confirmation that the data has arrived at its destination), all of DECnet's pipelining goes out the window, as far as you're concerned. It is easy to see how all this can reduce the effective throughput of a 56 K line to 19.2K. (In our tests doing raw $QIOs to a DMV we found that things weren't nearly that bad; 35Kbit/sec throughput was more like it.) ================================================================================ Note 197.8 56Kbps 64Kbps : The Difference? 8 of 13 EISNER::RICE "Gary Rice" 17 lines 28-AUG-1988 09:29 -------------------------------------------------------------------------------- >No, it doesn't mean that. In fact the DMV11 can run a bit faster than 56 k, >though ours began to choke somewhat before 64 k. What it means is that >you won't see that much end-to-end throughput, as measured, for instance, >by setting up a program that writes 1000 512-byte blocks at one end, a program >that reads 1000 512-byte blocks at the other end, timing the run, and doing >the arithmetic. This is in fact a variation of the basis of my statement. I wrote a simple DECnet file transfer .COM file that sends a 1000 block file to a remote node and measures the time that it took to do it. I've run this file (in batch) over 24 hour periods from all sorts of systems connected in various ways. The results of DDCMP circuits (typically from a MVII to DECnet router/server (DMV-11 <==> DECSA) indicate an actual throughput of 14 - 16 K baud. Gary ================================================================================ Note 197.9 56Kbps 64Kbps : The Difference? 9 of 13 EISNER::MERRIMAN "George Merriman - CCA/NY" 11 lines 29-AUG-1988 11:23 -< FAL file transferes inefficient over any link. >- -------------------------------------------------------------------------------- > I wrote a simple DECnet file transfer .COM file that sends a 1000 block > file to a remote node and measures the time that it took to do it. This is not really a test of DDCMP -- you are including a lot of overhead imposed by higher layers. In fact, I wouldn't be surprised if you got the same order of magnitude throughput on an Ethernet link. I've realized a five-to-one improvement in throughput over simple FAL transfers by writing a specialized program optimized for the application file structures and using task-to-task communications. If you don't want to write the stuff yourself, look into DEC's Distributed File Service software. ================================================================================ Note 197.10 56Kbps 64Kbps : The Difference? 10 of 13 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 5 lines 29-AUG-1988 12:46 -< FAL IS THE PROBLEM >- -------------------------------------------------------------------------------- -< FAL file transferes inefficient over any link. >- DEC agrees with this. For DFS they trashed FAL because of it's poor proformance. ================================================================================ Note 197.11 56Kbps 64Kbps : The Difference? 11 of 13 EISNER::BYRNE_C "Charlie Byrne" 7 lines 29-AUG-1988 14:44 -< Are the proofreaders just lazy? >- -------------------------------------------------------------------------------- This has all been VERY useful and answers questions I've had for a long time. But... Why do DEC still mix specifying devices for 56K and 64K? Is it just because they think (as I did) that they are more or less the same and one tech writer puts down 56 and the next guy puts down 64? ================================================================================ Note 197.12 56Kbps 64Kbps : The Difference? 12 of 13 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 15 lines 29-AUG-1988 14:52 -< use more + bigger buffers >- -------------------------------------------------------------------------------- >> -< FAL file transferes inefficient over any link. >- In RSX11M+ land we ALWAYS tuned up FAL and NFT from their default buffering level of 2 up to 6, and made the buffers bigger. Now, under the recent releases, I seem to recall that the DEC default buffersize had been greatly enlarged and I could not crank the number of buffers up to 6, but could get to 4. I think all that was controlled by presetting variables in [137,10]decpre.cmd. These changes made a big difference on a customer's private net that had 2 sat. hops before traffic landed in the far east. ================================================================================ Note 197.13 56Kbps 64Kbps : The Difference? 13 of 13 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 23 lines 7-SEP-1988 11:27 -< Are you sure it was delivered ? >- -------------------------------------------------------------------------------- re: 197.7 >Remember that there is a time delay from the issuance of a $QIO write >to the movement of data to the board, and another delay before the data >gets sent on the wire. The receiving DMV will take a little time to >CRC the packet and send the ACK, and your DMV will take a little time >>>to process the ACK and tell the host that it can complete the $QIO >>>write. I remember distinctly hearing that the functionality of DECNET was change back in V4.0 such that a $QIO write request completed "successfully" once the packet was transferred to from your process to NETACP !! (Under V3.x your description of delivery to remote host was correct). If you wanted "guaranteed delivery" the remote process had to send your process an "ACK". This was tested and verified by a knowledgeable member of MI VAX LUG on a application that was running DECNET to guided vehicles in a plant over radio links so that "guaranteed delivery" was mandatory and had to be coded in when they went to V4.x !!!  ================================================================================ Note 198.0 What is "Pick"? 23 replies EISNER::HASSINGER "Bob Hassinger" 18 lines 26-AUG-1988 15:26 -------------------------------------------------------------------------------- I think this is really a software question but since the only place we currently have the it would fit seems to be SOAPBOX and since it does involve the development of an integrated hardware/software system, here goes, at least for the purpose of getting a better pointer to another location: A co-worker came to me today asking for any information on something called "Pick". It seems she is doing development on a PC controlled machine for doing rehabilitation and physical therapy tasks. Someone told her she "should be using Pick". She is unclear if this is an OS, some kind of network, a data base or whatever. It seems to me I have heard of something with a name like this but I can not recall anything at this point. Any clues, info or pointers to better conferences? Bob H ================================================================================ Note 198.1 What is "Pick"? 1 of 23 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 15 lines 26-AUG-1988 15:58 -< Pick is an OS >- -------------------------------------------------------------------------------- Pick is an operating system that can be on a wide variety of hardware. A very enthusiastic user gave a talk at the NY Metro lug a year or so ago. I don't remember its salient points, but someone else surely will. I categorize it and its enthusiastic users the way I do various Mumps people I know. If you are in a big rush, call Bob Tinkelman in our NY office, as he is the Metro lug speaker coordinator and will remember. ================================================================================ Note 198.2 What is "Pick"? 2 of 23 EISNER::COY "Dale E. Coy" 13 lines 26-AUG-1988 17:13 -< More >- -------------------------------------------------------------------------------- It's either (1) an operating system that's also a data base, or (2) a data base system that's also an operating system. It actually seems to be very nice (you'll properly notice the similarity to MUMPS) if your main concern is data base type stuff - you can write and maintain your OS code in the same "language" you use to write and maintain your data base. If memory serves, the PC implementation is a "database" called Revelation, which is quite powerful. You can also buy a plug-in Unibus board that lets you run Pick on your VMS system (sort-of coprocessor). ================================================================================ Note 198.3 What is "Pick"? 3 of 23 EISNER::HASSINGER "Bob Hassinger" 1 line 26-AUG-1988 18:59 -< Now that is SERVICE! Q to A in a half hour. Thanks... >- -------------------------------------------------------------------------------- ================================================================================ Note 198.4 What is "Pick"? 4 of 23 EISNER::KILGALLEN "Larry Kilgallen" 5 lines 26-AUG-1988 21:46 -< Prime uses it for their Prime Office product >- -------------------------------------------------------------------------------- The fact that some vendor has tainted it with an office automation implementation must be caution against using it! Seriously, CEO from DG and PROFS from IBM are both notably poor performers (as is Prime Office). ================================================================================ Note 198.5 What is "Pick"? 5 of 23 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 5 lines 27-AUG-1988 13:55 -< and let's not forget our favorite... >- -------------------------------------------------------------------------------- > Seriously, CEO from DG and PROFS from IBM are both notably poor > performers (as is Prime Office). As is All-in-One, for that matter. ================================================================================ Note 198.6 What is "Pick"? 6 of 23 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 6 lines 27-AUG-1988 15:02 -< All-in->1<, PLEASE. >- -------------------------------------------------------------------------------- >> As is All-in-One, for that matter. ^^^ "1" Not just nitpicking; the copyright holder of "All-in-One" (with the number spelled out) is getting increasingly upset. ================================================================================ Note 198.7 What is "Pick"? 7 of 23 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 6 lines 28-AUG-1988 01:07 -< DECUServe aims to please >- -------------------------------------------------------------------------------- >> -< Now that is SERVICE! Q to A in a half hour. Thanks... >- If you look carefully, you will find some DECUServe responses that are less than 10 minutes! Alan ================================================================================ Note 198.8 What is "Pick"? 8 of 23 EISNER::FULLER_S "Stuart Fuller" 3 lines 28-AUG-1988 11:09 -< Close, no cigar, tho >- -------------------------------------------------------------------------------- > Not just nitpicking; the copyright holder of "All-in-One" Close... It's ALL-IN-1 in UPPERcase. ================================================================================ Note 198.9 What is "Pick"? 9 of 23 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 1 line 28-AUG-1988 14:25 -< That's right -- Thank you. >- -------------------------------------------------------------------------------- >> Close... It's ALL-IN-1 in UPPERcase. ================================================================================ Note 198.10 What is "Pick"? 10 of 23 EISNER::COY "Dale E. Coy" 1 line 28-AUG-1988 16:50 -< Two in One 8-) >- -------------------------------------------------------------------------------- I believe that All-In-One is copyrighted by another corporation. ================================================================================ Note 198.11 What is "Pick"? 11 of 23 EISNER::TERRAZAS "Mike Terrazas" 24 lines 29-AUG-1988 09:57 -< Caveat Emptor >- -------------------------------------------------------------------------------- Back to the original question... :-) We have Pick running on several types of hardware, from IBM RT PCs and some other vendor (I have a mental block about this vendor) which both run pick natively, to VAXen and MicroVAXen running it on a co-processor card from Ultimate. Our experience says to look very carefully before you let yourself be led down this particular garden path. In general, less technical programmers prefer to develop under Pick, after all the native language is a variant of BASIC. The support staff of systems programmers is way out of proportion to doing the same type of stuff with a package like Oracle or Ingress. The DBMS/OS/What-is-it likes to bomb out with wonderful messages like "Group format error", which means you have to go in and patch up a bunch of links. They like to tell you the DBMS is relational, but that's not really accurate. It's really hierarchical, and you will have programmers and systems programmers worrying about those link fields to a great degree. I could fill a conference with reasons why you should avoid any of the configurations we have. On the other hand, we have 2 fanatics here who would shout me down and fill the disk extolling the virtues. Good luck. MikeT ================================================================================ Note 198.12 What is "Pick"? 12 of 23 EISNER::MERRIMAN "George Merriman - CCA/NY" 15 lines 29-AUG-1988 12:03 -< Let's wait for NON_DEC_SOFTWARE, please >- -------------------------------------------------------------------------------- > < Note 198.0 by EISNER::HASSINGER "Bob Hassinger" > > -< What is "Pick"? >- > I think this is really a software question . . . agreed! > the only place we currently have it would fit seems to be SOAPBOX Which points out the need for the proposed NON_DEC_SOFTWARE conference. As your question seems to be answered and the ALL-IN-1 question has been raised in SOAPBOX I think we should hold off on any further PICK discussion until NON_DEC_SOFTWARE comes on line. George Merriman, Hardware Help Co-moderator. ================================================================================ Note 198.13 What is "Pick"? 13 of 23 EISNER::TERRAZAS "Mike Terrazas" 9 lines 29-AUG-1988 19:07 -< Yeah, but... >- -------------------------------------------------------------------------------- >> -< Let's wait for NON_DEC_SOFTWARE, please >- >> been raised in SOAPBOX I think we should hold off on any further >> PICK discussion until NON_DEC_SOFTWARE comes on line. Okay George, but what about talking about things like the Ultimate co-processor boards? MikeT ================================================================================ Note 198.14 What is "Pick"? 14 of 23 EISNER::COY "Dale E. Coy" 7 lines 29-AUG-1988 22:40 -< Why not continue? >- -------------------------------------------------------------------------------- > I think we should hold off on any further > PICK discussion until NON_DEC_SOFTWARE comes on line. George, since this is obviously a legitimate topic for discussion on DECUServe, I can't think of a single GOOD reason to hold off on the discussion. Would you elaborate on why you think the discussion should not be continued (if the participants wish)? ================================================================================ Note 198.15 What is "Pick"? 15 of 23 EISNER::MERRIMAN "George Merriman - CCA/NY" 8 lines 30-AUG-1988 10:35 -< carry on here >- -------------------------------------------------------------------------------- > Would you elaborate on why you think the discussion > should not be continued (if the participants wish)? I got a note from Upper Management to that effect. Since then the same Upper Management Person suggested 3RD_PARTY_VMS_SOFTWARE as a suitable home as Pick will run on a VAX. On the other hand, I've had several people suggest that we should continue here, so I suppose we may as well. ================================================================================ Note 198.16 What is "Pick"? 16 of 23 EISNER::MERRIMAN "George Merriman - CCA/NY" 11 lines 30-AUG-1988 10:44 -< I don't think this is the place for MS/DOS >- -------------------------------------------------------------------------------- > Okay George, but what about talking about things like the Ultimate > co-processor boards? I don't know what this device is, but I assume it is one of these things that run MS/DOS on a VAX. Clearly any discussion relating to issues like getting the board hooked up to the bus belongs here. However, I don't see how the fact that this device exists is a good reason to turn HARDWARE_HELP into a forum for MS/DOS questions. Discussion on the integration of the board's operation with VMS probably ought to go in the VMS conference along with the other VMS driver discussions. ================================================================================ Note 198.17 What is "Pick"? 17 of 23 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 11 lines 30-AUG-1988 12:49 -< ultimate? co-processor >- -------------------------------------------------------------------------------- >>> < Note 198.16 by EISNER::MERRIMAN "George Merriman - CCA/NY" > >>> -< I don't think this is the place for MS/DOS >- >> I don't know what this device is, but I assume it is one of these >> things that run MS/DOS on a VAX. Funny, I must have it backwards, I thought it was a C-Mos MV3 card to run on either any MS-DOS machine or to snuggle in with QuickWare's 45mhz-ish 4 card replacement for the first 21 cards in your 11/70... ================================================================================ Note 198.18 What is "Pick"? 18 of 23 EISNER::COY "Dale E. Coy" 5 lines 30-AUG-1988 13:55 -< A different brand of cigar >- -------------------------------------------------------------------------------- >> I don't know what this device is, but I assume it is one of these >> things that run MS/DOS on a VAX. No, it's one of those things that runs PICK on a VAX. Absolutely no relationship between this and a PC. ================================================================================ Note 198.19 What is "Pick"? 19 of 23 EISNER::TERRAZAS "Mike Terrazas" 21 lines 30-AUG-1988 14:30 -< Yeah but... (2nd try) >- -------------------------------------------------------------------------------- > > Okay George, but what about talking about things like the Ultimate > > co-processor boards? > I don't know what this device is, but I assume it is one of these > things that run MS/DOS on a VAX. Clearly any discussion relating to > issues like getting the board hooked up to the bus belongs here. The Ultimate co-processor is a board which allows you to run Pick on a VAX, it runs the whole set of Pick users under it with only stub processes on the VAX cpu to handle peripherals. There's much more to it than just "hooking it up". That's why I asked in this forum. > Discussion on the integration of the board's operation with VMS > probably ought to go in the VMS conference along with the other VMS > driver discussions. Again, there are some legitimate HARDWARE questions, and since Pick *ain't* VMS and we usually need this hardware... where would we go? MikeT ================================================================================ Note 198.20 What is "Pick"? 20 of 23 EISNER::MERRIMAN "George Merriman - CCA/NY" 45 lines 30-AUG-1988 23:23 -< let's keep this topic on Pick >- -------------------------------------------------------------------------------- > < Note 198.19 by EISNER::TERRAZAS "Mike Terrazas" > > -< Yeah but... (2nd try) >- First let me apologize for not knowing what an "Ultimate co-processor board" is. Had I known that it related to Pick rather than MS/DOS I would not have phrased my answer to your question in terms of MS/DOS. However, I still do not believe that HARDWARE_HELP is a suitable forum for the sort of general questions and discussion about an operating system (ANY operating system) as in note 198.0. Do you suggest that since VAX ELN runs on the VAX Real Time Accelerator co-processor we should be discussing ELN questions here? For that matter a MV-II CPU board plugs into a Q-bus and runs VMS, but I don't think that means we should be discussing VMS issues here. As note 1.0 says: "The purpose of this conference is to discuss hardware problems, glitches, ECOs, diagnostics, how-to questions, etc." As to the question of what to do when there is no obvious conference for a note such as the one in 198.0, I don't have a good answer. As the discussion about Pick started here I guess we should continue it in this topic until the non-DEC software conference comes on line. On the other hand, we have generated 20 notes in this topic: a question about Pick, several good answers to that question, several notes on the spelling of ALL-IN-1 (TM) (did I get it right?) and quite a few notes on the charter of this conference and the administration on DECUServe in general. I think we should limit this topic to Pick and carry on the other threads elsewhere. If anyone wants to discuss the scope and definition of HARDWARE_HELP please feel free to start a new topic in this conference. Anything else should go to SOAPBOX, PROPOSALS, etc. George Merriman -- HARDWARE_HELP Co-moderator ================================================================================ Note 198.21 What is "Pick"? 21 of 23 EISNER::TINKELMAN "Bob Tinkelman - CCA" 6 lines 1-SEP-1988 17:38 -< The ultimate contact >- -------------------------------------------------------------------------------- > Any clues, info or pointers to better conferences? From my notes of the NY Metro LUG meeting last November, you could try to call Art Rosenfeld at 1-201-887-9222, director of marketing for Ultimate (PICK on the VAX, via a co-processor board). They'll at least send you lots of glossy brochures. ================================================================================ Note 198.22 What is "Pick"? 22 of 23 EISNER::LYMAN "Rich Lyman" 14 lines 8-SEP-1988 19:50 -< Information on Ultimate PICK implementation >- -------------------------------------------------------------------------------- I am the person at Ultimate doing the interface between the Ultimate co-processor and the VMS operating system. If you have any questions or requests for information, please send me a mail message so we can try to keep this conference clean. The information that has been discussed so far is essentially correct about the Ultimate co-processor. Pick is an operating system optimized for database applications. It uses an "english" type query interface and is very easy for dumb users to learn. The coprocessor resides in either a Unibus or a Qbus and uses VMS disks,tapes, printers and terminals as its peripherals. ================================================================================ Note 198.23* What is "Pick"? 23 of 23 EISNER::MERRIMAN "George Merriman - CCA/NY" 8 lines 10-SEP-1988 07:20 -< Pick topic started in OTHER_SOFTWARE_HELP >- -------------------------------------------------------------------------------- I've copied several of the notes in this topic to topic 6 in OTHER_SOFTWARE_HELP. I suggest that any further discussion related to the Pick operating system continue there. Further hardware-related discussion of the Ultimate co-processor should continue here. To add OTHER_SOFTWARE_HELP to your notebook use KP7 (SELECT). George Merriman -- HARDWARE_HELP Co-moderator  ================================================================================ Note 199.0 C.P.I. CT1D8C12 Printer Help. 1 reply EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 22 lines 27-AUG-1988 15:05 -------------------------------------------------------------------------------- We have recently inherited a band printer, a Computer Peripherals Inc. model CT1D8C12 that is about ten or twelve years old (I think). Of course, there were no documents, cables, or cards with it. I have never heard of the company and cannot find any reference to it. This was supposed to have been used as the system printer on a PDP-11/60 for a while and then on a VAX-11/780. It is a band printer with 96- and 64-character bands, it uses a two-inch wide ribbon (we got a bunch, in plain white boxes), and has a 50-pin signal connector (two rows of 13 and two rows of 12); looks like it might be old Dataproducts. The new manager at the place we got it from said that he had heard it was LP11 compatible but could not confirm it. It is in a chest- high cabinet painted DEC blue and ivory but has no DEC part number or model numbers. Can anyone help me with information? I would like to know what the signal interface is, and would like to get readable copies of whatever documentation (installation, user guide, service manual) might be around. Was this a DEC-marketed product? Thanks... Mark S. ================================================================================ Note 199.1 C.P.I. CT1D8C12 Printer Help. 1 of 1 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 12 lines 28-AUG-1988 01:13 -< Sounds like Dataproducts to me >- -------------------------------------------------------------------------------- >> ...has a 50-pin signal connector (two rows >> of 13 and two rows of 12); looks like it might be old Dataproducts. >> The new manager at the place we got it from said that he had heard >> it was LP11 compatible but could not confirm it. If it is the Dataproducts interface, it is LP11 compatible. For that matter, the Centronics interface is LP11 compatible too, and the instructions for making the cable are in this conference. The LP11 logic is switched from one to the other by tying one connector pin to either +5 or ground. Alan  ================================================================================ Note 200.0 Need help moving printer to MV2 2 replies EISNER::BRUCE "Barton F. Bruce - CCA/MA" 22 lines 29-AUG-1988 15:25 -------------------------------------------------------------------------------- We have a Printronix P-600 (Dec LXY-22) that has served well on a Unibus PDP-11 for years and wanted to move it to a MV2 and so put in a LPV-11 card, and using the same old cable we get garbage. A DEC field service type suggested he be allowed to order 2 different parts but said they would take a very long time to come and would then cost $5 each and $450 to install. He also said we should have the new FCC type cable, but our old Berg (at the card) to Winchester (Dataproducts interface at the printer) cable still works when connected to the 11. The parts he suggested are 2 resistor packs, one 220 ohms, and the other 330 ohms and they replace the 1k ones that are there in the printer. Printronix supplied part numbers for 3 major suppliers of these, and suggested any electronics supply house or their local dealer could get them instantly. That is easy, and I will do it, but I wonder if anyone else has had similar problems moving a printer from an M7258 card on Unibus to an LPV-11 on Q bus, or has had problems with the terminating resistors on a Printronix printer. Is this all I have to do, or is there something else I need to worry about? ================================================================================ Note 200.1 Need help moving printer to MV2 1 of 2 EISNER::CONROY "Alan Conroy" 5 lines 30-AUG-1988 11:12 -< Try a different kind of cable >- -------------------------------------------------------------------------------- I know that when we got our P600 for our PDP-11/70s that the "standard" cable wouldn't work. We had the same symptom: garbage. It was fixed by using a slightly modified cable. Sorry, but it has been so long that I can't remember what was done. The P600 has since left for other parts. Hope this helps. :-) ================================================================================ Note 200.2 Need help moving printer to MV2 2 of 2 EISNER::NELSON 14 lines 9-SEP-1988 09:09 -< How about a serial Interface >- -------------------------------------------------------------------------------- > We have a Printronix P-600 (Dec LXY-22) that has served well on a > Unibus PDP-11 for years and wanted to move it to a MV2 and so put > in a LPV-11 card, and using the same old cable we get garbage. We have several P-300 and P-600 printer (some with DEC labels on them) and have solved this problem by adding the serial interface to the Printronix. Our service company (not DEC) added it for less than $200. Since you already spent the money for the LXY11, this may not be what you're looking for but it's another way out. Good luck, Jay  ================================================================================ Note 201.0 Problems with TE16 on 11/70 6 replies EISNER::NORTON "Bill Norton" 20 lines 30-AUG-1988 11:13 -------------------------------------------------------------------------------- OK, how about some *real* hardware help? My 11/70 was removed from the service contract Aug 1st. On the 15th nightly backups started failing.(someone must have told it it was an orphan 8-) ) The physical symptom is that partway through the BRU, the TE16 becomes unresponsive, and RSX reports error -59. The reel won't unload until the drive(formatter?) has been power cycled, and the select light on the tape drive in use at the time of failure is stuck on. I've tried switching to the extra TM03 formatter, and put in a replacement RH70 controller. Nothing seems to help. (The second TE16 on the TM03 is powered-down. It has its own problems.) The serviceman is getting tired of offering *free* advice, (OTOH its been worth every cent! 8-) ) Can anyone suggest something else to investigate? ================================================================================ Note 201.1 Problems with TE16 on 11/70 1 of 6 EISNER::CONROY "Alan Conroy" 4 lines 30-AUG-1988 11:17 -< What a response! >- -------------------------------------------------------------------------------- > The physical symptom is that partway through the BRU, the TE16 > becomes unresponsive... I wasn't aware that TE16s EVER were anything but unresponsive! ;-) ================================================================================ Note 201.2 Problems with TE16 on 11/70 2 of 6 EISNER::NORTON "Bill Norton" 8 lines 1-SEP-1988 10:37 -------------------------------------------------------------------------------- C'mon, get serious. This venerable system still supports several users. The only problem we have with paying to get it fixed is that it could easily cost more than the CPU is worth, and more than a replacement 11/73 system, to fix just the tape drive. So before we s**t-can the whole thing, I figured I'd try to repair it. There must be *some* hardware types out there that know something about TE16's. ================================================================================ Note 201.3 Problems with TE16 on 11/70 3 of 6 EISNER::CONROY "Alan Conroy" 2 lines 1-SEP-1988 12:28 -< In all seriousness >- -------------------------------------------------------------------------------- You might want to consider buying a used TE16 from a broker. They can be had for cheap. Just as a last resort idea... ================================================================================ Note 201.4 Problems with TE16 on 11/70 4 of 6 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 8 lines 1-SEP-1988 16:26 -< 125 ips may cost same $ >- -------------------------------------------------------------------------------- >>> You might want to consider buying a used TE16 from a broker. They can be >>> had for cheap. Just as a last resort idea... Better yet (maybe) you can get a used TU77 for about the same price! if you are shopping for used old Dec drives. And you can get the latest tu 78 ECOs for mechanical tape threading problems so these might occasionally be able to self thread the first time from Easy-Load (tm) reels. ================================================================================ Note 201.5 Problems with TE16 on 11/70 5 of 6 EISNER::CONROY "Alan Conroy" 5 lines 2-SEP-1988 11:39 -< Say what? >- -------------------------------------------------------------------------------- > And you can get the latest tu > 78 ECOs for mechanical tape threading problems so these might occasionally > be able to self thread the first time from Easy-Load (tm) reels. The TU78 ECO works on the TU77? ================================================================================ Note 201.6 Problems with TE16 on 11/70 6 of 6 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 22 lines 2-SEP-1988 12:59 -< yes TU78 eco on TU77! >- -------------------------------------------------------------------------------- >> < Note 201.5 by EISNER::CONROY "Alan Conroy" > >> The TU78 ECO works on the TU77? The hardware only ecos to do with tape path improvements are usable on the tu77, as it is the same Pertec hardware. The improvements to the electronics to make it possible to read difficult to read tapes (e.g. SDC tapes) don't apply. The repositioning of interior pieces to improve interior cooling done in designing the tu79, also do not apply. I have none of these, so this is all hear-say, but from hopefully reliable sources. There are still DECies who were involved in the original OEM choice that lament not using the large STC drives that were being considered at the time. They were not disqualified on technical grounds, but on other perceptions of the supplying company. I assume many users lament too...  ================================================================================ Note 202.0 rtVAX 1000 1 reply EISNER::TINKELMAN "Bob Tinkelman - CCA" 247 lines 30-AUG-1988 11:31 -------------------------------------------------------------------------------- The attached description was extracted from the DEC Electronic Store.  DESCRIPTION Low cost, low overhead, and predictable realtime response. The rtVAX 1000 contains a single quad-height KA620 Single Board Computer, which includes the Realtime MicroVAX 32-bit central processor chip (with modified memory management). Also available are a floating point coprocessor chip, 1 Mbyte of on-board memory, Q22-bus interface, Q22 map for DMA transfers, interval timer, boot and diagnostic facility, console serial line unit and time-of-year clock with optional battery backup support. The rtVAX 1000 uses CD backplane interconnect and a 50-pin connector to communicate with up to two high-speed, tightly coupled memory expansion modules, each of which can contain 4 or 8 Mbytes of local memory. IS THE RTVAX 1000 FOR YOU? VAX architecture, Q-bus functionality, and VAXELN realtime environment in one package. The rtVAX 1000 Realtime Computing Platform is designed for dedicated realtime engineering and scientific applications in manufacturing, government and defense, medicine, communications, and laboratories. It combines the Q-bus system package, the KA620 single-board computer, the VAXELN target runtime environment, and a one-year Digital service warranty. The system's modularity makes it easy to add on and upgrade as necessary, using Q-bus options supported under the VAXELN target runtime environment. The rtVAX 1000 system features the time-tested VAX architecture with its powerful instruction set, its rich selection of data types, and the ability to share resources with other VAX systems in a networked environment. Q-bus functionality and the wide range of existing I/O devices from Digital and third-party suppliers are invaluable benefits of the rtVAX 1000 Realtime Computing Platform. VAXELN software provides the industry-standard VAX/VMS development environment and an executive for realtime applications that is optimized for speed, efficiency, and simplicity. HIGHLIGHTS . VAX/Q22-bus architecture and MicroVAX II performance are optimized for realtime computing. . VAXELN offers both development advantages and an efficient runtime executive. . The KA620 single board computer is the key to MicroVAX II performance. - Realtime MicroVAX CPU chip-78R32 CPU - MicroVAX 78132 floating-point unit - 1 Mbyte of onboard parity DRAM memory - Memory management: single-level virtual-to-physical address translation - Powerup bootstrap diagnostic ROM set - Console serial line, with rear-access baud-rate selection - Battery back-up hooks - Time-of-year clock . Main memory can be expanded to 16 Mbytes. . Flexible packaging styles include 19-inch rack, pedestal, tabletop, or H9642JA/JB cabinet. . Entry-level, network-node, and mass-storage configurations are available. . BA23 8-slot system box provides low cost and modular flexibility. THE Q-BUS GETS YOU TO MARKET QUICKLY The Q-bus is an acknowledged standard in realtime markets, offering bus speeds up to 3.3 Mbytes per second. The variety of realtime I/O devices already available and supported by Digital and other vendors helps bring complete solutions to market with the shortest possible development cycle. The rtVAX 1000 system supports the standard Q-bus quad form factor, providing investment protection for those moving from 16-bit Q-bus to 32-bit Q-bus solutions. ONE ARCHITECTURE FROM DEVELOPMENT LAB TO TARGET SYSTEM For years, VAX systems have been the machines of choice for developing applications on industry-standard processors. Developers have come to rely on VAX/VMS systems for the numerous high-level languages, utilities, and development tools that make their job easier. With the rtVAX 1000 computing platform, you don't have to leave the VAX family to develop powerful solutions at the system level. You can continue to use your VAX/VMS system from MicroVAX 2000 through VAX 8800 clusters for your development environment, with the rtVAX 1000 system as your runtime platform. VAXELN & RTVAX 1000--OPTIMIZED REALTIME APPLICATIONS The VAXELN runtime environment builds on Digital's strengths in communications and networking to provide an ideal base for connected and distributed computing applications. It consists of both a software development toolkit and an optimized executive for realtime applications. As a development environment, the VAXELN toolkit runs on any VMS or MicroVMS host to produce tight, executive-only software for runtime applications. VAXELN gives you all the advantages of the renowned VAX/VMS development environment to create applications with the performance and predictability realtime solutions require. The VAXELN toolkit gives you access to the compilers, editors,debuggers, and other utilities that have established VMS as a leading development environment. VAXELN software lets you debug systems remotely and symbolically from VAX/VMS host, significantly speeding the application development process. For general application code you can use existing or new FORTRAN, C, Ada, or EPascal software. For performance-sensitive code such as device drivers, you can use extensions to the EPascal compiler, Ada or C runtime libraries. These tools deliver the performance of assembly code with the debugging ease of high-level languages. The VAXELN toolkit includes an optimizing EPascal compiler, C, and FORTRAN runtime libraries, Ebuild (a menu-driven, system-building utility), a symbolic debugger, and a number of runtime software modules that together create a tailored realtime application environment. Available options include VAXELN/Ada with its runtime library and the Rdb/ELN relational database management system. The VAXELN development toolkit is ideal for distributed computing applications. VAXELN applications can span multiple systems across an Ethernet local area network (LAN). Processes executing in parallel on separate systems can transparently exchange information via the Ethernet using a message-passing mechanism built into VAXELN applications. This server arrangement allows an application to share vital computing resources among several realtime processors. KA620--THE HEART OF THE RTVAX 1000 SYSTEM The heart of the rtVAX 1000 Realtime Computing Platform is the KA620 single- board computer. The realtime KA620 single-board computer is powered by the rtVAX 78R32 microprocessor and the 78132 floating point coprocessor. Together they provide the MicroVAX subset of the native VAX instructions to achieve optimum speeds. The rtVAX 78R32 microprocessor chip implements Digital's proven MicroVAX II chip technology. The 78132 high-performance floating-point coprocessor functions as an extension of the primary processor's data paths. It handles in hardware the major VAX floating-point types: single-precision (F floating), double-precision (D floating), and extended double-precision (G floating). And, it operates on byte, word, long-word, and quadword data types. Overall, the 78132 coprocessor accelerates floating-point instruction speeds by approximately 50 times, relative to emulation. By handling signed-integer multiply and unsigned-integer divide operations, the coprocessor doubles even-integer multiply and divide. ENTRY-LEVEL PLATFORM All models include the KA620 single-board computer. BA23F is the tabletop, pedestal enclosure; BA23R is the rackmount enclosure. ---------------------------------------------------------------------------- 620QY-A2 BA23F 120V 620QY-A3 BA23F 240V 620QZ-A2 BA23R 120V 620QZ-A3 BA23R 240V NETWORK PLATFORM All models include the KA620 single-board computer and a DEQNA Ethernet-Q-bus synchronous communications controller. ---------------------------------------------------------------------------- DH-620Q1-AA* BA23F 120V DH-620Q1-A3 BA23F 240V DH-620Q2-AA* BA23R 120V DH-620Q2-A3 BA23R 240V MASS-STORAGE PLATFORM All models include the KA620 single-board computer, plus the half-height RD31, a 20-Mbyte Winchester disk, and the half-height RX33, a 1.2-Mbyte diskette. ---------------------------------------------------------------------------- DH-620Q1-CA* BA23F 120V DH-620Q1-C3* BA23F 240V DH-620Q2-CA* BA23R 120V DH-620Q2-C3* BA23R 240V DH-620Q5-AA* H9642 120V DH-620Q5-A3 H9642 240V *The VAXELN runtime license is included with all DH-AA, DH-CA and DH-C3 configurations. FLEXIBLE PACKAGING Flexible packaging provides convenient building blocks for any realtime application. The rtVAX 1000 system-built around the low-cost, flexible BA23 system box- is available in several configurations. Standard package configurations include disk and network models mounted in a rack, a pedestal (which can convert to tabletop use). or an H9642 cabinet. The base hardware systems are listed above. SPECIFICATIONS Physical Characteristics ------------------------- Rackmount Model: Height 13.3 cm (5.2 in) Length 64.3 cm (25.3 in) Width 48.3 cm (19.0 in) Weight 24.0 kg (53.0 lb) Pedestal Model: Height 64.2 cm (25.3 in) Length 72.6 cm (28.6 in) Width 25.4 cm (10.0 in) Weight 31.75kg (70.0 lb) Tabletop Model: Height 17.7 cm (7.0 in) Length 72.6 cm (28.6 in) Width 56.2 cm (22.1 in) Weight 29.5 kg (65.0 lb) Cabinet Model: Height 106.8 cm (42.0 in) Length 91.4 cm (36.0 in) Width 64.6 cm (25.4 in) Weight 133.81kg (295 lb) Electrical Requirements ----------------------- Input Line Voltage 120V 240V Frequency 60 Hz 50 Hz Running current 6.0A 3.0A Power Consumption (max) 345W 345W DIGITAL SERVICE IS AVAILABLE WORLDWIDE Digital provides extensive field service, software service, and education service to support our products. Worldwide, more than 30,000 Digital customer-service employees work to maintain hardware and software, to keep systems up-to-date, and to train users. Whatever your service requirement, Digital gives you a comprehensive solution. MORE INFORMATION Because of configuration variations, this system presently cannot be ordered through the Electronic Store. For additional technical or ordering information, please refer to Digital's Systems and Options Catalog or contact your local Digital Sales Representative or Digital Distributor. ================================================================================ Note 202.1 rtVAX 1000 1 of 1 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 20 lines 30-AUG-1988 13:14 -< DEQNAs - NOT again! >- -------------------------------------------------------------------------------- >>< Note 202.0 by EISNER::TINKELMAN "Bob Tinkelman - CCA" > >> -< rtVAX 1000 >- >> All models include the KA620 single-board computer and a DEQNA Ethernet-Q-bus >> synchronous communications controller. ^^^^^ Dec thinks they found a market for DEQNAs!... Dec has been busy telling us DEQNAs are no longer 'ok'. Some people are waiting to see if units on contract will be 'fixed' or 'replaced' free. Now Dec turns and tries to sell them again! This is actually probably because one group in DEC comitted to 'take' some number of the offending cards a long time ago, and now can't get out of the commitment. The producing group gleefully makes more cards. If upper management could only step in and 'fix' things, not only would the customers be happier, but they might have better feelings about DEC in general. Now how do we get this message to their upper management?  ================================================================================ Note 203.0 FAST MASSBUSS on your 11/84 No replies EISNER::BRUCE "Barton F. Bruce - CCA/MA" 24 lines 1-SEP-1988 16:45 -------------------------------------------------------------------------------- There IS an ECO to RH11s that will let them work with such fast drives as RM05, and RM03s. You don't have to use a spun down to 2400 from 3600 rpm RM02 just because you have an 11/84. You don't have to dump your RM05s imply because you don't have an RH70. It is several years old now and I assumed everyone knew about it, but in just this last week a used DEC dealer and a Dec FE, both needing the info, didn't know anything about it. The old RH11 was engineered in the good old days when 11/45s had 2 unibuses. The A one was the normal one, and the B one was special. The RH11 has logic in it to allow going into HOG mode (allowing the needed faster transfer) but conditioned to only allow it on the B bus. The ECO simply was an etch cut to allow HOG mode on either A or B busses, and a second cut to allow the RH11 to drop out of HOG mode if its BIG (in those days) 16 word silo is empty. The final version of the ECO was going to have the cut wires extended to a spare header to allow easy jumper insertion/removal.  ================================================================================ Note 204.0 Disaster recovery & DEC No replies EISNER::SCOPELLITI "Whatsa behind is uva no importan" 8 lines 7-SEP-1988 00:18 -------------------------------------------------------------------------------- We're investigating DEC's Recover-all service (DEC will replace anything destroyed by an 'act of God' [like floods, fire, etc.] in ultra-quick time). Has anyone had any experience using this service. i.e. has anyone suffered a loss and had DEC do its thing? In a similar vein, anyone have any comments about DEC responding to a disaster at a site that doesn't have Recover-all? Are they quick to replace the gear?  ================================================================================ Note 205.0 Home Brew DS500? 15 replies EISNER::BRUCE "Barton F. Bruce - CCA/MA" 71 lines 16-SEP-1988 00:52 -------------------------------------------------------------------------------- Could you brew your own DS500 or recycle on hand parts to enhance an existing one? Are you really limited by Dec's marketing / packaging / pricing decisions? From what I understand, the DS500 is nothing more than a BA213 Q bus box, an 11/53 (which size?) an ethernet board (DEQNA or DELQA?) and up to 8 of 16 or 8 port terminal mux cards that must be similar to the DHV/DHQ types, because they all work with the same driver. Could you use other brand comm cards? If you had an Emulex card to run 64 of their ports on remote 16 port panels, you would have 7 comm slots to spare. This stuff was expensive once, but if you already had it, could you recycle it? In fact you can get at least 16 ports WITH MODEM CONTROL on a quad card almost anywhere EXCEPT Dec. If you had an 11/73 with some brand x memory, would that work in a BA23 rather than an 11/53 in a BA213. Are there special boot roms? and where are they on a real DS500. If the code on such roms is proprietary and only comes with the DS500 H/W, and not with the DS500 S/W kit, would it be easy to make legal functionally equivalent roms. There are 2 unused BA213 slots in a DS500. Will more comm cards or Q-bus memory help? If you are using a quad 73/83 in C-D interconnect slots, PMI (11/83/84 type. or Clearpoint QED-2) memory should work much faster than vanilla Q bus memory, but must be ahead of the CPU on the bus. Is the 64 port with modem control, 128 without limit due solely to ports on their current cards, or to performance issues with 15mhz J11s and limited memory in either 11/53 cpu? Will the promised 32 port cards be without modem control, but otherwise supported on the current DS500 software? If CPU speed is a limit, will the 30+ mhz J11+ chips Harris HAS play immediately in the current 11/53 cards, or in the 11/73/83 quads or in the 11/73 duals or is there a lot of redesign necessary. Some might even ask how well they would play in an HSC70, but that is getting away from terminal servers. (I hear they play very well in some brand x boards!). This next suggestion is a 'cute' trick Bob Tinkelman once suggested when a customer with a large Bridge (brand) terminal server net needed to bridge between that net and a Dec terminal server net. The Bridge folks had a U bus card that plugged into your CPU and came out with a tranceiver cable to ethernet. To software, this card looked as though you had put in 64 ports of DEC terminal mux cards and had wired them into a 64 port Bridge terminal server. If Bridge had a Q bus card like that U bus one (they didn't then, may now), rather than plugging it into a host system, plug it instead into a DS500 box! Instant bridge between the two worlds. There may be other brand nets (Able?) that this "their Q in Dec's DS500" type trick might work for, too. This would help Dec get their servers in where other brand stuff is entrenched. Now the BIG question. IF you made such a box, with very little DEC in it, but paid DEC the lic. fee for DS500 S/W, would it all be legal, or could they legally refuse to sell you the license? What if you were a still-feeling-screwed LAT11 customer simply wanting to upgrade? (This is in hardware because that seems best to me. It could plausibly go elsewhere, and I don't care at all if moderators move it. ) ================================================================================ Note 205.1 Home Brew DS500? 1 of 15 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 16 lines 16-SEP-1988 09:11 -< STANDARD CARD >- -------------------------------------------------------------------------------- > If you are using a quad 73/83 in C-D interconnect slots, PMI > (11/83/84 type. or Clearpoint QED-2) memory should work much > faster than vanilla Q bus memory, but must be ahead of the CPU on > the bus. > If CPU speed is a limit, will the 30+ mhz J11+ chips Harris HAS > play immediately in the current 11/53 cards, or in the 11/73/83 > quads or in the 11/73 duals or is there a lot of redesign > necessary. Some might even ask how well they would play in an According to my friends at NaC the CPU board is a standard PDP-11/53 CPU card with 1.5MB. There are only two mods. The clock chip is 18MHZ rather than the standard 15MHZ. And of course the firmware has been changed. Other than that it is suppose to be the same. The CPU should offer 90 percent of the performance of an 11/83 with PMI. ================================================================================ Note 205.2 Home Brew DS500? 2 of 15 EISNER::MAYHEW "Bill Mayhew" 6 lines 16-SEP-1988 16:40 -< 3/4 of a grain of salt? >- -------------------------------------------------------------------------------- Slightly off topic... > According to my friends at NaC... Having seen this acronym before on RONNIE, also without definition, I am led to ask: Whatzit? ================================================================================ Note 205.3 Home Brew DS500? 3 of 15 EISNER::DELARISCH "Arnold S. De Larisch" 11 lines 16-SEP-1988 17:17 -< NaC = Networks and Communications >- -------------------------------------------------------------------------------- >> < Note 205.2 by EISNER::MAYHEW "Bill Mayhew" > >> -< 3/4 of a grain of salt? >- >> > According to my friends at NaC... >> Having seen this acronym before on RONNIE, also without definition, >> I am led to ask: Whatzit? NaC = Networks and Communications -Arnold ================================================================================ Note 205.4 Home Brew DS500? 4 of 15 EISNER::BYRNE_C "Charlie Byrne" 8 lines 19-SEP-1988 01:50 -< What is RONNIE? >- -------------------------------------------------------------------------------- >> < Note 205.2 by EISNER::MAYHEW "Bill Mayhew" > >> -< 3/4 of a grain of salt? >- >> Having seen this acronym before on RONNIE, also without definition, >> I am led to ask: Whatzit? -Arnold Speaking of WHATZIT, I'd like to be enlighted on what RONNIE is. ================================================================================ Note 205.5 Home Brew DS500? 5 of 15 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 11 lines 19-SEP-1988 08:32 -< OEM ONLY SYSTEM >- -------------------------------------------------------------------------------- > Speaking of WHATZIT, I'd like to be enlighted on what > RONNIE is. An approximation of the name is.... Remote Online Network Information system. It is the online information system for Digital's marketing partners (the OEM's). It includes Sales Update information, price queries, and other misc. information that is useful to someone selling DEC systems. ================================================================================ Note 205.6 Home Brew DS500? 6 of 15 EISNER::MAYHEW "Bill Mayhew" 10 lines 19-SEP-1988 13:43 -< Etymology of "RONNIE" >- -------------------------------------------------------------------------------- I believe the current translation is: Remote Online Network for News and Information Exchange "Remote" used to be "Resellers'", but changed when, I gather, DEC opened the system to more of its own sales staff. Rumor has it the original developer of the system was named Ronnie, and the acronym does have that character of "how can we find words to match this acronym?" to it... ================================================================================ Note 205.7 Home Brew DS500? 7 of 15 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 19 lines 21-SEP-1988 15:54 -< I like the real one ! >- -------------------------------------------------------------------------------- Can't help you with the "home brew" version but I would like to say as the owner of a **REAL** DS500 I have been very please so far. For all of you who are tired of ... swapping NVRAMs between DS200's when one croaks or ... have to use a PC to download the params again or ... have a pseudo-terminal programs that does it for you or ... have bought the Terminal Server Manager ... The DS500 comes with a configuration program that that let's you modify **ALL** the params on all the ports !! A quick reboot (or load) and your on your way !! ================================================================================ Note 205.8 Home Brew DS500? 8 of 15 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 56 lines 21-SEP-1988 20:53 -< Someone home-brewed already >- -------------------------------------------------------------------------------- Since I first posted the .0 note here there has been a steady trickle of additional information that seems of interest. First off, Able has already made their own DS500 using DEC's software. It is their EH128, and comprises an 11/53, a DEQNA, a BA11-MA, and their own Q bus card that looks like 128 ports WITH modem control and exits via twisted pair at 1mhz (not T1) to their EasyWay net. Apparently all they did was to use modified proms that do not do the same boot up diagnostics that a normal DS500 would do. They sell you their box and with it a license to run the DEC server software. Their part of the agreement with DEC includes that they buy CPU, Ethernet controller, and BA11 box all from DEC. Normal DS500 S/W is what gets downloaded, and a site that already has a DS500 distribution kits doesn't need anything else. Their 11/53 is only the 15mhz one as opposed to the 18mhz one used in the real DS500. The DELQA cards WON'T work with current DS500 S/W. They ship DEQNAs. This arrangement between Able and DEC (probably via Dec's Hamilton Avnet distributor) is not available to Emulex... The DS500 CX... cards are DHQ11 class that are like DHUs, not DHVs. With the proms Able is using, but in a bigger box, as the one they are using has little spare space, one could apparently run DHV class generic cards of any brand (maybe). The DS500 apparently will reject the non CX... cards on bootup, but can be coaxed into then continuing and the driver will then work with them. The modified proms eliminate this. Another vendor, Micom, has a 128 port Q bus card #1480 that lets you exit in their T1 format on twisted pair at 1.544 mhz and go either to up to 4 of their 32 port panels, or into a big Micom terminal switch and then off to the 32 port panels. If this card looks like a DHQ, it might plug right into a DS500, else you would need to do the prom trick Able has done. If you have a big Micom net and Dec Lat types have been trying to stick their foot in your door, you could ask them to sell you a DS500 with the same type of proms Able uses but none of the otherwise compulsory cards (normally they MAKE you buy 2 comm cards for starters - but that would take you past 128 ports). That is only if you are not inclined to experiment. I have just heard from Emulex that my old 16 port CP11 and CP12 panels that used to work with U bus CS11s or CS32s will not get support on the Q-bus CS04 (at least for a long while) that only works with the newer CP34 16 port panels. These are all smart and expensive remote 16 port panels several of which daisy chained to a relatively inexpensive controller card sitting on your bus. The Unibus cards can use either old or new panels, the Q bus card can't. This was what I was hoping to use in MY home brew box when I started this topic. Tough luck... Does anyone have any other ideas? ================================================================================ Note 205.9 Home Brew DS500? 9 of 15 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 23 lines 22-SEP-1988 02:04 -< A different DS 500 view >- -------------------------------------------------------------------------------- We've added several DS500s after having extensive experience with DS100s and DS200s. As opposed to Jack's experiences, we've had massive problems with ports changing settings on the fly. Believe it or not, we've had to set all ports to AUTOBAUD to make the things work. If we left them at a fixed speed, they'd dynamically change! Is this a feature? :-{ We're also experiencing hung ports among other niggling problems. DEC regional support will be around here soon to see what the h**l is going on. Meanwhile, the DS200s right next to the DS500 just cruise along. We have TSM - absolutely love it. Particularly the ability to upload a server's settings to a command file, and command files in general. DEC recommends having two load hosts for DS500s (that way if one is down and the server reboots you're still on the air). Well.. that gets you into the image file synchronization problem - you gotta be sure to make the same change to both images or copy one to the other or.. yuk. I'd just as soon use the 500s to measure gravity and install DS200s. Wonder if a DS500 falls at the same rate as a cannon ball? ================================================================================ Note 205.10 Home Brew DS500? 10 of 15 EISNER::PRIGOT "Jonathan M. Prigot" 11 lines 22-SEP-1988 11:19 -< BACKUP (load) HOSTS >- -------------------------------------------------------------------------------- > DEC recommends having two load hosts for DS500s (that way if one > is down and the server reboots you're still on the air). Well.. > that gets you into the image file synchronization problem - you No it doesn't! The DS500 knows who loaded it the last time and expects that host to reload it. If the primary host can't do it, it goes down the list you specified with DEFINE SERVER BACKUP HOSTS. If it doesn't "hit" that way, THEN it does a multicast to all hosts. If you haven't DEFINEd your backup hosts, it still looks first to the host that loaded it and failing (90 seconds) that, will multicast. ================================================================================ Note 205.11 Home Brew DS500? 11 of 15 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 4 lines 22-SEP-1988 11:27 -< DS500=DECSA? >- -------------------------------------------------------------------------------- One interesting fact I found out from CSC is that the DS500 code is the same as the DECSA ETS code, with a little conditional assembly in it to allow for the difference between the DECSA's 11/24 and the DS500's 11/53. ================================================================================ Note 205.12 Home Brew DS500? 12 of 15 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 7 lines 22-SEP-1988 12:39 -< DEC paranoid? >- -------------------------------------------------------------------------------- One of the changes to the boot proms that Able had to do was apparently to eliminate a timing test that only succeeded on an 18mhz 11/53. It failed if those proms were stuck in a vanilla 15mhz card! Is Dec afraid of what users might do, or is this simply valid hardware checking...? ================================================================================ Note 205.13 Home Brew DS500? 13 of 15 EISNER::KILGALLEN "Larry Kilgallen" 4 lines 22-SEP-1988 13:34 -< Sometimes paranoia is justified >- -------------------------------------------------------------------------------- > succeeded on an 18mhz 11/53. It failed if those proms > were stuck in a vanilla 15mhz card! Is DEC afraid of errors Field Service might make? ================================================================================ Note 205.15 Home Brew DS500? 15 of 15 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 21 lines 30-SEP-1988 02:05 -< now LAT <-> CASE, too >- -------------------------------------------------------------------------------- >>> < Note 205.8 by EISNER::BRUCE "Barton F. Bruce - CCA/MA" > >>> -< Someone home-brewed already >- >>> First off, Able has already made their own DS500 using DEC's >>> software. It is their EH128, and comprises an 11/53, a DEQNA, a >>> BA11-MA, and their own Q bus card that looks like 128 ports WITH >>> modem control and exits via twisted pair at 1mhz (not T1) to >>> their EasyWay net. The CASE Communications DCX networks can't go directly to DS500 look-alike boxes, but get there simply and fast via an Able gateway box (called "REMOTEgate") that connects Able's EasyWay twisted pair line (coming out of the EH128 noted above) to two CASE ARQ2 protocol lines (each running at up to 80,000 kbps). These are the backbone links between CASE nodes, and so this little 3.5" high box looks to the rest of the CASE network like a 128 port node that would consume a good piece of rack space! Add the EH128 above, also 3.5" high, and a VERY short piece of their twisted pair cable, and you have connected 128 Lat ports on your Ethernet to your global Case network using 7" of rack space.  ================================================================================ Note 206.0 LN03 saga 2 replies EISNER::MAYHEW "Bill Mayhew" 67 lines 20-SEP-1988 19:31 -------------------------------------------------------------------------------- It seems my LN03 is a very sick puppy. I'm reporting a recent saga of its ills (and fixes) for the benefit of others, and hoping that someone will come up with an explanation for the current killer problem. A week or so ago, we started noticing that the printer would intermittently "stop printing"... that is, you'd get a solid white band across the middle of a page. The band might be just about any width, indeed sometimes it goes on for pages, itself. Usually the printer "starts printing again", right where it should be; it's not just moving excess paper, it *thinks* it's printing, there's just no image on the page. This was curious but not particularly severe at the time. Now it is VERY severe and it is the outstanding problem that's leading Field Service to replace the whole unit. Around the same time this began, we started using new letterhead, which is printed on 24-lb. stock. The paper was tested for LN03 compatibility before ordering, but no one (mentioning no names in this small company -- oh, all right, it was me!) bothered to try to print a *full* page. It turns out that the bottom couple of lines were "fading", or printing very faintly, when using this paper. However, the published spec says 24-lb. paper is OK. After a bit of prodding, and one ineffective service rep, the local office sent someone out who knew what he was doing and fixed this. It seems that part of his "fix" involved adjusting the overall print image so that it is now higher on the page -- printing a test-status sheet results in something that's nowhere near centered, vertically -- but, well, who cares, right? Now we have not used *much* of that paper in it at all (probably less than 50 pages so far). Over the weekend, as the white-band problem deteriorated, a new phenomenon appeared. The printer lost track of its baud rate. Power down/up and it thought it was at 2400, when the switches were set for 9600. Printing the status page showed the switches in a different position than they were, physically. Monday morning we logged a call, and by the end of the day had a new logic board and a new engine driver board, and the technician had identified the "back door" of the machine (metal plate covering the bottom right rear), which holds a PCB with the switches, as the culprit on the baud-rate issue. Figured out that we could fuss with a cable connector, while powering the machine up, with a good chance that it would "see" 9600 (this had to be done with the machine open, of course). Got a new back door this morning, and that problem is indeed fixed. But the white-band problem lives on. Anyone ever seen this before? Sometimes the band is only half-white -- you can see ghosts of the text that should have printed. Overall, the text on the entire page is a little lighter than it should be. We've done the whole user maintenance kit routine twice, including new OPCs. Of course, Mr. Murphy is hard at work and this all happened while having to print 2 reports of 900+ pages each (plus several others in the 1-200 range -- a very unusually high workload for us, the machine is 2.5 years old and has 49.6K pages on it), *each of which had to be gone through by hand to identify pages that had to be reprinted.* (ugh) Tidbit: according to one of the more knowledgable service reps, DEC is coming out with a 15-ppm 2-drawer printer, otherwise similar to the LN03 (Ricoh engine and all that). One of the other more knowledgable service reps says that something called the "LN04" has been cancelled. Those two rumors may cancel each other out, I dunno. ================================================================================ Note 206.1 LN03 saga 1 of 2 EISNER::PRIGOT "Jonathan M. Prigot" 6 lines 22-SEP-1988 11:10 -< Similar problem(?) >- -------------------------------------------------------------------------------- We once had a similar (?) problem in that text would not reliably print across the width of the page. I think that we saw a light band too, since I remember changing the OPC. The problem turned out to be that toner was not feeding reliably from the drawer. FS had never seen this before, but changing the toner drawer was the fix. ================================================================================ Note 206.2 LN03 saga 2 of 2 EISNER::MAYHEW "Bill Mayhew" 18 lines 22-SEP-1988 17:15 -< Saga continues >- -------------------------------------------------------------------------------- Interesting. We are now on our 3rd LN03 -- the replacement unit DEC brought yesterday dies after about 50 pages with a "call field service, code 1" error in the front panel, which the book says is a fuser error. A technician swapped out all the allegedly related parts with our original unit, and waited around while we printed 100 pages and it was fine -- and of course it then died on page 155, same symptoms. That replacement unit was factory-refurb'ed, with 132,000 copies on the odometer. (Our original had just under 50K.) We now have a loaner which was scrounged up which is running, but which sounds like it's on its last legs, and a 4th unit is on order, hopefully to arrive tomorrow. The tech replaced the high voltage assembly today on the original machine in an effort to solve the original banding problem, but it was not successful. However since I now have 2 machines here, I'll try swapping drawers after we get the pent-up demand behind us, assuming this unit that's squealing in my ear survives the evening.  ================================================================================ Note 207.0 VAXrelated FCO's No replies EISNER::RENES "Stuart Renes, DFWLug Chair" 42 lines 27-SEP-1988 14:50 -------------------------------------------------------------------------------- The following are FCO's released for the months of July and August, 1988: CI750-I006 Another VAXcluster node randomly receives incorrect data in data-packed originating from AVX750/CI750, when a VAX750 single-bit or double-bit main memory (MS750-XX) error is encountered by the CI750. This FCO upgrades the CI750 to Rev. F1 and incorporates ECO: L0009-MK003. A quick check is to look for new part 23-121C6-00 at location E26 on the L0009 module. LPS40-R001 The PrintServer 40 availability to the customer has been severely reduced due a variety of problems affecting rate of jams and print quality. Symptoms include jamming, dirty printer, and print quality degredation. FCO LPS40-R001 will upgrade all field units to rev. level B1. VT330-I001, VT340-I001 Text/Graphics terminals contain firmware version V1.1. However, several problems have been identified as firmware bugs and have been corrected in version V2.1. This FCO calls for the replacement of the VT330/340 ROM cartridge labeled "VT330/340 V1.1" with the new ROM cartridge labeled "VT330/340 V2.1" and 70-22623-04 Rev. C01. 85XBA-I007, 871BA-I003, 882BA-I005 The PC38N console subsystem has a number of problems which can cause systems to crash, and the Console to "Hang" or crash. This FCO changes the console microcode to V8.0 and will be printed on the PC38N screen when the console is booted. This FCO affects 85,87 and 88 series VAX processors. 85XBA-I008, 871BA-I004, 882BA-I006 The current MCL (F1001) can only isolate down to the failing mother board. This fixes the problem of the 64 meg. memory arrays not always being able to isolate a problem down to the failing FRU (daughter board) when the failure is intermittent memory problems. This FCO replaces the F1001 with the F1021 module. This FCO affects 85,87 and 88 series VAX processors.  ================================================================================ Note 208.0 RL01s Dead Forever? 5 replies EISNER::ALLEN_C 9 lines 27-SEP-1988 19:04 -------------------------------------------------------------------------------- RL01s Dead Forever? I know you guys are going to laugh me off the service but, since I have use for this information, I'll ask: Has Digital stopped supporting RL01 drives and, while you're at it, tell me if a person got his distribution for RSTS on tape but only had RL01s, will Digital arrange to transfer the information for him? I have already suggested that my user should get himself another (large) drive but, for the moment, what would you do? ================================================================================ Note 208.1 RL01s Dead Forever? 1 of 5 EISNER::KENNEDY "Terry Kennedy" 3 lines 27-SEP-1988 19:45 -< Not a RSTS supported distribution medium >- -------------------------------------------------------------------------------- RL01 isn't a supported distribution medium for RSTS, but RL02 is. This goes back at least to V8. You could probably do it, but it wouldn't be pretty. Ask in the RSTS_OS conference if you need more info. ================================================================================ Note 208.2 RL01s Dead Forever? 2 of 5 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 15 lines 28-SEP-1988 04:24 -< the BIG ones are cheap >- -------------------------------------------------------------------------------- >>> RL01 isn't a supported distribution medium for RSTS, but RL02 is. This >>> goes back at least to V8. Certainly the RL02 controllers played RL01s, but I don't remember if there was an RL01 controller that won't play an RL02. So I assume the same controller plays RL01, or RL02 drives, and RL02 drives are dirt cheap these days. Probably $200-300 from the rip-off folks and less (if you can find them) elsewhere. The connector is that very junior Massbuss like job that quickly unplugs, so switching drives around is simple, but I think the controller does 4 drives, anyway so you daisy chain them and stick the terminator in the last. Your customer should love you for helping him get double density disk drives... ================================================================================ Note 208.3 RL01s Dead Forever? 3 of 5 EISNER::HAHN "DECUServe Vice Chair" 3 lines 28-SEP-1988 07:02 -< RL01 and/or RL02 >- -------------------------------------------------------------------------------- Yes, RL01 and RL02 can be daisy chained up to four (4) drives per controller. Even the oldest Q-bus 16bit controller takes RL02 drives. We even mix them on same system. ================================================================================ Note 208.4 RL01s Dead Forever? 4 of 5 EISNER::KOZAM 5 lines 28-SEP-1988 21:49 -< RLV11 = 18-bit DMA >- -------------------------------------------------------------------------------- > per controller. Even the oldest Q-bus 16bit controller takes ^^ I thought that the oldest controller was the RLV11 and it definitely is an 18 bit device. ================================================================================ Note 208.5 RL01s Dead Forever? 5 of 5 EISNER::GLEASON "CyberPunk" 11 lines 29-SEP-1988 17:57 -< the extreme case.. >- -------------------------------------------------------------------------------- And...with a patch to the DLDRIVER, you can read and write RL01s on 11/730 using the IDC controller (but we still haven't figured out how to init one there). Lee K. Gleason Control-G Consultants 2416 Branard #D Houston TX 77098 713/528-1859  ================================================================================ Note 209.0 DF242 Firmware Bug 4 replies EISNER::MCCORMICK_B "Bob McCormick" 39 lines 28-SEP-1988 22:36 -------------------------------------------------------------------------------- I reported this problem to Colorado Customer Support on 2-Feb-1988. I got a call last week from my Field Service Unit Manager indicating that some new EPROMS would be coming my way shortly... Connect a DEC DF242-CA using a DEC async modem cable to a DSRVB-BA RS232 modem control DECserver 200. Set the port for 2400 baud. Set the modem to perform speed buffering, autobaud off, and speed to 2400 baud. Users at any speed have no problem dialing into the modem. They can connect to the VAX, and if they are not disabled in the UAF, they can use the system no problem. The port(s) that the DF242's are connected to are also defined as a service (called DF242, neet, eh?). Many users have the need to dial out by using reverse lat technique: they CONNECT DF242, and then issue the DIAL T91413551212 command. Unfortunately the Attached: 2400 message never comes back to their terminal. Connect the DF242 to a TXA device or directly to a terminal and you get the message. Connect a DF112 to the DS200 and you also get the attached message. Seems like there are a number of bugs in the DF242... It took DEC almost four months to acknowledge this: the DECserver 200 keeps its receiver off till some spec specified number of milliseconds after the modem passes CD. So according to spec, the modem needs to wait that amount of time BEFORE blurping out the Attached message. It appears that the current DF242 ROMs wait aprox 4 milliseconds, ie no time at all... I've been told that this problem most likely also exists in the rack mount equiv boards (but they are called DF124-??'s) although I can't confirm it. I've got some on order; I'll post more info here as it develops. Anyone else find/see this problem with DEC modems, or other vendor's modems with attempts to dial out not receiving the modem's status messages? ================================================================================ Note 209.1 DF242 Firmware Bug 1 of 4 EISNER::NORTON "Bill Norton" 6 lines 29-SEP-1988 09:28 -< Ditto >- -------------------------------------------------------------------------------- Our Black Bod MD825A's seem to act the same, you have to *guess* at call progress when dialing out from a DS200. Plus its MNP doesn't seem to be compatible with the DECUServe modems... Is the problem maybe something to do with the Rockwell chipset everyone seems to use these days? ================================================================================ Note 209.2 DF242 Firmware Bug 2 of 4 EISNER::MCCORMICK_B "Bob McCormick" 8 lines 6-NOV-1988 19:57 -< My problems fixed, at last >- -------------------------------------------------------------------------------- Last week my Field Service engineer came out with some updated chips (ROMS) for the DF242. I believe these were V1.5 ... and the problems I identified and logged back on the first of February were fixed... Fixed, but it took MUCH TOO LONG! I hear there are other things that still need fixing, and this is 'in process' ... ================================================================================ Note 209.3 DF242 Firmware Bug 3 of 4 EISNER::MCCORMICK_B "Bob McCormick" 14 lines 14-DEC-1988 20:41 -< DF124-CA also doesn't work with DS200 >- -------------------------------------------------------------------------------- Well, we got the DF100 rack mount box, and installed a number of the new DF124-CA Scholar Plus rack mount boards. These are supposed to be the rackmount versions of the DF242 desktop product. As you can expect, these modems DO NOT display the Attached message when connected to a DECServer 200... same problem as described in the root of this thread... Again, the problem has been reported to DEC, and I'm waiting to see what happens. If like last time, look for another update in about 6 months. BTW, the ROM upgrade for the DF242 bugger doesn't work in the DF124-CA. And the DF124-CA is nothing like the previous DF124 modem! ================================================================================ Note 209.4 DF242 Firmware Bug 4 of 4 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 13 lines 15-DEC-1988 07:24 -< Additional info >- -------------------------------------------------------------------------------- I received updated proms for both the DF224-AA and the DF242-CA. The DF224-AA's (scholar) prom is numbered TD24-J200-061 and fixed the problem where the scholar would not work correctly with the DECServer 200 with alternate speed set. The DF242-CA's (scholar-plus) prom is numbered DF242 V1.5B and does correct the attached message problem when dialing out of a DECServer 200. I can not confirm if the MNP problem is corrected though because I am currently not implementing this feature. A side effect of V1.5B is that you may NOT type ahead of the DF242-CA anymore. For example after pressing B to wake up the modem you could start typing your dial sequence before the ready prompt now you may not.  ================================================================================ Note 210.0 MUXserver 100 & DECmux II No replies EISNER::MCCORMICK_B "Bob McCormick" 10 lines 29-SEP-1988 12:04 -------------------------------------------------------------------------------- MUXserver 100 EQ-01525-01 DECMux II EQ-01524-01 FCO's (or is it ECO's) that the local DEC offices have made to these equipment due to a number of problems, including handling of errors over the synchronous link between the two boxes. Note: These changes (which are some EEPROMS) must be made in the equipment at _both_ ends at the same time. I have also found that the parameters need to be reprogrammed after they have swapped chips.  ================================================================================ Note 211.0 EIA Loopback Connectors? 4 replies EISNER::LEDERMAN "Bart Z. Lederman" 14 lines 30-SEP-1988 20:34 -------------------------------------------------------------------------------- This isn't one of the world's earth-shattering problems, but it's got me intrigued. I have in my desk three EIA loopback connectors, used for testing DEC MUXes (like terminal servers, DHU11s, etc.). They are all female connectors with gray plastic hoods, and have almost identical part numbers molded into them: 12-15336-00 12-15336-02 1215336-08 Are these all functionally equivalent? If not, how do I find out which connector goes with which piece of equipment? ================================================================================ Note 211.1 EIA Loopback Connectors? 1 of 4 EISNER::KOZAM 6 lines 30-SEP-1988 20:42 -< DIfferent Revs for Different Plastic Cases >- -------------------------------------------------------------------------------- I've used the different loopback connectors interchangeably and never had a problem. The revision numbers have more to do with the style of plastic case. Some are squared off, some are triangular, some have a little chain that you use to twirl it around while you try to figure out why the equipment doesn't work... ================================================================================ Note 211.2 EIA Loopback Connectors? 2 of 4 EISNER::DELARISCH "Arnold S. De Larisch" 18 lines 3-OCT-1988 16:34 -< All DB-25 Loopbacks are NOT created equal! >- -------------------------------------------------------------------------------- >> < Note 211.1 by EISNER::KOZAM > >> -< DIfferent Revs for Different Plastic Cases >- >> I've used the different loopback connectors interchangeably >> and never had a problem. The revision numbers have more to do with the >> style of plastic case. Beg to differ ... The REV numbers ARE Significant ... for example the -02 variant is for testing DS200 WITH modem control. There is a special loop-back for IBM PC's which DEC gives you when you buy DECnet DOS (-04 variant). In any case, take a Break-out-box and check them ... THEY ARE a bit different! -Arnold ================================================================================ Note 211.3 EIA Loopback Connectors? 3 of 4 EISNER::RENES "Stuart Renes, DFWLug Chair" 5 lines 4-OCT-1988 15:14 -------------------------------------------------------------------------------- Yeah, but up until now I wondered what I was supposed to do with the ones with chains on the end! Thanks :-)) ================================================================================ Note 211.4 EIA Loopback Connectors? 4 of 4 EISNER::LEDERMAN "Bart Z. Lederman" 2 lines 4-OCT-1988 20:02 -< What I would do: >- -------------------------------------------------------------------------------- I thought it was needed so you could bold down the loopback connector so the FE couldn't walk off with it!  ================================================================================ Note 212.0 TK50 RECALL 8 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 31 lines 7-OCT-1988 11:59 -------------------------------------------------------------------------------- I N T E R O F F I C E M E M O R A N D U M Date: 6-Oct-1988 12:58pm EST From: E. F. Berkley Shands SHANDS Dept: VAX SIG Subject: TK50 recall Our local field service office just informed me about a TK50 cassette recall. The tapes are abrasive, they litterally EAT the drive. A drive using one such tape can die in as little as a day. One of our department machines had a case (24 tapes), used daily. They had lots of new tk50's. I think that we could spread the word (the FS memo was dated JUNE 88) and just now filtered down. The tape id's are as follows 'n' is any digit - both numbers must match Outside center hub Inside center hub 9n74n and nn5n 9n74n and nn0n 9n80n and nn09n or nn09nn 9n80n and nn10n or nn10nn example 91743 2858 Contact the supplier of the tapes, they will replace them. Berk ================================================================================ Note 212.1 TK50 RECALL 1 of 8 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 10 lines 7-OCT-1988 15:31 -< SDC tapes suspect too? >- -------------------------------------------------------------------------------- Is it known if any of these tapes were used by SDC? Or is does this recall only apply to tapes bought as blank tapes. Do we need to look at all recent SDC tapes and any in the near future? We don't buy the damned things, and only recycle obsolete SDC tapes. Fortunately we don't have a TK70, so what SDC ships us is reusable. ================================================================================ Note 212.2 TK50 RECALL 2 of 8 EISNER::MCCORMICK_B "Bob McCormick" 3 lines 7-OCT-1988 17:55 -< Last time... >- -------------------------------------------------------------------------------- When the TK50 first came out there was a recall of tapes with certain numbers and white leaders. I had a few SDC tapes that WERE recall tapes at that time... ================================================================================ Note 212.3 TK50 RECALL 3 of 8 EISNER::RENES "Stuart Renes, DFWLug Chair" 6 lines 9-OCT-1988 16:56 -< I found two SDC ones! >- -------------------------------------------------------------------------------- I just found two such (BAD) tapes in my collection of obsolete SDC release TK50 tapes. Stuart. :-( ================================================================================ Note 212.4 TK50 RECALL 4 of 8 EISNER::TERRAZAS "Mike Terrazas" 10 lines 17-NOV-1988 09:46 -< Update, Please >- -------------------------------------------------------------------------------- Is there any more on this? Our DEC sales rep is giving management here a song and dance about all of these tapes having been succesfully gathered in and replaced. Personally, I find that a little hard to believe, but if there is just one person out there who can document just one tape that fits the description that DEC has not recalled, I would appreciate hearing about it so that I can raise awareness levels here about believing everything vendors may say. Like the old saying: "How can you tell when your DEC salesman is lying to you?" ================================================================================ Note 212.5 TK50 RECALL 5 of 8 EISNER::DELARISCH "Arnold S. De Larisch" 21 lines 17-NOV-1988 17:20 -< Only effected 9274n nad 9280n Tape Batches! >- -------------------------------------------------------------------------------- < Note 212.0 by EISNER::KILLEEN "Jeff Killeen DECUServe Chair" > -< TK50 RECALL >- >> Subject: TK50 recall >> Outside center hub Inside center hub >> 9n74n and nn5n >> 9n74n and nn0n >> 9n80n and nn09n or nn09nn >> 9n80n and nn10n or nn10nn >> >> example 91743 2858 I read the actual Memo that was distributed thru our regional office (Atlanta). The only tapes that were effected were the 9274n and 9280n for the CENTER HUB marking with the above INSIDE CENTER HUB as indicated above. Also, there was a note that no SDC (Software Distribution Center) tapes were suspect. -Arnold ================================================================================ Note 212.6 TK50 RECALL 6 of 8 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 9 lines 21-NOV-1988 09:57 -< Some SDC tapes affected >- -------------------------------------------------------------------------------- RE: < Note 212.5 by EISNER::DELARISCH "Arnold S. De Larisch" > -< Only effected 9274n nad 9280n Tape Batches! >- | Also, there was a note that | no SDC (Software Distribution Center) tapes were suspect. Interesting. We just had three SDC (distribution kits) replaced happily by DEC because their serial numbers were included in the list. ================================================================================ Note 212.7 TK50 RECALL 7 of 8 EISNER::DELARISCH "Arnold S. De Larisch" 18 lines 21-NOV-1988 12:44 -< SDC tapes 9274n & 9280n or 9n74n & 9n80n Batches? >- -------------------------------------------------------------------------------- >> < Note 212.6 by EISNER::TILLMAN "Brian Tillman, Smiths Industries." > >> -< Some SDC tapes affected >- >> | Also, there was a note that >> | no SDC (Software Distribution Center) tapes were suspect. >> >> Interesting. We just had three SDC (distribution kits) replaced >> happily by DEC because their serial numbers were included in the >> list. I too have serveral SDC tapes in the effected batch numbers. They, DEC, are unwilling to replace them since the memorandium from the Atlanta Center explicitly EXCLUDED SDC tapes. I'm considering pressing the issue. A question Brian, Did you tapes fall into the 9274n and 9280n or 9n74n and 9n80n category? -Arnold ================================================================================ Note 212.8 TK50 RECALL 8 of 8 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 1 line 21-NOV-1988 16:15 -< Sorry, Arnold, I don't remember >- --------------------------------------------------------------------------------  ================================================================================ Note 213.0 What's YOUR favorite tape? 25 replies EISNER::ZIMMERMAN 4 lines 7-OCT-1988 16:33 -------------------------------------------------------------------------------- I would like to know everyone's opinion on what the best 9 track magnetic tapes are. I need to make a sizeable purchase and would like other's input. thanks ================================================================================ Note 213.1 What's YOUR favorite tape? 1 of 25 EISNER::COY "Dale E. Coy" 2 lines 7-OCT-1988 22:07 -< This is a serious question - no comedy intended >- -------------------------------------------------------------------------------- Best for what? What are the most important qualities you would like to get in a tape? You can't have them all, so pick carefully. ================================================================================ Note 213.2 What's YOUR favorite tape? 2 of 25 EISNER::ANGLEY "Kevin Angley" 10 lines 9-OCT-1988 14:10 -< Black Watch! (Now come the nay-sayers) >- -------------------------------------------------------------------------------- Before you get two hundred replies that say "anything BUT Black Watch", I will say that I have used 2400 and 3600 foot Black Watch tapes successfully for years. I like them. No particular reason, just that I've never experienced any data loss and they don't get mangled as much as other tapes I have tried. I guess my company (Memorex Telex) would like me to point out that, were I to be buying tapes today - post-merger (which we are not since we switched to 8mm), I would probably choose a Memorex brand. ================================================================================ Note 213.3 What's YOUR favorite tape? 3 of 25 EISNER::RENES "Stuart Renes, DFWLug Chair" 8 lines 9-OCT-1988 17:02 -< GRAHAM Epoch 480 >- -------------------------------------------------------------------------------- We use GRAHM Epoch 480 tape here. Haven't had any problems with it on 1600 to 6250 drives (TU78 and TU81+). I believe this tape is one of only two approved for use by NASA at the Cape (I don't remember the other approved one). Stuart. ================================================================================ Note 213.4 What's YOUR favorite tape? 4 of 25 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 66 lines 9-OCT-1988 20:32 -< NEVER use BACKCOATED tape - here is why... >- -------------------------------------------------------------------------------- >>>> < Note 213.2 by EISNER::ANGLEY "Kevin Angley" > >>>> -< Black Watch! (Now come the nay-sayers) >- >>>> Before you get two hundred replies that say "anything BUT Black >>>> Watch", NAY-SAYER #1: DON'T use it. AND there is a VALID reason. Aside from the extra crud you may need to clean off, or any extra head surface abraded off, there is a simple reason to not use any backcoated tapes. If you look at a CDC Keystone drive (9218x family, also marketed by some other company for 3 times the price as TU8x drives) you will notice that the tape ONLY touches the head, ONLY. It does NOT touch those corner posts! It floats around them on an air film, and the back pressure on that air supply is related to tape tension. There in NO capstan. Tape motion past the head is done by directly controling the relative motion of the supply and the takeup reels while watching tape tension, and doing all that while the circumference of each pack of tape on its reel keeps changing. NOW THE BUMMER. As you crank on one reel and hold back some on the other you expect, and in fact MUST HAVE, the desired tape velocity across the head. BUT the back-coated tapes SLIP MORE layer to layer and you can haul more than you expected (and at higher velocity) off the supply reel, OR move LESS than expected onto the takeup reel as the darn tape slips and winds itself yet tighter around the hub. Then remember those zillion and one reversals needed by BACkup, even with /BUF=5, failing to keep the streamer streaming and forcing repositioning. More chances for more slipping. You say you have a vacuum column drive! And you have NO friends with streamers, not even a casual LUG tape swap possible, ever! Actually it is a clever move, NOBODY who knows better will want to 'borrow' tapes. I have heard they work very well on Bouroughs (spelling?) drives. They probably also work well on the ONLY serious drive DEC ever sold, a TU72. But it is a shame to insult such a drive with this tape. (If you still have a MASSBUSS VAX and someone is scrapping an old 10/20 nearby, GRAP a quality TU72. There is a VMS driver for them, written not to give you access to a quality drive, but to woo the 10/20 users into migrating to VMS while keeping old peripherals.) The TU72 is an STC drive. Yes, I have some backcoated tapes, and I have a VACUUM column tri-density drive. (No, not DEC's only tri-density which was a 200, 556, 800 bpi TU10.) I certainly did NOT buy the tapes. But I have a very special use for them. There are some people who always ask for the latest SIG tape, and promise to return an equal tape, but NEVER return anything. Slowly my backcoated tape pile shrinks. ================================================================================ Note 213.5 What's YOUR favorite tape? 5 of 25 EISNER::CONROY "Alan Conroy" 14 lines 10-OCT-1988 11:12 -< Blackwatch is good, except... >- -------------------------------------------------------------------------------- > Before you get two hundred replies that say "anything BUT Black > Watch", I will say that I have used 2400 and 3600 foot Black Watch > tapes successfully for years. I like them. No particular reason, I also used Blackwatch for several years and had no problems. They seemed to be a good, solid tape. However, I have run into certain drives (one made by Cipher) that cannot sense tape motion on these tapes because of the flat black coating. The drive (and others I hear) needs a reflective surface on the tape or it assumes the tape isn't moving and then - drive error! It's unfortunate that the industry and 3M aren't together on this because I never had a blackwatch tape go bad on me. Sure, they are a little more "dirty", but that was never a consideration one way or the other for me. However, because of these incompatibilities, I no longer use the blackwatch. ================================================================================ Note 213.6 What's YOUR favorite tape? 6 of 25 EISNER::DELARISCH "Arnold S. De Larisch" 17 lines 10-OCT-1988 11:45 -< 3M Blackwatch is #1 in my Shop! >- -------------------------------------------------------------------------------- I too am a Blackwatch advocate. I've YET to have one go bad on me! Further, I purchased 50 DEC brand 2400 foot reels. The DEC tape S U C K S 14 of the suckers went bad in the period of about a year! My field service rep complains anytime I hand him a Blackwatch tape to test the drive. I GLADLY hand him DEC tape for testing (I love to watch him eat his words). I have used BLACKWATCH 3600', 2400' and 600' reels on TE16s and TSV05s (O.K. Stop Laughing!) ... without a single problem! If you have to store anything for any significant lenght of time (greater than 1 year) I recommend 3M Blackwatch (assuming your tape drive can handle it!). -Arnold O.K. ... start throwing the rotton Fruit and Vegies! 8-{) ================================================================================ Note 213.7 What's YOUR favorite tape? 7 of 25 EISNER::MERRIMAN "George Merriman - CCA/NY" 2 lines 10-OCT-1988 14:19 -------------------------------------------------------------------------------- Topic 182 in this conference also discusses various mag tape brands. ================================================================================ Note 213.8 What's YOUR favorite tape? 8 of 25 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 23 lines 10-OCT-1988 16:51 -< Memorex and Graham >- -------------------------------------------------------------------------------- We have a multivendor installation here, with Prime, Honeywell, HP, and VAX equipment; using CDC, Kennedy, Pertec, and Unknown Vendor tape drives (and we interchange with an IBM site running the Drive of the Month). We use 800, 1600, 3200, and 6250 BPI densities. There are only two vendors from whom we now purchase tapes -- Memorx Cubic HD or Quantum II, and Graham Ultimag. We have found that these specific tapes (and Memorex MRX V) are apparently burnished or polished on their oxide sides before shipment, since they will show zero errors on BACKUP when new; most others won't until they have been run two to five times through the tape drive, shedding minute amounts of oxide each time. We have a few hundred MRX V tapes but no longer order them since they do not seem to hold up as long before starting to accumulate soft errors. We use no 3M tapes and have cycled most of our existing ones out, since the binder will soften when held in stationary contact with the heads on three of our drives for more than about a minute (high write current). We supply all of these drives from a common pool of tapes, and find it impractical to assign specific tapes to specific drives. All of the tapes we now buy have shown extremely low rates of data loss when stored in controlled environments (not under the telephone, for example) ================================================================================ Note 213.9 What's YOUR favorite tape? 9 of 25 EISNER::OSUDAR "John Osudar" 5 lines 11-OCT-1988 19:16 -< another Memorex vote >- -------------------------------------------------------------------------------- Memorex Cubic HD's have worked best for us. Our procurement people have switched several times over the past few years, and each time we find problems with their new choice and get them to go back to stocking the Cubic HD's. (Of course, we haven't bought any new tapes lately, as we've been recycling old ones.) ================================================================================ Note 213.10 What's YOUR favorite tape? 10 of 25 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 3 lines 12-OCT-1988 00:27 -< A Black Watch vote >- -------------------------------------------------------------------------------- We've been using Black Watch for a while now... Some warnings from DEC service, but we haven't seen any problems at all (TU81+) on several systems. ================================================================================ Note 213.11 What's YOUR favorite tape? 11 of 25 EISNER::HASSINGER "Bob Hassinger" 9 lines 12-OCT-1988 12:33 -< Only thing wrong with Black Watch seems to be DEC Field Service >- -------------------------------------------------------------------------------- We have been using Black Watch tape here for a long time because our corporate EDP center buys it in large quantity so we get a good price. My only problem with it has been the fact the DEC Field Service people seem to blame everything under the sun on it BUT, they have NEVER been prepared to put ANYTHING in writing - which we would need to address their complaints to EDP, the vendor, etc. I think we are not alone in this experience. Bob H ================================================================================ Note 213.12 What's YOUR favorite tape? 12 of 25 EISNER::LEDERMAN "Bart Z. Lederman" 28 lines 12-OCT-1988 13:28 -< It's not just the tape. >- -------------------------------------------------------------------------------- We have used Graham Epoch, BASF (various), 3M (various including Black Watch), DEC, and I don't know what else. Probably, anything which was low priced at the time the P.O. made it into our purchasing department. All have been used on DEC TU10s at 800 BPI (which everyone else swears at), Kennedy 9000, and other drives, with little or no problems on any brand. *************** BUT ******************* We have our own tape cleaners and certifiers (commercial units anyone can buy: the point is we have our own on-site). ALL tapes carrying important information go through the cleaner and certifier before each data cycle. If there are excessive bad spots the tape is taken out of use: or, if the spots are near the beginning or end, as they often are, the bad portion is cut off and a new leader or trailer formed, which also eliminates problems with tapes running off the end of the reel with large block sizes as we ALWAYS have a long enough trailer. In addition, our own staff maintains the tape drives, checking alignment, cleaning the drives, etc. We don't depend on the mystery guest FE of the week. All this is an expense, but a small expense compared with the cost of losing the data on the tapes. (By the way, the systems also write all critical data to two tapes simultaneously on two drives: we take no chances.) ================================================================================ Note 213.13 What's YOUR favorite tape? 13 of 25 EISNER::BURNET "John Burnet" 16 lines 14-OCT-1988 11:33 -< Info on tape-cleaning services solicited >- -------------------------------------------------------------------------------- Re. < Note 213.12 by EISNER::LEDERMAN "Bart Z. Lederman" > > We have our own tape cleaners and certifiers (commercial units anyone > can buy ... Can you tell us where we can acquire such devices, how much they cost, how they work, etc.? Alternatively, does anyone know of any good commercial tape-cleaning services available separately from off-site-storage services? (We used to use DEC for off-site storage of our backup tapes, but we don't any more.) We have lots of tapes that we've taken out of use since they need to be cleaned, and we've been putting off finding a service bureau to do it. I don't know if the expense of a tape-cleaning device would be justified at our site, since the number of tapes we have is in the hundreds rather than the thousands or tens of thousands. Thanks... ================================================================================ Note 213.14 What's YOUR favorite tape? 14 of 25 EISNER::CONROY "Alan Conroy" 8 lines 14-OCT-1988 11:58 -< Tape Cleaning Equipment >- -------------------------------------------------------------------------------- Most vendors of magnetic tape media sell cleaners/evaluators. Call a big time tape manufacturer and ask for the name of their distributor in your area. Call this company and you will probably find that they sell these machines. Thus far, I have found their costs to be prohibitive ($6000 average, sometimes a good deal can be found for $2000). I have either rented a machine (the people that sell them sometimes rent them too), or used some outside service. Of course, some people go through a LOT more tapes than we do and the cost of one of the babies is justified to them. ================================================================================ Note 213.15 What's YOUR favorite tape? 15 of 25 EISNER::LEDERMAN "Bart Z. Lederman" 35 lines 14-OCT-1988 15:20 -< What we use. >- -------------------------------------------------------------------------------- I agree that owning your own tape cleaner and certifier may only be for people who: a), own a lot of tapes, or b), are very interested in tape data integrity. If you happen to be in a company with a separate MIS department using mainframes and your (or other) departments using DEC, then the two of you together may have enough tapes to justify a cleaner. (I sometimes think that 10.5" open reel magnetic tape is the only industry standard that actually means anything: you really can put, at least physically, any tape on any drive.) The certifiers we have now are by Computer-Link: the sticker on the back says they're in Burlington MA and the model is 2400. I don't know much else about them, except that you load the tape on the front, thread it through, push some buttons to select the data density and some other options, and push start. It does the rest, and will even print out the error log on an internal adding-machine size paper tape strip. You do have to change the cleaning blade and wiper cloth occasionally. In addition to physically cleaning the tape, it actually writes 9 (or 7) track data on the tape at a given density (800, 1600, 6250) and reads it back. You can adjust the read threshold to make the certification more critical, if you choose. We used to have units made by KYBE, but got rid of them. I don't know if there was any reason other than that they were old and worn out: they may have been a little less automatic than the units we have now, and they were larger physically. I don't know if anyone makes a unit which cleans only. If they did, it would be less expensive. You could sort of certify tapes on the tape drive: the old User Mode Diagnostics on RSX-11M had a mode where it would just write the entire tape and read it back, noting any errors. Better than nothing, I suppose, and it will get rid of any old data at the end of the tape if you don't have a bulk eraser. ================================================================================ Note 213.16 What's YOUR favorite tape? 16 of 25 EISNER::PRIGOT "Jonathan M. Prigot" 4 lines 17-OCT-1988 09:25 -< Scotch 703 >- -------------------------------------------------------------------------------- We use Scotch 703 Blackwatch tape here. Before you $set flame/intensity=thermonuclear, this is a non-backcoated variant of their tape. We have had no problems at 1600 or 6250 on TU77s and a TA81. ================================================================================ Note 213.17 What's YOUR favorite tape? 17 of 25 EISNER::MCBRIEN "Jeff McBrien" 17 lines 28-OCT-1988 15:29 -< 3M Blackwatch! >- -------------------------------------------------------------------------------- My vote is for Blackwatch. I've seen 3M take quite a beating on this, now and in the past, and I would like to straighten a few things out. According to our labs, 3M Magnetic Media Division, the residue that comes off the back of the tape is normal with ALL tapes. It just so happens that ours is black and therefore you can see it. ALL other tapes shed this residue as well the only problem is that you can't see it. Personally, I would like to see what is clogging up my drive. As for the fact that Blackwatch acts differently on the drive hub, I can't comment one way or another on that. We have used 3M Blackwatch tapes for years (obviously) without an incident. Jeff McBrien - 3M ================================================================================ Note 213.18 What's YOUR favorite tape? 18 of 25 EISNER::KENNEDY "Terry Kennedy" 14 lines 29-OCT-1988 03:01 -< Normal tape can't shed! >- -------------------------------------------------------------------------------- > According to our labs, 3M Magnetic Media Division, the residue that > comes off the back of the tape is normal with ALL tapes. It just > so happens that ours is black and therefore you can see it. But... 'Regular' tape is *not* back-coated. All you have is the Mylar (plastic) tape base, with some binder and oxide on the data side of the tape. If your non-back-coated tape is shedding Mylar residue, your tape is dying. Otherwise, there should be no shedding on a non-back-coated tape. Lastly, DEC specifically states that *no* back-coated tape is currently accepted by them on the TU8x drive series. They don't seem to have any stand (at least specifically stated) with the other drive models. ================================================================================ Note 213.19 What's YOUR favorite tape? 19 of 25 EISNER::HASSINGER "Bob Hassinger" 4 lines 31-OCT-1988 11:14 -< If it was in writing it was a first! Normally ONLY verbal.... >- -------------------------------------------------------------------------------- > Lastly, DEC specifically states that *no* back-coated tape is currently >accepted by them on the TU8x drive series. They don't seem to have any stand >(at least specifically stated) with the other drive models. ================================================================================ Note 213.20 What's YOUR favorite tape? 20 of 25 EISNER::KENNEDY "Terry Kennedy" 7 lines 1-NOV-1988 00:56 -< 'nuff said >- -------------------------------------------------------------------------------- > If it was in writing it was a first! Normally ONLY verbal... TU80 Subsystem User's guide, p/n EK-0TU80-UG-002, October 1983: Page 1-2: "... To date, there are no back-coated magnetic tapes that meet Digital's specification. Back-coating has been found to cause such prob- lems as tape slippage, auto-load failures, false EOT..." ================================================================================ Note 213.21 What's YOUR favorite tape? 21 of 25 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 12 lines 1-NOV-1988 18:16 -< DEC approves of someone >- -------------------------------------------------------------------------------- Comment from a battle-scarred system manager: Ask DEC what tapes _DO_ meet Digital's specifications, that are _NOT_ sold by DEC. We've got tons of Black-watch tapes being used on TU81+ drives with no problems to speak of. OK, OK... so one tape ticked me off and I torched it. Nobody would have known, if the smoke hadn't tripped the fire system and dumped the halon. Besides, whose idea was it to rig the halon system into the building sprinkler system. And was it my fault the water poured into the company three floors below and soaked their boxes of computer paper, making them sorta explode. Sigh. ================================================================================ Note 213.22 What's YOUR favorite tape? 22 of 25 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 0 lines 8-DEC-1988 01:05 -< See 242.0 for related comments >- -------------------------------------------------------------------------------- ================================================================================ Note 213.23 What's YOUR favorite tape? 23 of 25 EISNER::KENNEDY "Terry Kennedy" 8 lines 29-JUN-1989 18:42 -< 3600' tapes >- -------------------------------------------------------------------------------- Someone asked (in 388.5) about 3M tapes because they come in 3600' lengths. Without mentioning *our* brand preference, I know of at least 2 other vendors for 3600' tapes: Memorex Seismic 3600 and CDC Total Service 3600. I'd also mention that you may have a problem with 3600' tapes; the software timeout may be to short to allow a full rewind from the end without an error. ================================================================================ Note 213.24 What's YOUR favorite tape? 24 of 25 EISNER::CONROY "Alan Conroy" 12 lines 1-JUL-1989 04:29 -< 3600' tapes >- -------------------------------------------------------------------------------- > Someone asked (in 388.5) about 3M tapes because they come in 3600' >lengths. Without mentioning *our* brand preference, I know of at I believe they also come in 2400' lengths. > I'd also mention that you may have a problem with 3600' tapes; the >software timeout may be to short to allow a full rewind from the end >without an error. I haven't used 3600' tapes for years! They appear to be made from a thinner material that is less resilient to stretching and other mangling (at least the brands I remember). I'd recommend against it. ================================================================================ Note 213.25 What's YOUR favorite tape? 25 of 25 EISNER::MCGLINCHEY "DECUS Board of Directors" 5 lines 25-OCT-1989 17:36 -< Nobody did this yet!! >- -------------------------------------------------------------------------------- Well, _MY_ favorite tape is "The Best of The Statler Brothers" -- Glinch  ================================================================================ Note 214.0 DFA01 help/suggestions needed No replies EISNER::MCCORMICK_B "Bob McCormick" 18 lines 7-OCT-1988 17:59 -------------------------------------------------------------------------------- We are configuring DFA01's on our MicroVAX 3XXX series machines. We use DF242-CA on other configurations, attached to a modem control port on a mux (ie DHQ11 in MicroVAX II, DMB32 on BI machine) The DF242-CA has speedbuffering - Nice! Set the speed between the CPU and the modem and forget it. You can dial in at any speed. Dialing out (a la SET HOST/DTE or Kermit) you can command the modem to dial your favourite number. The DFA01's don't have the speed buffering capability. We've tried a number of different settings... but are not totally happy with each situation (we don't want to consider purchase of DEC's software package for the device). Anyone have any experience or comments regarding the DFA01? We would like to honour the following: - Dial in at 1200 or 2400 bps - Dial out (without too much hassle) at 1200 and/or 2400 bps  ================================================================================ Note 215.0 How can you salvage data from a crashed RA-81? 2 replies EISNER::GODFREY "Gary Godfrey" 7 lines 10-OCT-1988 14:04 -------------------------------------------------------------------------------- We have had a head crash on an RA-81 disk drive. The drive does not even spin up. I know there must be some companies out there who can salvage whatever data that may be left on the disk. Does anyone know of such a company? Has anyone had any experience with such companies? Any other helpful information will be greatly appreciated. ================================================================================ Note 215.1 How can you salvage data from a crashed RA-81? 1 of 2 EISNER::MAYHEW "Bill Mayhew" 6 lines 10-OCT-1988 15:14 -< Lead #1 >- -------------------------------------------------------------------------------- Indeed I do know of someone who does this. Try calling Terry Compton at COMP3, Inc. 614-876-3682, and tell him I referred you. I have no connection with this company except that Terry is one of those electronic friends I've never actually met! I don't know how much he can do for you since there was a real head crash, but it's worth a call. ================================================================================ Note 215.2 How can you salvage data from a crashed RA-81? 2 of 2 EISNER::CONROY "Alan Conroy" 10 lines 11-OCT-1988 12:18 -< Sorry, but no good news >- -------------------------------------------------------------------------------- If your head REALLY crashed, then at least some of your data will never be recovered. Hindsight is 20/20 as they say, but here is a good reason for regular and frequent backups... Anyway, a really crashed disk will not be recovereable by you, regardless of the amount of bit-twiddling you may want to do. You'll have to pay some- one else to do it for you (if there is anyone willing to take on an RA81 HDA). I know that certain trade magazines have had advertisements for these services, but I don't recall which one. Basically, someone will have to take the platters out of the HDA, and somehow read what is left of the disk (NOT an easy task!) Also be prepared to pay a LOT.  ================================================================================ Note 216.0 NonDEC PostScript Printers 15 replies EISNER::TINKELMAN "Bob Tinkelman - CCA" 10 lines 11-OCT-1988 14:13 -------------------------------------------------------------------------------- We are planning to buy a Postscript printer (not necessarily the LN03R) and to use DEC's "ScriptPrinter software" (whatever its official name is). When we looked into this once before, we verified that DEC would sell the software unbundled from the LN03R. I'm concerned about compatibility issues and am anxious to hear from anyone who has attempted to run nonDEC Postscript printers using DEC's translators and symbiont. Most interesting would be to hear from anyone using the new Apple Laserwriters (IINT and IINTX) or the 12 ppm Dataproducts LZR 1260. ================================================================================ Note 216.1 NonDEC PostScript Printers 1 of 15 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 21 lines 11-OCT-1988 20:43 -< Same DEC, different story. >- -------------------------------------------------------------------------------- Interesting. I was told by a DEC type from the group that did the ScriptPrinter symbiont at the DECUS before last that it went out of its way to check that it was really talking to a LN03R and would choke if it was someone elses printer. They also claimed that it came only with the LN03R and wasnt available unbundled. Good ol' Digital again I guess. Is there a particular reason that you are enamored of the DEC symbiont? I have about 40 PostScript printers on my cluster and Im using the Pacer Software PacerPrint symbiont. It does several conversions on the fly (I think the same ones as DEC's but Im not sure), is multithreaded, works with different vendors printers, is quite reliable, and is reasonably priced. In terms of printers, we use QMS PS800s, PS810s, and PS800IIs. I did a comparison of the QMS printers to the LN03R, the old LaserWriter, and a couple others. The QMS printers were faster (particularly compared to the LaserWriter), cheaper, and had about the best print quality. They also have about the widest family of PostScript printers available, including a color one. ================================================================================ Note 216.2 NonDEC PostScript Printers 2 of 15 EISNER::TINKELMAN "Bob Tinkelman - CCA" 10 lines 11-OCT-1988 22:29 -< "Unbundled" .nes. "Cheap" >- -------------------------------------------------------------------------------- RE: < Note 216.1 by EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" > > Interesting. I was told by a DEC type from the group that did the > ScriptPrinter symbiont at the DECUS before last that ... > came only with the LN03R and wasnt available unbundled. From the DEC Electronic Store: QLZFQ-GZ ScriptPrinter V1.1 Doc Set $95 QL-VF9A9-AA ScriptPrinter V1.1 License $1122 ================================================================================ Note 216.3 NonDEC PostScript Printers 3 of 15 EISNER::HASSINGER "Bob Hassinger" 1 line 12-OCT-1988 12:26 -< There should be an "H-kit" too. >- -------------------------------------------------------------------------------- ================================================================================ Note 216.4 NonDEC PostScript Printers 4 of 15 EISNER::HARVEY "Jack Harvey" 11 lines 13-OCT-1988 08:03 -< Has anybody bought it? >- -------------------------------------------------------------------------------- >>>From the DEC Electronic Store: >>>QLZFQ-GZ ScriptPrinter V1.1 Doc Set $95 >>>QL-VF9A9-AA ScriptPrinter V1.1 License $1122 From July U.S. Price list: QL-VF9AA-AA ScriptPrinter License/W $1,155 QA-VF9AA-H5 ScriptPrinter TK50 $525 QA-VF9AA-HM ScriptPrinter 16MT $525 ================================================================================ Note 216.5 NonDEC PostScript Printers 5 of 15 EISNER::BURNET "John Burnet" 13 lines 24-APR-1989 20:07 -< What Postscript printers are good? Qume Scripten? TI 2115? >- -------------------------------------------------------------------------------- Does anyone have any experience with the Qume "Scripten" Postscript printer? We are considering buying one as part of an electronic-publishing system (VAXstation-based). Qume doesn't have as big a presence in the laser-printer market as do HP, DEC, Apple, and others, and we know virtually nothing about this printer. What print engine does it use? Is the HP Laserjet emulation accurate? Any gotchas we should know about? Etc. One alternative is the TI Omnilaser 2115, although I've heard bad things about the 2115 (here and on the Pageswapper system). How does the 2115 stack up against the Qume or other Postscript printers that can print faster than 8 ppm? I guess what I'm really asking is more general: What Postscript printer would *you* recommend for use in a VMS environment, and why? ================================================================================ Note 216.6 NonDEC PostScript Printers 6 of 15 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 17 lines 24-APR-1989 22:21 -< QMS >- -------------------------------------------------------------------------------- > What Postscript printer would *you* recommend for use in a VMS > environment, and why? We have a bunch (~45) of QMS PS800 and PS810 PostScript printers. These correspond to the Laserwriter I and II. They use the Cannon print engine and have the 13 or 35 fonts. At the time we made the decision to use the QMS printers, they outperformed the the Laserwriter (printing "typical" stuff with timedpages.ps loaded) and we were able to get a better price on them. We have been quite happy with them. The print quality is good, they have been very reliable and they have worked with all of the applications packages we have tried. QMS also has a quite broad line of PostScript printers, including a dual bin Cannon based one (PS820), a 20 page per minute one (PS2200) and a color printer (ColorScript 100). We have one of each of these either installed or on order. QMS also sells a downloadable DECtech font. (We use it with MEC's MASS-11). ================================================================================ Note 216.7 NonDEC PostScript Printers 7 of 15 EISNER::BURNET "John Burnet" 9 lines 27-APR-1989 12:20 -< TI Omnilaser is it >- -------------------------------------------------------------------------------- Thanks for the info about QMS. We had considered their printers, and they do have some good products. We're kind of partial to the Ricoh write-white engines, though, after having used DEC and Talaris printers. It looks like we'll be going with the TI Omnilaser 2115. It seems to have everything we need, service in this area is no problem (Qume service would be a problem), and as an added bonus we'll be able to use the same supplies that we use with our Talaris 1590s. The TI is more expensive than the Qume, but it's also faster. ================================================================================ Note 216.8 NonDEC PostScript Printers 8 of 15 EISNER::RICE "Gary Rice" 13 lines 28-APR-1989 08:37 -------------------------------------------------------------------------------- > -< TI Omnilaser is it >- One word of caution with the TI unit: The installation guide states that "If you are not going to use the printer for more than 4 hours, turn it off." We failed to heed that caution. 6 months and about 2000 copies later, the print engine died and had to be replaced (and US with no maintenance contract). Be sure to turn the printer off ... Gary ================================================================================ Note 216.9 NonDEC PostScript Printers 9 of 15 EISNER::BURNET "John Burnet" 9 lines 28-APR-1989 18:24 -< Premature Omnilaser death -- ?!? >- -------------------------------------------------------------------------------- > Be sure to turn the printer off ... Thanks for the warning! I'm baffled, though, as to why the engine should die after being left on for a long time. The engine is made by Ricoh (forgot the model number), and it's also used in the Talaris 1590. We have had two 1590s running 24 hours/day for almost a year now with no problems. Do you know just why this is happening? Does the TI electronics keep the imaging belt running continuously, or what? ================================================================================ Note 216.10 NonDEC PostScript Printers 10 of 15 EISNER::RICE "Gary Rice" 10 lines 29-APR-1989 09:56 -------------------------------------------------------------------------------- > Do you know just why this is happening? Does the TI electronics keep the > imaging belt running continuously, or what? Sorry to say that I wasn't there when the service guy showed up. All I have is words from a user (who watched the repair process - an exchange) that re-iterated the warning in the manual. Gary ================================================================================ Note 216.11 NonDEC PostScript Printers 11 of 15 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 2 lines 30-APR-1989 11:32 -< Some people get lemons, others, lemonade >- -------------------------------------------------------------------------------- Well, our OMNI has been running 24 hours per day for the better part of a year. No problem. ================================================================================ Note 216.12 NonDEC PostScript Printers 12 of 15 EISNER::OSUDAR "John Osudar" 13 lines 30-AUG-1989 15:41 -< DEC CPS$SMB with non-DEC Postscript printers? >- -------------------------------------------------------------------------------- I'd like to revive this topic in its original form, i.e. running DEC's ScriptPrinter software to drive non-LN03R Postscript printers. (We have several LN03R's, but have recently acquired a couple of other Postscript printers. We'd like to have similar features available on all of them.) If I try to start up a queue on a non-DEC Postscript printer (specifically, a Dataproducts LZR1260), the symbiont aborts with a CONTERMINATED status (connection terminated abnormally). I assume that it's trying to verify in some way that the printer is an LN03R -- but maybe it's just trying to establish two-way communications and is failing to do so. Has anyone out there tried doing this? Has anyone succeeded? Or does anyone know for certain whether or not it can be done? ================================================================================ Note 216.13 NonDEC PostScript Printers 13 of 15 EISNER::OSUDAR "John Osudar" 37 lines 26-SEP-1989 19:59 -< non-DEC Postscript printers >- -------------------------------------------------------------------------------- Here I am, answering my own questions again... Yes, it *is* possible (at least with V2.0 of the software) to use DEC's Postscript symbiont to drive a non-DEC real Postscript (i.e. from Adobe vs. "PS-compatible") printer. To do this, you have to overcome two problems: One is that the "password" on an LN03R is a character string, while on all other PS printers I've seen it's an integer. It takes a patch to the symbiont to get around this. (CPS$SMB uses SYS$FAO to format stuff; it formats the password string with "(!AS)". Changing this string to the numeric password, e.g. " 0 ", is what's required. Note that there's more than one "(!AS)" in CPS$SMB; the one at or around location 000008DE seems to be the correct one. Also note that a patched CPS$SMB *cannot* be used with an LN03R -- so call it something else. And if you patch CPS$SMB, as with patching any other DEC software, the consequences are all yours... :-) The other is the fact that LN03R's have printer-specific stuff that's different than for other PS printers. The main difference I saw (in terms of what CPS$SMB requires to think it's talking to an LN03R) was "papersize" (defined in statusdict), which returns the paper tray size. A definition like this: statusdict begin /papersize { (lettertray) true } def end when done outside the server environment (i.e. after an "exitserver") will do the trick. (This would be for a Dataproducts LZR1260; other printers may have other quirks.) My original problem (CONTERMINATED) appears to stem from having the printer on a LAT port (and on an Emulex P4000 terminal server); the printer works fine on a DMB32 port. Of course, since V2.1 of the software is now out, there's no telling how much of this will still be true (or necessary)... ================================================================================ Note 216.14 NonDEC PostScript Printers 14 of 15 EISNER::CHILD "John Child" 7 lines 28-SEP-1989 15:54 -< Using Scriptprinter symbiont w/ a Laserwriter >- -------------------------------------------------------------------------------- This is a repeat of a reply in the DESKTOP_PUBLISHING conference. I happen to have a copy of the Scriptprinter software, and rather than go out and buy another symbiont, I'd like to try to re-use the "free" symbiont I've got to run Apple Laserwriters off the network. Has anyone done this? Is it just a question of modifying DEC's device control library, or must one patch the symbiont, as John suggests above? ================================================================================ Note 216.15 NonDEC PostScript Printers 15 of 15 EISNER::OSUDAR "John Osudar" 15 lines 28-SEP-1989 19:55 -< yes, symbiont mod required >- -------------------------------------------------------------------------------- The "serverdict begin exitserver" string is generated by the symbiont, using $FAO -- it is *NOT* built into some module in the device control library. (I looked!) So you *do* have to modify the symbiont. You also have to deal with the "quirks" of the particular printer. For example -- the AST Turbolaser/PS-Plus3 requires that the "error reporting to host for Postscript" be enabled each time the printer power is turned on (and that has to be done with an "executive mode" command -- all this stuff is AST-specific.) If anyone else has figured out a better (or even different) way to accomplish the same thing, I'd like to hear about it -- but when I asked the question, I got no answers (so I went and did it myself. :-)  ================================================================================ Note 217.0 Modems and DECServers 13 replies EISNER::CONROY "Alan Conroy" 5 lines 14-OCT-1988 12:07 -------------------------------------------------------------------------------- I am having a problem with connecting a Hayes 2400 modem to our DECServer. The problem seems to lie in how the server and modem handle DTR. The DECServer manual and the modem manual talk about on-to-off transitions of DTR. My question for y'all is what voltage (relative to signal ground - pin 7) is a "On" state for DTR, and what is "Off"? ================================================================================ Note 217.1 Modems and DECServers 1 of 13 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 16 lines 14-OCT-1988 13:38 -< RS-232 voltages >- -------------------------------------------------------------------------------- >> My question for y'all is what voltage (relative to signal >> ground - pin 7) is a "On" state for DTR, and what is "Off"? Positive voltage = Binary zero = Signal "space" = Control On Negative voltage = Binary one = Signal "mark" = Control Off Thus, a positive voltage on pin 20 means that DTR is "On". RS-232-C states that the voltages shall be in the range of 3.0 to 25.0 volts. Any voltage in the range of minus 3.0 to plus 3.0 volts is undefined. There may be a later update than RS-232-C, but I don't have a copy. Alan Frisbie ================================================================================ Note 217.2 Modems and DECServers 2 of 13 EISNER::CONROY "Alan Conroy" 30 lines 14-OCT-1988 20:12 -< DTR strangeness >- -------------------------------------------------------------------------------- Thanks, Alan, that's what I was afraid of. Well, here's a problem for all you techies out there... We have a DECServer 200/MC which has been connected to some UDS Fastalk 2400 baud modems. For various reasons we haven't been pleased with the Fastalks and got a good deal on some Hayes Smartmodem 2400s. When I connected one, it didn't work. I made sure it was configured the same as the UDS but it made no difference. I tried another Hayes and had the same problem. The symptoms were that the modem would hang up almost as soon as it answered the phone. (this is with the AT&D3 setting). So I disabled the DTR sense on the Modem (setting AT&D0) and it answered fine. I could even connect to the server. However, when I logged off, the modem never hung up. This makes sense because if the modem doesn't see the DTR drop, then it won't hang up. Since the variable here was the AT&D setting, which has to do with DTR, I figured that something strange was going on with DTR. So I connected a line monitor between the modem and server and guess what? I found that DTR drops twice! Once during connection and once during disconnection. The DTR drop at disconnection time is normal and documented, however it shouldn't drop during connection. I reconnected a UDS with the line monitor and found that the DTR still dropped but the UDS ignored the first drop, and hung up in response to the second. Both drops in DTR last about 3 or 4 seconds. I called CSC and they told me to make sure that the modem wasn't pulling DTR low. So I disconnected that signal and the DTR still dropped. Obviously not the modem. So I called CSC back and am still waiting to hear from them. Anyone have any idea what is happening here? This is LAT V5.1, DECServer software V2.0. Barring any solution on the DEC front, anyone know how to make the Hayes ignore the first DTR drop but not the second one? Thanks in advance. ================================================================================ Note 217.3 Modems and DECServers 3 of 13 EISNER::KENNEDY "Terry Kennedy" 8 lines 14-OCT-1988 23:53 -< More info needed... >- -------------------------------------------------------------------------------- > Anyone have any idea what is happening here? This is LAT V5.1, DECServer > software V2.0. That first drop is *wrong*. Verify that pins 4 & 5 are both connected, as well as 8. For more help, upload a copy of SHO PORT n CHAR as well as a detailed cable pinout. Also, the rest of the observed signals would help. There will probably be a delay in my answering, as I'm off to the Symposium... ================================================================================ Note 217.4 Modems and DECServers 4 of 13 EISNER::PRIGOT "Jonathan M. Prigot" 14 lines 17-OCT-1988 09:40 -< Signal sequence >- -------------------------------------------------------------------------------- The rest of the signals (CTS, CAR, and DSR) are very important for this. Check out the diagram on Page 8-14 of the VMS V5 "I/O User's Guide Part 1." (It is also in the V4 documentation, I just don't know the page number.) From my experience (admittedly with a DHU11), the sequence should be: 1. DECserver raises and holds DTR and RTS. Modem holds DSR, CAR (carrier), and RTS low. 2. Modem raises RI (ring indicate) and DSR. (Phone rings) 3. Modem raises CAR and CTS. (Remote carrier detected) 4. Communications takes place. 5. After logout, VMS drops DTR and RTS, and expects the modem to drop DSR, CTS, and CAR. ================================================================================ Note 217.6 Modems and DECServers 6 of 13 EISNER::CONROY "Alan Conroy" 17 lines 17-OCT-1988 15:49 -< And the winning combination is... >- -------------------------------------------------------------------------------- > The rest of the signals (CTS, CAR, and DSR) are very important for > this. Check out the diagram on Page 8-14 of the VMS V5 "I/O User's Right you are! CSC called me back today and suggested the following commands to set up the modem properly: AT&D3&W AT&C1&S1&W ATS0=1&W Apparently you need the EXACT combination of settings, or no go. The apparent problem (since this fixed it) was that the modem wasn't sending the proper signals to the DECSerer (DSR, DCD, and RI are the biggies). This confused mr. DECServer into dropping DTR for a while, with attendent problems. With this setting, DTR doesn't drop during connection. Note that CSC also told me that DTRWait HAD to be set for the Hayes Smartmodem 2400, but I found that the setting made no difference for me, either way. ================================================================================ Note 217.7 Modems and DECServers 7 of 13 EISNER::CONROY "Alan Conroy" 6 lines 20-OCT-1988 11:50 -< Cabeling info. >- -------------------------------------------------------------------------------- I have found that the cable going between the Hayes Smartmodem 2400 and the DECServer must NOT pass pin 23. This is because the Hayes apparently connects pins 23 and 12. Pin 12 is used for signalling the modem speed. Pin 23 is used for signalling the server speed. Since the Hayes connects them, the server causes pin 12 to change when it changes pin 23. Thus it never detects the modem speed correctly. ================================================================================ Note 217.8 Modems and DECServers 8 of 13 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 37 lines 23-OCT-1988 00:05 -< This could also be a "chatty modem" problem >- -------------------------------------------------------------------------------- Another thing that can cause the EXACT symptom you reported is the "chatty modem" syndrome. Many Hayes-like modems (I haven't used a real Hayes model, so I can't say about them) send status codes like "RING" and "CONNECT 2400" to the local DTE, not only when dialling out, but also when a call comes in. The failure sequence goes like this: 1. Call comes in. 2. Modem notices it and sends "RING" to the DTE (ie to the VAX). 3. VAX autobauds and sends something -- anything, it doesn't matter -- to the modem. 4. The modem sees stuff coming from the DTE before it sees carrier from the remote modem, which makes it choke. It says "NO CARRIER" to the local DTE, and hangs up the line (dropping DSR). 5. In response to the dropping of DSR, the VAX drops DTE, just to make sure that the modem hangs up. The cure, on those modems that support it, is ATQ1, which enables Quiet mode. Dialout users must be trained to send the modem an ATQ0 before starting their call and an ATQ1 when they're finished. The Telebit Trailblazer has a nice solution: "Partial quiet" modes, in which the modem will be quiet except on outgoing calls. (It also has a nifty feature whereby it automatically resets itself to the "saved" configuratoin whenever the local DTE drops DTR. The "saved" stuff is stored in nonvolatile RAM. Now, if it just had a password so that J. Random Luser couldn't mess up the saved settings!) This problem does not occur on DMF32 modem ports, which send no data to the host until the modem raises Carrier Detect. But then you have to dial blind on outgoing calls, unless you can tell the modem to "manually" raise CD for an outgoing call (the Trailblazer, and at least one Racal- Vadic model, support this). ================================================================================ Note 217.9 Modems and DECServers 9 of 13 EISNER::CONROY "Alan Conroy" 11 lines 24-OCT-1988 13:14 -< Hayes isn't chatty >- -------------------------------------------------------------------------------- > "chatty modem" syndrome. Many Hayes-like modems (I haven't used a > real Hayes model, so I can't say about them) send status codes > like "RING" and "CONNECT 2400" to the local DTE, not only when > dialling out, but also when a call comes in. The failure sequence Yes. That's the reason we went with real Hayes modems, because they don't do that. For a dial-in/out application, the "ctty modem" type of equipment just doesn't work. Either you use quiet mode which makes dial-out use difficult, or you don't and you can't dial-in. I haven't worked out ALL of the bugs yet, but the Hayes modem's are the best I've yet seen for this application. ================================================================================ Note 217.10 Modems and DECServers 10 of 13 EISNER::MCALLISTER "Brian McAllister" 29 lines 25-OCT-1988 14:06 -< You can have you HAYES and user-friendly too... >- -------------------------------------------------------------------------------- > Either you use quiet mode which makes > dial-out use difficult, or you don't and you can't dial-in. I haven't > worked out ALL of the bugs yet, but the Hayes modem's are the best I've > yet seen for this application. You might want to look at VENTEL modems. They have both a "HAYES" mode (which seems very close to the real thing, and does not have the problem of being too 'chatty'), and a "VENTEL" mode, which give a very friendly (and verbose) user interface, complete with a dialing directory. The best thing is that, with the modem set to one mode, you can toggle into the other with a single command. This is nice, because you can leave them in 'HAYES' so people with PCs and dialing software can use 'AT' commands, but people with terminals have easy access to the user-friendly command interface. These come in 300/1200 and 300/1200/2400 versions, with speed-buffering and (intelligent) MNP on the 2400 version. One caution: When we tried leaving them in "VENTEL" mode, they would "lock up" frequently (>once a day), requiring power-off to recover. It seemed that it might be some kind of auto-baud problem (the "VENTEL" mode uses carriage-return), but we couldn't make it go away and Ventel had no explanation. The problem disappeared completely when we left them in 'HAYES' mode, even though the 'VENTEL' mode is frequently accessed. NOTE: My only connection with Ventel is as a happy customer. These modems work well, and seem to be quite reliable as well. ================================================================================ Note 217.11 Modems and DECServers 11 of 13 EISNER::CONROY "Alan Conroy" 6 lines 26-OCT-1988 14:01 -< Ventel modems >- -------------------------------------------------------------------------------- > You might want to look at VENTEL modems. They have both a Thank you. That's good to know. My initial experience with a Ventel modem (about 4 years ago), was exactly this type of application and it failed miserably. From what you say it sounds like they've worked the bugs out of their latest offerings. ================================================================================ Note 217.12 Modems and DECServers 12 of 13 EISNER::MARTIN_T 37 lines 5-APR-1989 16:54 -< Another Hayes/DS200 Timing Problem >- -------------------------------------------------------------------------------- I've encountered an interesting timing problem trying to connect a Hayes- compatible modem to a DECserver 200/MC. I've setup the server port as follows: Port 1: Modem #1 Character Size: 8 Primary Speed: 2400 Flow Control: XON Alternate Speed: 1200 Parity: None Modem Control: Enabled Access: Dynamic Local Switch: ^A Backwards Switch: None Name: MODEM_1 Break: Local Session Limit: 4 Forwards Switch: None Type: Soft Enabled Characteristics: Autoprompt, Broadcast, Dialup, DTRwait, Inactivity Logout, Input Flow Control, Loss Notification, Message Codes Output Flow Control, Password, Verification The port will not sense the fallback speed unless the modem is strapped to track the incoming carrier (versus setting CD high forever). Yet, this same setting causes the server to disregard (i.e. throw away) the modem's CONNECT message intended for the user. This leads to blind dialing and an inability to programmatically automate dialing procedures. I've talked to DEC and the modem manufacturer and received the same explanation that I had already reasoned: it's a timing problem between the two devices. Apparently the server will not allow any data transfer until all pins have been satisfied and the modem is transmitting its response codes prior to that satisfaction. Other than an inability to receive the CONNECT status message, the connection works fine. Does anyone have any ideas or successful Hayes-compatible/DS200 modem recommendations? The modems I have tried are the EVERCOM EVEREX EV-945 and a LEADING EDGE 24E. Terry J. Martin State of Florida (904) 487-1455 ================================================================================ Note 217.13 Modems and DECServers 13 of 13 EISNER::TINKELMAN "Bob Tinkelman - CCA" 6 lines 6-APR-1989 00:40 -< Answer posted elsewhere >- -------------------------------------------------------------------------------- RE: < Note 217.12 by EISNER::MARTIN_T > > Other than an inability to receive the CONNECT status message, the > connection works fine. See note 19.37  ================================================================================ Note 218.0 Environmental Monitoring System 3 replies EISNER::TINKELMAN "Bob Tinkelman - CCA" 60 lines 23-OCT-1988 08:19 -------------------------------------------------------------------------------- Attached is a Digital press release announcing EMS.  DIGITAL ANNOUNCES NEW REALTIME ENVIRONMENTAL MONITORING SYSTEM MAYNARD, Mass. -- October 17, 1988 -- Digital Equipment Corporation today announced an electronic computer site surveillance system designed to provide timely warnings against environmental threats such as fire, water, extreme temperatures and excess humidity. The Environmental Monitoring System (EMS) is an advanced, microprocessor-based monitoring system built to handle the total electronic environmental surveillance of any computer site and its supporting equipment. It uses from one to 112 external sensors/ probes to provide realtime sampling of physical environmental changes up to 5,000 feet away, and compares these changes to user-established thresholds. "The 5,000 foot range of the EMS allows you to monitor computer rooms in multiple computer sites," says John Yurcak, Product Manager. "Your EMS unit can reside in one building and monitor additional computer sites almost a mile away from that building." The EMS responds immediately to any environmental abnormality with both a visual display and an audio announcement on its connecting terminal, allowing a quick response to equipment and data-threatening environmental events. An optional software package available for purchase with the EMS allows the system to send mail messages and to build historical environmental data files that can reveal environmental patterns leading to equipment breakdown. The EMS can function as a standalone unit, or can be daisy-chained together with up to seven additional EMS units to form a monitoring network. The EMS can also be daisy-chained to Digital Power Conditioning System Plus or Power Distribution System Plus Units to expand the monitoring capabilities of an existing network. When alarms occur in a configuration of more than one unit, the alarm is displayed simultaneously on the unit detecting the alarm condition and on the master unit. "The EMS is also an economical solution to an organization's environmental surveillance needs," says Yurcak. "It is readily installable in computer sites with conventional wiring systems." The EMS is available immediately on a worldwide basis. Digital Equipment Corporation, headquartered in Maynard, Massachusetts, is the world's leading manufacturer of networked computer systems and associated peripheral equipment, and is the leader in system integration with its networks, communications, software, and service products. #### CORP/89/409 ================================================================================ Note 218.1 Environmental Monitoring System 1 of 3 EISNER::TINKELMAN "Bob Tinkelman - CCA" 2 lines 23-OCT-1988 08:25 -< Cross reference to REMS >- -------------------------------------------------------------------------------- For a description of REMS, the associated Digital software product, see the DEC_SOFTWARE conference, topic 122. ================================================================================ Note 218.2 Environmental Monitoring System 2 of 3 EISNER::ANGLEY "Kevin Angley" 4 lines 7-NOV-1988 17:37 -< A Yugo alternative to DEC's Mercedes >- -------------------------------------------------------------------------------- Yawn ... my $59 Radio Shack Sensaphone works just fine. Calls me up in the middle of the night to say its hot. Love it. (But then again my needs are fairly simple - I just want to know when the air breaks or the power fails). ================================================================================ Note 218.3 Environmental Monitoring System 3 of 3 EISNER::RENES "Stuart Renes, DFWLug Chair" 20 lines 8-NOV-1988 13:41 -< Nightmare on Elm Street? >- -------------------------------------------------------------------------------- I had one of those (*^^&@*(&damn sensaphone modules once and had (cleverly??) programmed several telephone numbers for it to call when trouble was sensed. One of the numbers was mine and another was my bosses (at home!) Seems like when this little box sensed a power loss or whatever it dialed my number (I didn't answer - clever huh?) and then dialed up my boss (who like me, was at work). But his wife was home and after receiving 20 or so (as she put it _CRANK calls from some joker with a SICK voice) she called hubby who came down to the computer room to see what was happening. Needless to say, his wife was UNIMPRESSED with our little CPU squealer! It now lives quietly (unplugged) in a cabinet! Stuart. :-)  ================================================================================ Note 219.0 LN03X-DE barcode cartridge for LN03 No replies EISNER::TINKELMAN "Bob Tinkelman - CCA" 63 lines 25-OCT-1988 10:29 -------------------------------------------------------------------------------- Attached is the Digital Press Release announcing the LN03X-DE, a barcode cartridge for the LN03 printer.  MAYNARD, Mass. -- October 27, 1988 -- Digital Equipment Corporation today announced availability of a new barcode font cartridge for Digital's LNO3 laser printers. The cartridge produces all 43 data characters (0-9, A-Z, 6 symbols and space) of industry standard Code 39, the alphanumeric bar code symbology most widely used in manufacturing, health care, and many government agency applications. "Bar coding is one of the most accurate means of data entry," said Joe Meany, Marketing Manager, Terminals and Printers Group, Digital Equipment Corporation. "It is a low-cost, highly effective way of marking products to streamline their identification, and is used extensively in virtually every industry today." The bar code symbols can be used to identify the product's price, description, manufacturer, and other information for accurately counting, tracing, and tracking items. LN03X-DE Cartridge The LN03X-DE font cartridge contains two bar code heights: 18-point (bars only), 36 point (bars with characters below bars). The 18- and 36-point heights can be combined to create a variety of bar heights. The LN03X-DE is designed for use with Digital's LN03 and LNO3 Plus laser printers, desktop units that print on 8.5"x11" cutsheet, transparencies, special labels and European A4 paper at 8 pages per minute. The LN03 printer features 16 resident fonts, 96 Kbytes of memory expandable to 256 Kbytes, and plug-in connections for two ROM and/or RAM font cartridges. The LN03 Plus also has space for two option RAM or ROM font cartridges, but contains 17 resident fonts, and 256 Kbytes of memory expandable to 1Mbyte. The LN03X-DE sells for $299. Bar Code Software "High-quality bar code printing software is key to the successful bar coding solution," said Meany. Among the companies offering bar-code printing software is Integrated Software Design, Inc. (ISD), based in Mansfield, Massachusetts. ISD recently introduced a VMS version of its memory-resident program ON-TAP/VMS. Formerly available only for MS-DOS-based personal computers including Digital's VAXmate personal computer, ON-TAP is now compatible with Digital's VAX-based systems. ON-TAP can generate all U.S. and European bar code symbologies. It can be used to add bar codes to letters, spreadsheets, checks, forms, invoices, mailing labels, product ID labels, shipping labels, and purchase orders. ON-TAP/VMS supports most of Digital's dot matrix and laser printers, requires no special programming ability to use and is available directly from ISD. Digital Equipment Corporation, headquartered in Maynard, Massachusetts, is the world's leading manufacturer of networked computer systems and associated peripheral equipment, and is the leader in systems integration with its networks, communications, software, and service products. #### VAXmate, VMS, LN03 and LN03 PLUS are trademarks of Digital Equipment Corporation. ON-TAP is a registered trademark of Integrated Software Design, Inc.  ================================================================================ Note 220.0 DELUA PDP-11 OS Download Problems / ECO's / FCO's 20 replies EISNER::SIMONS "Paul, Not that CONVEX!" 7 lines 27-OCT-1988 18:10 -------------------------------------------------------------------------------- I still can't quite believe this. Does anybody out there download operating systems over DELUA's ? I'm trying to download an OS between two PDP-11/70's using DELUA's. The trigger command works, but I don't have any boot ROMs installed. The load command doesn't work at all. I talked with Colorado Support, they verified my predicament and SPR'ed it. Am I missing something ? Don't new terminal servers use DELUA's ? ================================================================================ Note 220.1 DELUA PDP-11 OS Download Problems / ECO's / FCO's 1 of 20 EISNER::DELARISCH "Arnold S. De Larisch" 8 lines 27-OCT-1988 19:08 -< NOPE! DECservers still use DEQNAs I believe! >- -------------------------------------------------------------------------------- >> Don't new terminal servers use DELUA's ? Nope! I believe that the DS500 is still using DEQNAs because of the very problem you just ran into! Can someone verify this? -Arnold ================================================================================ Note 220.2 DELUA PDP-11 OS Download Problems / ECO's / FCO's 2 of 20 EISNER::MCCORMICK_B "Bob McCormick" 6 lines 27-OCT-1988 20:06 -< DELQA I think I saw... >- -------------------------------------------------------------------------------- Boy, I could have swarn that the DECserver 500 we installed on Tuesday had a DELQA in it. I also know that the DECserver 550 announcement indicated that it would ship with (correct me if I'm wrong) a DESQA, which isn't really anything more than the DELQA (half height, remember?) with a DESTA on the other half height space (all modules in the BA200 series box take up the full slot even if they are half height). ================================================================================ Note 220.3 DELUA PDP-11 OS Download Problems / ECO's / FCO's 3 of 20 EISNER::MAYHEW "Bill Mayhew" 3 lines 27-OCT-1988 20:20 -< UNIBUS vs. Q-bus >- -------------------------------------------------------------------------------- Hold on, the original questioner was asking about a DELUA, which is a UNIBUS device, not a DELQA etc., which are Q-bus devices. ================================================================================ Note 220.4 DELUA PDP-11 OS Download Problems / ECO's / FCO's 4 of 20 EISNER::MCCORMICK_B "Bob McCormick" 8 lines 27-OCT-1988 21:55 -< Ah... two questions?! >- -------------------------------------------------------------------------------- > Don't new terminal servers use DELUA's? No. DECserver 500/550 use Q-Bus DELQA/DESQA devices. Other terminal servers such as DS100, DS200, MS100 are 'single board' servers; ethernet is built in. U running M+ ... what versions? The Unibus DELUA works under VMS if memory serves me correctly! ================================================================================ Note 220.5 DELUA PDP-11 OS Download Problems / ECO's / FCO's 5 of 20 EISNER::FULLER_S "Stuart Fuller" 2 lines 27-OCT-1988 23:54 -------------------------------------------------------------------------------- FWIW, DECSA boxes (DECnet router, SNA gateway, etc) use a DEUNA, which is the older 2 board set. ================================================================================ Note 220.6 DELUA PDP-11 OS Download Problems / ECO's / FCO's 6 of 20 EISNER::FULLER_S "Stuart Fuller" 12 lines 27-OCT-1988 23:59 -------------------------------------------------------------------------------- What about the LOAD command doesn't work? There was an ECO to the one of the boards in the 11/70 cpu years ago to allow "remote booting". BAsically, it was a set of jumpers on one of the boards (M8131?) that allowed you to set up a boot address, of either 165024 or 173024, so that at power up, the cpu would access one of those addresses, instead of location 24. I don't have the details here at home with me, but if you like, I'll see what I can find out about it. Stu ================================================================================ Note 220.7 DELUA PDP-11 OS Download Problems / ECO's / FCO's 7 of 20 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 20 lines 28-OCT-1988 01:44 -< what supported when!? >- -------------------------------------------------------------------------------- The previous version of DS500 s/w ONLY worked with DEQNAs. S/W V1.1, that is now shipping, supports DELQAs and is necessary for the DS550 that has 1.5 meg + a DESQA. The DESQA (as noted above is a DELQA + DESTA) has a switch to select AUI or BNC type cables. If some V1.1 s/w user would try an 11/73 cpu + q memory + the DS550 boot proms in a seperate boot board as a quick test replacement for the 11/53 in their DS5x0 box, I would LOVE to hear the results... I seem to recall that DELUA support came to M+ around V3.0. It does work on a 3.0 machine we have, and I am pretty sure it did NOT work under 2.1 (but you could use the 3.0 era Decnet for 2.1 with a little hacking...). The DECSAs always came with the 2 card DEUNAS, but has anyone tried a DELUA? or better yet what about trying any of the hop-up cpu cards to replace the lowly 11/24 therein? Maybe some timing would break. ================================================================================ Note 220.8 DELUA PDP-11 OS Download Problems / ECO's / FCO's 8 of 20 EISNER::ERSKINE "Chris Erskine" 8 lines 28-OCT-1988 06:15 -< Working on 11/24's >- -------------------------------------------------------------------------------- We have used DELUA's on 11/24's for Shop Floor Gateways. These beast are running a 11/S which is down loaded from our VAXen. The only problem that I recall using them was when 2 or more VAXen are setup to load them. The get into a bidding war which confuses the PDP. We did get a new set boot ROM's to go with it if I recall. Chris ================================================================================ Note 220.9 DELUA PDP-11 OS Download Problems / ECO's / FCO's 9 of 20 EISNER::SIMONS "Paul, Not that CONVEX!" 19 lines 31-OCT-1988 09:45 -< Jumpers?! >- -------------------------------------------------------------------------------- <<< EISNER::DUA0:[NOTES$LIBRARY]HARDWARE_HELP.NOTE;1 >>> -< HARDWARE_HELP >- ================================================================================ Note 220.9 Downloading Blues 9 of 9 EISNER::SIMONS "Paul, Not that CONVEX!" 10 lines 31-OCT-1988 09:36 -< Jumpers!? >- -------------------------------------------------------------------------------- > There was an ECO to the one of the boards in the 11/70 cpu years > ago to allow "remote booting". BAsically, it was a set of jumpers > on one of the boards (M8131?) that allowed you to set up a boot > address, of either 165024 or 173024, so that at power up, the cpu I talked with my hardware guy, and he didn't know about this. Our 11/70's are a mix of old and new, so this may very well apply. Please supply any additional info! He recognized the two addresses as applying to the M9301YH and the M9312 boot boards, and wondered if the jumpers went to those. BTW, we are using M+ 4.0, and our DELUA's are brand new, and what does FWIW mean? ================================================================================ Note 220.10 DELUA PDP-11 OS Download Problems / ECO's / FCO's 10 of 20 EISNER::FULLER_S "Stuart Fuller" 27 lines 31-OCT-1988 10:38 -------------------------------------------------------------------------------- The jumpers in question were on the M8130 module of the 11/70 cpu. They set up the "start vector" (SV), and I believe they are normally wired to location 24 (the power up/down interrupt vector). For "push button boot", and "boot on power up", the SV must be jumpered to 773224 (see the M8130 prints, rec C or later, sheet 6). You should also have the following revs: M9301-Yx at least rev K (I suspect that the M9312 would have the required revs already) M8130 at least rev C M8138 at least rev C1 or rev E M8136 at least rev E backplane at least rev J Have your hardware guy reference the 11/70 techtips, TT-4 for more details FWIW := For What It's Worth. We need to get Bill Mayhew to get his list of TLAs put together!  TLA := Three Letter Acronym Stu ================================================================================ Note 220.11 DELUA PDP-11 OS Download Problems / ECO's / FCO's 11 of 20 EISNER::SIMONS "Paul, Not that CONVEX!" 7 lines 22-NOV-1988 13:58 -< More Info? >- -------------------------------------------------------------------------------- > Have your hardware guy reference the 11/70 techtips, TT-4 for more > details I talked to him and he said that, because we do our own maintenance, we don't have access to "techtips". Is there another way to get the information ? He checked the rev levels from your note, and we are OK in that respect. ================================================================================ Note 220.12 DELUA PDP-11 OS Download Problems / ECO's / FCO's 12 of 20 EISNER::FULLER_S "Stuart Fuller" 5 lines 23-NOV-1988 21:47 -------------------------------------------------------------------------------- I'll see if I can look up the info next week, and post it here. Sorry, I didn't know you weren't under DEC maintenance! Stu ================================================================================ Note 220.13 DELUA PDP-11 OS Download Problems / ECO's / FCO's 13 of 20 EISNER::SIMONS "Paul, Not that CONVEX!" 15 lines 23-JAN-1989 16:43 -< The answer?! >- -------------------------------------------------------------------------------- The results of the SPR are back. It would seem that DELUA's are a product of CSSE. We installed our PDP 11/70's in 1980. The system was developed using equipment purchased in 1976. Quite a few of the interfaces we purchased at that time were from CSS (DV-11, DQ11, PCL11-A, etc.). The experience has served to ingrain in us a fervent desire to avoid any products from CSS (CSSE as they are now called). And now we find that DELUA's are from them! Anyway, according to DEC Engineering, our problem has been fixed by a ROM change. We were told in our SPR to contact FS, and they would provide us with the new ROM's. After some grief, we were told that an ECO will *not* be made available, that only new boards will be furnished, and, for 60% of list, we can get boards that might do what DELUA's are advertised to do! This is, of course, not reasonable in our case. Their argument was that not enough sites use DELUA's to warrant an ECO. How can we get DEC to move on this and supply us the ROM's? ================================================================================ Note 220.14 DELUA PDP-11 OS Download Problems / ECO's / FCO's 14 of 20 EISNER::MAYHEW "Bill Mayhew" 6 lines 23-JAN-1989 17:24 -< CSS != CSSE ? >- -------------------------------------------------------------------------------- Somebody check me on this, please, but I am 90%+ certain that CSS is NOT the same as CSSE. CSS still exists ("Computer Special Systems" -- basically the 'custom' folks) under that acronym. CSSE was once deciphered for me as something completely different, but I don't remember what. The Digital Reference Service contains a list of CSS products, and the DELUA is not among them. ================================================================================ Note 220.15 DELUA PDP-11 OS Download Problems / ECO's / FCO's 15 of 20 EISNER::KENNEDY "Terry Kennedy" 22 lines 23-JAN-1989 20:46 -< Mainly a terminology problem, I think... >- -------------------------------------------------------------------------------- > ... After some grief, we were told that an ECO will *not* > be made available, that only new boards will be furnished, and, for > 60% of list, we can get boards that might do what DELUA's are advertised > to do! This is, of course, not reasonable in our case. Their argument > was that not enough sites use DELUA's to warrant an ECO. How can > we get DEC to move on this and supply us the ROM's? Huh? I think somebody got their terminology mixed up when they talked to you. Indeed, ECO's are *not* furnished to the field. An ECO may be for something like a vendor change on an IC, for example. ECO's which *do* have an impact on functionality are released to the Field as FCO's. Thus, an FCO always has a corresponding ECO, but not necessarily the other way 'round. Now that we've gotten the terminology out of the way, do you have a part number (probably EQ-something or FA-something) for the fix? I'll try to look it up tomorrow anyway. There *has* been FCO activity on the DELUA in the last year, so it may be in my list. Remember that 'available' may mean by special order (as opposed to DECdirect), but if you are on DEC FS contract, then it's *their* problem to order it. Anyway, I'll look it up tomorrow... ================================================================================ Note 220.16 DELUA PDP-11 OS Download Problems / ECO's / FCO's 16 of 20 EISNER::SIMONS "Paul, Not that CONVEX!" 4 lines 24-JAN-1989 16:39 -< Never a part # >- -------------------------------------------------------------------------------- Alas, the lack of a part number is exactly our problem. The fixes will come out as a new rev DELUA. DEC wants us to trade in our DELUA's, give them $12,000 *more*, and hey, maybe they'll do what they were supposed to in the first place! ================================================================================ Note 220.17 DELUA PDP-11 OS Download Problems / ECO's / FCO's 17 of 20 EISNER::KENNEDY "Terry Kennedy" 10 lines 25-JAN-1989 07:24 -< Possible EQ part number >- -------------------------------------------------------------------------------- According to the Fiche, the current DELUA rev is CS F. That fixed some bugs in 8300's and 11/84's, but no mention of the download problem, although I'm *sure* I saw it mentioned somewhere in the fiche. Of course, searching N thousand frames just isn't going to get done... Anyway, this means that DEC will ship you a rev F DELUA if you order one today. So, the FCO to get to {ev F *may* help you. I'd suggest you try ordering one to see if it's helpful (either through Field Service or DEC Direct). The part number is EQ-01501-01, at MLP of $75.00, and *hooray* it's a 30-day stocked part. ================================================================================ Note 220.18 DELUA PDP-11 OS Download Problems / ECO's / FCO's 18 of 20 EISNER::KENNEDY "Terry Kennedy" 3 lines 27-JAN-1989 20:35 -< Could also try DECmailer >- -------------------------------------------------------------------------------- One other thing you could try is DECmailer (if you have an account). The flat-rate repair and upgrade to current rev is $375.00 for a DELUA (I looked it up in the new price book yesterday). ================================================================================ Note 220.19 DELUA PDP-11 OS Download Problems / ECO's / FCO's 19 of 20 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 11 lines 27-JAN-1989 21:49 -< getting latest ECO another way >- -------------------------------------------------------------------------------- I guess the 'ethics' of this should be self evident. I will word this carefully, and this should NOT be taken as a suggestion, just a story once told. I have heard of a site that needed an upgrade to some card but F/S had refused because it wasn't a 'mandantory' one. A LARGE electrolytic cap suitable charged accidently fell so its leads connected to some poor IC's tender legs... ================================================================================ Note 220.20 DELUA PDP-11 OS Download Problems / ECO's / FCO's 20 of 20 EISNER::KENNEDY "Terry Kennedy" 3 lines 28-JAN-1989 00:22 -< FS spares may not be at current rev level >- -------------------------------------------------------------------------------- I'd just point out that FS spares, while supposed to have parts at the latest revision level, may not be. There are a number of circumstances which cause this, but that's a better topic for SOAPBOX...  ================================================================================ Note 221.0 LLF01 - Long line fiber printer interface No replies EISNER::TINKELMAN "Bob Tinkelman - CCA" 41 lines 2-NOV-1988 11:49 -------------------------------------------------------------------------------- Attached is the Digital press release announcing the LLF01, a fiber interface for (some) existing parallel printers.  REMOTE HIGH SPEED PRINTER CAPABILITY AVAILABLE Maynard, Mass. -- October 31, 1988 -- Digital Equipment Corporation today introduced a new bus independent printer cable interface, based on fiber optic technology, which represents a unique Digital solution for the remote placement of its medium and high speed parallel line printers. The LLF01, long-line fiber optic interface, is compatible with Digital's Q-Bus, VAX BI Bus and UNIBUS controllers and Digital LG01, LG02, LP27 and LP29 parallel printers. These printers produce hardcopy at speeds ranging from 600 to 2,000 lines per minute. The LLF01 plugs into standard data ports of the printer and host computer controller for a seamless transition from short line to long line transmission. With this state-of-the-art communications device, parallel line printers can now operate up to 1,500 feet from the host computer, whereas the normal distance has been less than 30 feet. This means printers can be moved outside the data processing site, to free up valuable floor space in the computer room and eliminating paper dust. For increased convenience to hardcopy users, these line parallel printers can now be placed at different locations within a building, in dedicated printer rooms or in departmental work areas where large volume hardcopy is required. When compared to conventional coaxial data cables, fiber optic cable technology provides greater reliability of data transmission over long distances, and increased security of confidential data from unwarranted access. The LLFO1 also features a comprehensive self-test capability for easy diagnostics and fault isolation. Digital Equipment Corporation, headquartered in Maynard, Massachusetts, is the world's leading manufacturer of network computer systems and services. Digital's VAX computer family integrates the enterprise, from the individual and work group to the whole organization. #### VAX BI Bus, Q-Bus, LLF01, LG01, LG02, LP27, LP29 and Q-BUS are trademarks of Digital Equipment Corporation.  ================================================================================ Note 222.0 Feedback on Qbus GCR tape systems wanted! 3 replies EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 14 lines 4-NOV-1988 16:13 -------------------------------------------------------------------------------- Does anyone have experience with third-party 9-track tape subsystems with MicroVAXes? We are looking at a subsystem available from Digital Basics. CDC 92185-02H drive (the same drive DEC uses in the TU81E), Dilog DQ153 controller. This gives 1600/6250 capability for about what DEC charges for 1600 only ($13K, including cabinet). I also want to consider a Cipher drive (one of those front-load autoload things). Is anyone successfully using one of these? Note that this will NOT be our primary backup medium; we're getting an 8mm drive for that. ================================================================================ Note 222.1 Feedback on Qbus GCR tape systems wanted! 1 of 3 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 60 lines 4-NOV-1988 19:06 -< 6250 on MV2 >- -------------------------------------------------------------------------------- >>>< Note 222.0 by EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" > >>> -< Feedback on Qbus GCR tape systems wanted! >- >>>Does anyone have experience with third-party 9-track tape subsystems >>>with MicroVAXes? Sure, we have been using a Kennedy 9400 800/1600/6250 for some time. It is on a Dilog DQ142, and, via an Iverson switch, is also available to PDP11 using a Dilog Unibus card. Good vacuum column drive once they ironed out its first year bugs some years ago. Only gripe is lack of MS: boot support from DEC. >>>We are looking at a subsystem available from Digital Basics. CDC >>>92185-02H drive (the same drive DEC uses in the TU81E), Dilog DQ153 >>>controller. This gives 1600/6250 capability for about what DEC >>>charges for 1600 only ($13K, including cabinet). Suprised you are paying that much, but maybe the cabinet... You DID say DQ153. There MAY be a problem. There IS a problem with the DQ/DU142 (and ...132s) that are a TS emulation, and the 92185 drive. The TS drive can read backwards, the CDC streamer can't. The read backwards emulation in done by guess what: BACKSPACE, READ-FORWARD, BACKSPACE. Vms BACKUP apparently makes use of read backwards, I think for /VE, on drive types that it thinks should support it, and performance is HORRIBLE as the streamer takes forever. The DQ153 is a TMSCP emulation, and hopefully BACKUP should behave better, BUT I think there may still be some problem with the DQ153. Carl Friedburg has been heavily involved with this and has been trying to get Dilog to help solve the problem, as he and at least one customer of his are affected using xx142s. Call Dilog and get their tech support type that specializes in tapes, and ask if he is the one Carl has been talking to. If so, you should be able to get reasonable answers about problems/fixes. The DQ153 can declare itself to be a TU81, and a MV2 can't boot it! but can also lie and say it is a TK50, and then becomes bootable! Your choice. Remenber that Dilog designed the TU80 controller for DEC (it is a DU132 that only does 1 instead of 4 drives). Dilog also did the TK25 controller, so they do have potential access to info inside DEC other venders may not. >>>I also want to consider a Cipher drive (one of those front-load >>>autoload things). Is anyone successfully using one of these? We had one here on a PC, and had temporarily plugged it into a DQ142 on a MV2, and it ran well, but only had it there for a few hours. We had its transfer rate cranked way up (it is a cached drive that allows transfer speed throttling) and we were able to keep it streaming with BACKUP apparently better than I would have expected after frustrating experiences with real TU81s. No proof, just an impression. A customer has a Kennedy 9610 (800/1600/3200/6250) streamer they are quite happy with. ================================================================================ Note 222.2 Feedback on Qbus GCR tape systems wanted! 2 of 3 EISNER::MCCORMICK_B "Bob McCormick" 33 lines 4-NOV-1988 21:29 -< That's what we ordered! >- -------------------------------------------------------------------------------- I purchased the package you indicate from Digital Basics some time ago. What I wanted was a 1600/6250 tape drive that looked like a TU81. This was 5 months ago. I was shipped the tape drive, and the DQ152 controller... I was sure supprised when it looked like a TS11... and, of course, I plugged it into a MicroVAX II running MicroVMS V4.5. Come to find out they did not have access at that time to the board that really would do the emulation. So, I had to get the driver. About 3 weeks ago I received a box with a ROM and some jumpers in it. I'm told that if the following steps are done I'll be able to turn my CDC drive and TS11 style controller into what I wanted: - change ROMS - change jumpers - change something in the tape drive - execute some config code (on the ROM)? I called Digital Basics 'cuz I only had prelim manual for DQ152, and new that I really wanted athe DQ153. Their tech expert has told me that we should be OK with the 152 etch and getting it to look like the DQ153 (can anyone confirm/deny this?) Anyway, we have yet to do this, as there is a project under way on this system that precludes my ability to take it down during the day, which I have to do so that I can have someone at Digital Basics walk me through the procedure of what has to be done! So, I'm somewhat frustrated... I wanted (what was advertised) a TU81 at 1/2 the price of the DEC drive... and have waited some time to get there... and have to rely heavily on them to get us converted over... I don't have a good feeling about it (the TS11 to TU81 cvt) ================================================================================ Note 222.3 Feedback on Qbus GCR tape systems wanted! 3 of 3 EISNER::GOETZ "Keith Goetz" 12 lines 4-NOV-1988 22:12 -< how about a Kennedy 9610? >- -------------------------------------------------------------------------------- >>>Does anyone have experience with third-party 9-track tape subsystems >>>with MicroVAXes? Sure, I've been using a Kennedy 9610 800/1600/3200/6250 for about two years. It is on a Dilog DQ142. It has been a very good drive. Sounds a little like a vacuum cleaner as it loads the tape but after that it's great. It is compact, quiet, low in power consumption (and also heat production) and pretty fast too. Do I sound like a satisfied customer? With my University discount, these are about $8k each.  ================================================================================ Note 223.0 6200/8650 comparison No replies EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 20 lines 10-NOV-1988 16:18 -------------------------------------------------------------------------------- Just a simple question to start a religious war(?): I have been looking at the DEC sales info about 6200 machines and the 8650, and it seems to say something along the lines of: 6210 = 0.5 x 8650 6220 = 1.0 x 8650 6230 = 1.5 x 8650 6240 = 2.0 x 8650 Given the price differential one leans heavily towards the 62xx solution, as long as the systems scale well. Now, all of you that have these new beasts, does the 62xx series appear to scale as well as DEC claims? Just to keep everyone from guessing in a total vacume our environment at PBS is about 200 users distributed between 2 8650s, and growing with general office automation (ALL-IN-1), data base, and general computing usage. We currently show multiple entries in the CPU queue (anywhere from 4 to 12 on average per cpu during prime hours).  ================================================================================ Note 224.0 Aviv SCSI-KLESI Interface for tape drives 2 replies EISNER::OSUDAR "John Osudar" 13 lines 10-NOV-1988 21:27 -------------------------------------------------------------------------------- Does anybody out there have any experience with, or know anything about, the Aviv Corp. SKI-series adapters, which allegedly allow various tape drives to be connected to DEC's BI KLESI controller (the one used for a TU81-Plus)? There was an article on the front page of the 31-Oct-1988 Digital News, but it contained some obvious inaccuracies (e.g. references to "a TU81-class disk drive") and not nearly enough detailed information about the product itself. It claimed that the SKI adapter allows various nine-track, helical scan, and IBM 3480 tape drives to be connected to DEC's BI controller. I've heard of Aviv before, but know virtually nothing about the company. Anybody have any good/bad things to say about them? Thanks for your help... ================================================================================ Note 224.1 Aviv SCSI-KLESI Interface for tape drives 1 of 2 EISNER::RAY_T 16 lines 12-SEP-1989 16:10 -< AVIV says 80 units in the field now >- -------------------------------------------------------------------------------- I have before me a quotation for AVIV's 8MM Helical Scan Cartridge Tape system Model TFS AB8202-RM for DEC VAX 6000 and 8000 series BI-Bus systems. The quotation stipulates that DEC's KLESI-B tape adapter is required. Their price for this unit is $16,350. The New Jersey Board of Public Utilities, where I work, currently uses a VAX 8530 with a TU81+. AVIV is sending a technical rep down to discuss this quotation with us. I will keep you posted as things develop. In the mean time, their product description states that their system allows up to 9 gigabytes of unattended backup on four drives, the SKI controller is TMSCP compatible. the AB 8202 tape unit daisy chains off the TU81 or plugs directly into the KLESI BI adapter. Data transfer from each drive to the SKI controller is at the rate of 1.5MB/sec. During write, the SKI bursts up to 128 records to a 256Kbyte buffer for each drive. The AB 8202 consists of two drives, which can be dual ported to two separate hosts. ================================================================================ Note 224.2 Aviv SCSI-KLESI Interface for tape drives 2 of 2 EISNER::MAYHEW "Bill Mayhew" 5 lines 12-SEP-1989 20:32 -< Japanese Connection? >- -------------------------------------------------------------------------------- > AVIV's 8MM Helical Scan Cartridge Tape system Model TFS AB8202-RM ^^^ My. This sounds suspiciously like the 8mm subsystem | that DEC-Japan is selling, since it is known to come | from AVIV and its model number is TK820---------------------|  ================================================================================ Note 225.0 UDA,KDA,KDB support for RA60,RA70,RA81,RA82 2 replies EISNER::BYRNE_C "Charlie Byrne" 25 lines 11-NOV-1988 16:31 -------------------------------------------------------------------------------- Can I put an RA70, or RA82 on a UDA controller running off an 8650? (I really wouldn't want to put the RA70 there, but checking anyway.) From VAX Systems and Options 1988 Oct-Dec: Page 5.3: "Standard Interconnect storage devices include... SA600, SA482, RA70, RA82, RA81, and RA60" "ANY DRIVE OR STORAGE ARRAY CAN BE CONNECTED TO ANY CONTROLLER PORT ...The drives and storage arrays can be connected to the HSC-70,HSC50,KDB50,KDA50, and UDA50." However, page 5.15, RA 70 section: "Currently it is supported by the KDA50 controller on MV35/3600 computers". Is this just a packaging issue? Worse yet on page 5.19 under RA-82 they have a nice chart of configuring Information; however, only inof for the KDA and KDB is provided. Is DEC trying to sneak away from it's architectural commitments to the Unibus based UDA? It looks like they have a grandiose speil about their great architecture but you better check the implementations avaialable. (or am I missing something) ================================================================================ Note 225.1 UDA,KDA,KDB support for RA60,RA70,RA81,RA82 1 of 2 EISNER::KENNEDY "Terry Kennedy" 14 lines 11-NOV-1988 22:28 -< RA82 supported, RA70 works but not supported >- -------------------------------------------------------------------------------- > Worse yet on page 5.19 under RA-82 they have a nice chart of > configuring Information; however, only inof for the KDA and > KDB is provided. SA482 (4 x RA82) is sold and supported on UDA's. I have no reason to believe that RA82's sold as add-ons would not be. Check out the PDP-11/84 packaged system w/ UDA and RA82... Regarding the RA70, it is a storage component rather than a storage device. You need to feed it regulated DC power, and bring in the SDI port via ribbon cable, rather than BC26 cables. Yes, it works, but as DEC doesn't have an enclosure/power supply, don't expect support. *IF* DEC brings out a box similar to the RD5x-D (leprechaun) box for the 70, then they might support it... ================================================================================ Note 225.2 UDA,KDA,KDB support for RA60,RA70,RA81,RA82 2 of 2 EISNER::MCCORMICK_B "Bob McCormick" 7 lines 12-NOV-1988 19:30 -< RA70's in 3xxx expansion cab >- -------------------------------------------------------------------------------- You can get the 3500/3600 "expansion box" which has 11 Q-Bus slots and space for two RA70 disks. These disks terminate in the standard SDI cable ports, such that they can be cabled back to a 3500 or 3600 system. Sure, they'll work on any DSA compliant controller - that was the whole point of the architecture: be able to produce and market a family of disks (in this case RAnn) without having to rewrite drivers or build new controllers.  ================================================================================ Note 226.0 Graphics monitors, output standards, scan frequencies 5 replies EISNER::BURNET "John Burnet" 22 lines 11-NOV-1988 18:46 -------------------------------------------------------------------------------- Can anyone tell me whether there exists a formula to calculate a video monitor's horizontal scan frequency, given the resolution at which the monitor is being used? I suspect that the relationship is actually the other way around (the resolution is determined by the scan frequency), but due to my complete ignorance of video hardware :-), I don't even know where to start trying to figure this out. It would also be nice to know how the video bandwidth (pixel frequency) is calculated -- is it just directly related to the resolution (horizontal x vertical), or what? In lieu of a formula for the horizontal scan frequency, or in addition, can anyone say definitely what the scan frequency of a VAXstation 2000 (GPX) is? I assume that the monitor is driven by analog RGB RS-170A signals -- correct? I know that the resolution is 1024 x 864. Is the scan frequency such that it would be compatible with video output devices (monitors, film recorders) that accept a 64 kHz scan frequency and normally work at 1280 x 1024 resolution? See the next topic for an explanation of why I need this information, and the next reply for a summary of the information I do have about various video resolutions and PC graphics standards. Thanks... ================================================================================ Note 226.1 Graphics monitors, output standards, scan frequencies 1 of 5 EISNER::BURNET "John Burnet" 46 lines 11-NOV-1988 18:48 -< Some common resolutions, PC graphics standards, etc. >- -------------------------------------------------------------------------------- Here's some information I have collected about various graphics monitors' and PC boards' resolutions and other stuff. I'd appreciate any contributions to this list (replacements for the question marks where I didn't know a particular value), and any additional information people might have. 1. PC graphics standards (TTL RGB or monochrome) ================================================ Concurrent Bit planes Device Resolution colors [log2(max(colors,2))] ------ ---------- ---------- ------------------ MDA 720 x 350 1 + bg 1 Hercules 640 x 200 1 + bg 1 CGA 320 x 200 4 2 EGA 640 x 350 16 4 PGC 640 x 480 256 8 VGA, "high" (glork!) resolution 640 x 480 16 4 VGA, low resolution 320 x 200 256 8 MCGA, "high" resolution 640 x 480 1 + bg 1 MCGA, low resolution 320 x 200 256 8 Mac Jagger (Stan K-B's phrase) 540 x 342 1 + bg 1 Mac II 640 x 480 256 8 There are other, lower-resolution EGA modes I wasn't interested in; I haven't tried to make this list complete. 2. Analog video =============== Horizontal Resolution scan freq. (video bandwidth) ---------- ---------- ----------------- 512 x 512 ? ? VGA, etc. 640 x 480 31.5 kHz 40 MHz 1024 x 768 48 kHz? 80 MHz VAXstation 1024 x 864 ? ? 1280 x 1024 64 kHz 110 MHz? 107.5? 1600 x 1248 ? ? 1664 x 1248 ? ? 2730 x 2048 ? ? The resolutions higher than 1280 x 1024 I'm including only because I've seen monitors capable of those resolutions advertised and I'm curious about the scan frequencies and video bandwidths necessary to use them. ================================================================================ Note 226.2 Graphics monitors, output standards, scan frequencies 2 of 5 EISNER::BRIDGE "Adam Bridge" 5 lines 12-NOV-1988 14:22 -< VAXstation 2000 stats for table >- -------------------------------------------------------------------------------- The VAXstation 2000 monochrome display has a video bandwidth of 70 MHz, black negative, composite video. I don't have the info on the color monitor. ================================================================================ Note 226.3 Graphics monitors, output standards, scan frequencies 3 of 5 EISNER::HASSINGER "Bob Hassinger" 30 lines 14-NOV-1988 11:31 -< A little about scan rate, bandwidth and resolution... >- -------------------------------------------------------------------------------- A good question. It is hard to get really good information, partly I think because of all the specsmanship that goes on in promoting video displays. I believe horizontal scan frequency is not a direct indication of "resolution". It tells you something about the number of scan lines per second displayed. With that and the frame/field rate I can compute the number of scan lines per frame or field (depending on the use of interlace). That, in turn tells me something about the vertical resolution. Horizontal resolution depends on the video bandwidth. Dividing the video bandwidth by the horizontal scan rate should give you an indication of how much horizontal resolution you will get (i.e. how much bandwidth is available per scan line). Overall, it would seem video bandwidth would be the most interesting spec. It tells how much information you can display. As long as an assumption that the product makes a good division of the available bandwidth between horizontal and vertical resolution, it should tell you something about what you will see on the display overall. Horizontal scan rate is a less direct indication. For example, the horizontal scan rate on my Amiga's analog RGB monitor is the same as the standard NTSC composite video you use on your TV set. The resolution is very different however because the NTSC signal is limited to a bandwidth in the range of 5Mhz while the RGB signals can, and typically do, have a much higher bandwidth. The resulting vertical resolutions of the two signals and displays is the same, 500 odd lines, but the horizontal resolution is typically two or three times as great using the RGB signals as long as the monitor is capable of displaying that much bandwidth. ================================================================================ Note 226.4 Graphics monitors, output standards, scan frequencies 4 of 5 EISNER::MERRIMAN "George Merriman - CCA/NY" 7 lines 14-NOV-1988 13:44 -< look for video bandwidth and frame rate >- -------------------------------------------------------------------------------- > Overall, it would seem video bandwidth would be the most interesting > spec. It tells how much information you can display. Video bandwidth vs. frame rate is what you really want, as your analysis points out. Disregarding interlace and other overhead, dividing bits/sec (bandwidth) by frames/sec (frame rate) gives you bits/frame. ================================================================================ Note 226.5 Graphics monitors, output standards, scan frequencies 5 of 5 EISNER::BURNET "John Burnet" 0 lines 21-NOV-1988 10:19 -< .3 & .4 -- thanks... have info; will ponder >- --------------------------------------------------------------------------------  ================================================================================ Note 227.0 Film recorders for color graphics 3 replies EISNER::BURNET "John Burnet" 30 lines 11-NOV-1988 18:50 -------------------------------------------------------------------------------- I'd like to start a discussion on film recorders. We're looking at several different units with the thought of buying one for slide- and print-making. We'll need to feed it from at least three different sources: a PC VGA board; a high-resolution (1280 x 1024) PC graphics board from Number Nine that generates analog RGB output at a 64 kHz scan rate; and a VAXstation 2000 (GPX, 8-plane). We won't have the VAXstation until some time next year, which is why I asked about its video-out scan rate in a previous note. We've been looking at the Matrix (formerly Dunn Instruments) analog film recorders, especially the 6132 (32 kHz scan rate) with a scan converter to let it accept 64 kHz input. This looks like a great unit, since according to the sales literature I have (I know -- don't believe sales literature :-), it will accept 32 kHz TTL RGB input as well as analog RGB. That makes it perfect for us, since we'll need to feed it TTL from a VGA board and analog RGB from the #9 board and the VAXstation. The camera in the Matrix units also accepts three different backs -- one for 35 mm slide film, one for 4 x 5 Polaroid film, and one for 8 x 10 Polaroid. Does anyone have any experience with this unit, or with others that have similar capabilities? Can you recommend any manufacturer in particular? Are there any film recorders or manufacturers we should stay away from because of poor performance or support? We have looked at digital-only film recorders with rasterizers (the rasterizer is usually a PC board that takes HPGL input and drives the recorder). We have ruled those out, since we'll need to put raster images on film (from Autoshade, MOVIE.BYU, VAXstation screens, etc.) -- the digital recorders/rasterizers will only accept vector-graphics input. Thanks in advance for all information. ================================================================================ Note 227.1 Film recorders for color graphics 1 of 3 EISNER::KENNEDY "Terry Kennedy" 0 lines 11-NOV-1988 22:30 -< Why do you want to use TTL from a VGA (rather than analog)? >- -------------------------------------------------------------------------------- ================================================================================ Note 227.2 Film recorders for color graphics 2 of 3 EISNER::ERSKINE "Chris Erskine" 1 line 14-NOV-1988 06:48 -< In fact 'Standard IBM VGA' does not have TTL output >- -------------------------------------------------------------------------------- ================================================================================ Note 227.3 Film recorders for color graphics 3 of 3 EISNER::BURNET "John Burnet" 14 lines 14-NOV-1988 14:27 -< VGA: analog RGB out? (Thanks for the correction) >- -------------------------------------------------------------------------------- I hadn't been aware that the VGA produces an analog signal. I assume that it's available at the back of the board as separate R, G, and B outputs (sync on green or separate sync) -- is this right? I'm just so used to configuring PCs with EGA boards and monitors that I had assumed that the VGA would follow the EGA's lead of using an RGB TTL output signal. This may change the situation a little bit and open up a wider field in which more analog-only film recorders can compete, although the Matrix unit still looks like a good deal for our purposes. One thing I forgot to mention in .0 is that we are definitely shopping at the *low* end of the film-recorder market... preferably around (or under) $12k for a complete recorder, interface, etc.  ================================================================================ Note 228.0 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 23 replies EISNER::MCDOUGALL "Bob McDougall" 8 lines 14-NOV-1988 14:40 -------------------------------------------------------------------------------- We have a Fujitsu Eagle Disk Drive running on our 780 mass bus using the Emulex SC7000 Controller. The device was bought from and installed by Emulex and is now maintained by DEC. Emulex has just informed us that Version 5 of VMS does not support the SC7000 and as a result we need to purchase a UNIBUS Controller. What has happened to the Emulex promise of upward compatibility? If my memory serves me right Emulex has supplied a DRDRIVER, at no cost in the past. Comments? ================================================================================ Note 228.1 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 1 of 23 EISNER::KENNEDY "Terry Kennedy" 18 lines 14-NOV-1988 19:49 -< Your options >- -------------------------------------------------------------------------------- I didn't know Emulex *ever* promised *any* sort of compatibility, but that's another issue... I assume your drive looks like a RMxx or RPxx with a different geometry (number of cyls, heads, sectors)? If so then there are two possible prob- lems: 1) You definitely need to patch the drive geometry in the standard DEC driver. 2) The Emulex controller may not fully/properly emulate the DEC version. If the only problem is 1), and you have a fiche kit, then you should be able to continue using the drive and controller by doing your own support. If you are bit by 2), I suggest you get a Unibus controller (if your CPU is 11/7xx or 86xx) or a SMD-to-SDI adapter (for other big VAXen). I also suggest you consider the vendor's reputation for tech support, even if Emulex gives you a low price to 'make it up to you'. ================================================================================ Note 228.2 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 2 of 23 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 12 lines 14-NOV-1988 21:14 -< NEVER AGAIN WILL I BUY EMULEX >- -------------------------------------------------------------------------------- > is 11/7xx or 86xx) or a SMD-to-SDI adapter (for other big VAXen). I also > suggest you consider the vendor's reputation for tech support, even if > Emulex gives you a low price to 'make it up to you'. We can now add the SC7000 to the list of hardware Emulex has sold customers that will not run after a major O/S version upgrade. That list includes the TC02, TC05, CS02. I also assume that Emulex will tell you to take a flying leap if you ask them to pay the cost of their half-baked engineering. If any of you out their are still buying Emulex gear please send me mail I have a deal on Bridge in Brooklyn - trust me it is a LAN Bridge. ================================================================================ Note 228.3 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 3 of 23 EISNER::MAYHEW "Bill Mayhew" 10 lines 14-NOV-1988 23:52 -< How old is the SC7000? >- -------------------------------------------------------------------------------- I basically agree with Jeff, but... About what vintage is the SC7000? I really don't remember the numbers, but I was involved in legal action involving one of their controllers several years ago and if this is the one, I can probably shed some light on some of the specific "holes in the emulation" (to use Emulex's own words at the time)... i.e. explain what some of the technical problems are. However I can see no technical reason why they couldn't support *any* heretofore-VMS-compatible product with V5, _if they wanted to_. Worst it can take is a new driver, right? ================================================================================ Note 228.4 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 4 of 23 EISNER::HASSINGER "Bob Hassinger" 28 lines 16-NOV-1988 10:44 -< NEVER deal with Emulex! & How I run SC7000 under V5 >- -------------------------------------------------------------------------------- I too have mentioned here and elsewhere on more than one occasion my dissatisfaction with Emulex. My problems in fact relate explicitly to the SC7000, it's deficiencies, and the lack of support and commitment from Emulex. I still agree with the comment about NEVER dealing with Emulex and I recommend anyone else considering dealing with them take careful consideration of the comments in this thread! As it so happens my SC7000 IS functioning (as well as it ever did anyway) under VMS V5 with a CDC 9771. Apparently the only reason for this is that I had the wisdom back at the beginning to insist that the emulation *I* got was of two real, standard size RM05s rather than the customary one oversized drive. The cost of this has been to not use a fraction of the total drive capacity but the benefit is that I have never had to depend on those *&$#*&^^ Emulex patches. I have always been able to use DEC's standard RM05 driver. In fact, I have never seen one of Emulex's patches or modified drivers in all the time I have had the product. In any case, my experience suggests that at least in the simple, non-SMP case, the RM05 driver is still compatible with the SC7000 emulation so all that should be required is the geometry table patches for the driver or, alternatively to switch to a standard size emulation. Depending on the version of the firmware you should be able to change some switches on the SC7000 and pick one that uses a reasonable fraction of the drive. Tell me more about "known holes in the emulation...". It might provide clues to some of the things I see and don't understand. ================================================================================ Note 228.5 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 5 of 23 EISNER::CONROY "Alan Conroy" 14 lines 16-NOV-1988 13:21 -< Don't patch if you don't have to >- -------------------------------------------------------------------------------- > this is that I had the wisdom back at the beginning to insist that the > emulation *I* got was of two real, standard size RM05s rather than the > customary one oversized drive. The cost of this has been to not use a > fraction of the total drive capacity but the benefit is that I have > never had to depend on those *&$#*&^^ Emulex patches. I have always > been able to use DEC's standard RM05 driver. In fact, I have never > seen one of Emulex's patches or modified drivers in all the time I have > had the product. I, too, have followed this policy as much as possible. It certainly is not worth the hassle to gain the extra (relatively small) amount of disk space. I also did not tell anyone else that there was any extra potential capacity on the disk. What management doesn't know in this case, doesn't hurt them. ================================================================================ Note 228.6 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 6 of 23 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 18 lines 16-NOV-1988 20:23 -< EMULEX has SC7000 VMS 5.0 support >- -------------------------------------------------------------------------------- >>Emulex has just informed us that Version 5 of VMS does not >>support the SC7000 and as a result we need to purchase a UNIBUS >>Controller. What has happened to the Emulex promise of upward >>compatibility? If my memory serves me right Emulex has supplied >>a DRDRIVER, at no cost in the past. EMULEX has the expanded DB/DR drive size support for VMS 5.0 for your SC7000, right now. I don't have one of these SC7000s, but do have the CS02 blacklisted in .2 but that works fine except for having 8 phantom ports. I called EMULEX to find out what the bad news in .2 was, and it seems to only be a problem on the MV3 cpus. AND proms are now in the mail that will remove my 8 phantom CS02 ports. The guy I spoke to was NOT the Big VAX DISK guy, but did say the SC7000 fix is available and users should call 1-800-854-7112 ext 3384, and ask for Mark who IS the Big VAX DISK guy. ================================================================================ Note 228.7 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 7 of 23 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 9 lines 16-NOV-1988 20:30 -< QUESTION >- -------------------------------------------------------------------------------- > I don't have one of these SC7000s, but do have the CS02 blacklisted in > .2 but that works fine except for having 8 phantom ports. I called > EMULEX to find out what the bad news in .2 was, and it seems to only be > a problem on the MV3 cpus. AND proms are now in the mail that will > remove my 8 phantom CS02 ports. Was the Board over a year old? And if so did you have to pay for the firmware upgrade? ================================================================================ Note 228.8 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 8 of 23 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 55 lines 16-NOV-1988 22:12 -< I didn't pay - they aren't horrible. >- -------------------------------------------------------------------------------- >>Was the Board over a year old? Heavens yes! warantee expires 12/86 is glued to one of the big chips. >>And if so did you have to pay for the firmware upgrade? Of course not, this is Emulex, NOT DEC. Over the years I have had a fair number of new prom sets from them, and can't ever remember paying, but it has come close a few times. Sure, technically he is supposed to be paid after the first year. After all I am not paying him through-the-nose-monthly the way I have to pay DEC to get ECOs, but, also unlike DEC, Emulex seems to care about PR. I didn't make a point of pointing out it was that old, but did say I just got a similar ECO for a CS04 (that is truely only weeks old) and could he also fix my CS02. I am sure he knew what I was doing. Some ECOs you may feel are more fixing a built in screw-up, others are newly created by DEC changing something. The former you almost deserve free forever. The latter, well.. They are afterall trying to provide us viable alternative or even superior sources to what we need, so give them a chance. We all have war stories about these different vendors (and that includes DEC). There are always problems. Some products are GOOD, some BAD, some so-so. Some companies do better than others. I definitely have an Emulex Q-bus SMD disk controller that I have FOUL words for, but I have lots of their stuff I really LIKE and that serves us well. Even the one I don't like served well, just not as well as I had hoped. I assume they will continue to produce products I will want to and should buy, and unfortunately also some that hopefully someone warns me about first. Also remember that if DEC sues them, they must be wearing some kind of white hat. Without them imagine how DEC would tromp on us even more. I still remember an old rule: NEVER BUY an E..... TAPE controller. They may now have one I would like, but I use Dilog tape controllers. I will still listen if they say they have a better one. In the last 2 years I have seen a couple of truely better - you guessed it - new MOUSE TRAPS! (no, NOT the cruel torture GLUE ones). If you think Emulex if wrong about ECO policies, or shouldn't cheat on a what should be a balanced line driver by using a single ended driver and a resistor network, or should be a better Unibus citizen, you obviously will talk with your dollars and that DOES count, but sometimes faster is to try dear FRED COX (I think that is his name) letters. Maybe dust off and recycle some old dear KEN ones - often similar problems. It shortens the feedback time, and these little guys can reposition much faster than something the size of DEC. ================================================================================ Note 228.9 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 9 of 23 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 18 lines 16-NOV-1988 22:43 -< YOU MUST BE BLESSED BRUCE >- -------------------------------------------------------------------------------- > Of course not, this is Emulex, NOT DEC. Over the years I have had > a fair number of new prom sets from them, and can't ever remember > paying, but it has come close a few times. Well you and I have doing business with different Emulex Corp's. Before they were banned from any our new systems three years ago we have invested 2,475 in firmware upgrades. We required firmware upgrades on SC03, TC02, and TC05 boards - in all cases when they were out of warranty they charged us $225 per upgrade. Yes friends that is over 20 percent of the original board price. Futhermore they list all their firmware upgrades as optional ECO's and not Mandatory ECO's. The only firmware or ECO upgrades I have ever paid for have been from Emulex. I have never had to pay for any other upgrade because Emulex was the only company sleazy enough to list them as optional ECO's. The maintainance companies I do business with did not supply optional ECO's at that time. ================================================================================ Note 228.10 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 10 of 23 EISNER::DOHERTY "Bob Doherty" 21 lines 17-NOV-1988 10:41 -< We've had no problem >- -------------------------------------------------------------------------------- We have been using a patched DR driver for Eagles on an Emulex SC750 (emulating an expanded RM80) and the patches required are REALLY trivial. The only reason I can see to use the actual Emulex patches is to allow coexistence of Emulex and DEC Massbus boards on the same machine. The only thing we patch is the geometry table to give the correct geometry for the expanded RM80. This patched driver has been running satisfactorally for 2 months now and we've seen nothing untoward from it. I can understand an abstract avoidance of using patched drivers, however, simply patching the geometry table barely qualifies, IMHO. BTW, if the only thing that needs to to changed is the geometry table, it would probably be posssible to just change the values in the UCB of the drives involved, and then use the virgin driver. I have difficulty believing that, at this point, there is going to be a hardware incompatibility between the Emulex Massbus emulators and the DEC Massbus drivers. I can readily believe, however, that Emulex wants to get out of the business of supplying gratis patches to its, in many cases, former customers. It would be interesting to find out if that, is in fact, the case. ================================================================================ Note 228.11 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 11 of 23 EISNER::MCDOUGALL "Bob McDougall" 9 lines 25-NOV-1988 15:44 -< RM05 emulation, specifics please >- -------------------------------------------------------------------------------- > < Note 228.4 by EISNER::HASSINGER "Bob Hassinger" > > -< NEVER deal with Emulex! & How I run SC7000 under V5 >- > As it so happens my SC7000 IS functioning (as well as it ever did > anyway) under VMS V5 with a CDC 9771. Apparently the only reason for > this is that I had the wisdom back at the beginning to insist that the > emulation *I* got was of two real, standard size RM05s rather than the > customary one oversized drive. This sounds good, how do you go about setting it up? How get it to emulate two real, standard size RM05s? ================================================================================ Note 228.12 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 12 of 23 EISNER::HASSINGER "Bob Hassinger" 17 lines 25-NOV-1988 21:21 -< Buy it with that emulation ... >- -------------------------------------------------------------------------------- The SC7000 has a set of DIP switches that specify which of a number of emulations to use. The firmware on the board has the required information so the controller can do this. For the most part I think it is a matter of geometry, interleave factors and so on. On the Massbus side I don't know that there is much that needs changing for any of the emulations. The key was to insist that I wanted this particular geometry in my emulation when I bought the controller. They committed to supply it although at the time it was not one of the "standard" emulations in the SC7000. They failed to do so at first. When we complained they told us they had fired the guy who had promised it. Eventually they did come through with the modified PROMs. I have no way of checking but I have the impression they have gone through various additional ECOs to the PROM set so there may even be more emulations on the current version, (whatever it is). ================================================================================ Note 228.13 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 13 of 23 EISNER::MAGINNIS "Jim Maginnis" 15 lines 13-JAN-1989 05:23 -< Emulex SC31/Eagles, DECUS has source code >- -------------------------------------------------------------------------------- I have a 11/750 with an Emulex SC31/Eagles combination that I also have been informed is now longer being supported. There are also a couple of RA80's. I bought the 4.5 driver (UMxx). Sales says that several customers have purchased source code ($500.00) to attempt modifying the driver for VMS 5 themselves (though they do not maintain any customer lists). Tech support, however, said that the source code had "been donated to DECUS" (but could not provide any contacts, etc.). Both say that the only "supported" solution to to spend $1750.00 on an upgrade cost to a SC33. Any specific help ??? 228.10>> We have been using a patched DR driver for Eagles on an Emulex SC750 Might exchanging the SC31 for a SC750 be reasonable ??? Thanks. ================================================================================ Note 228.14 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 14 of 23 EISNER::KENNEDY "Terry Kennedy" 37 lines 13-JAN-1989 21:42 -< Doing battle with Emulex >- -------------------------------------------------------------------------------- > I have a 11/750 with an Emulex SC31/Eagles combination that I also have > been informed is now longer being supported. The Emulex 'Disk and Tape Products Handbook' (no part number, dated 1984), says (on Pages 80-81, under 'VAX Software Support') "All VMS/UX software packages have been developed and are supported in-house by the Emulex soft- ware support group. All elements of every package carry a 90-day design warranty plus update service for additions and revisions made, including any required by new releases of VMS software. Update service is available beyond 90 days at nominal cost." Note that this does not say "until we feel like not doing it any more". Given that the depreciation cycle was at least 5 years, and 1984+5=1989, you might want to make a legal issue of it, or at least *threaten* to make a legal issue of it. We had a similar SC31 issue with Emulex which was settled, but I can't discuss the terms of the settlement (neither can they). Try it - but you may have to give the impression you're prepared to play hardball. > Both say that the only > "supported" solution to to spend $1750.00 on an upgrade cost to a SC33. Do you mean UD33? If it was me, and I had to buy a new controller, I wouldn't throw good money after bad. By the way, the UD33 has a bug which will *prevent* it from working with future software releases. Emulex has ignored my comments on this. No, I can't say which release of what soft- ware, but I can say that we're talking a few years out. Do you want to be in the same boat again? > Might exchanging the SC31 for a SC750 be reasonable ??? I don't think so, but then again, I'm opinionated. Bearing that in mind, I say: MSCP is the only way to go when putting foreign disks with different geometries from DEC disks on the system. If your disk looks *exactly* like a DEC disk (RM05 vs CDC 9766) you *might* want to go with an emulating con- troller *if* you can get it at a steal. Note that the older controllers will not handle the transfer rates of the newer drives, so they are a dead end. ================================================================================ Note 228.15 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 15 of 23 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 68 lines 14-JAN-1989 19:01 -< consider Dilog's Dx256 MSCP family >- -------------------------------------------------------------------------------- > Do you mean UD33? If it was me, and I had to buy a new controller, I > wouldn't throw good money after bad. By the way, the UD33 has a bug which > will *prevent* it from working with future software releases. Also I believe (but am not positive) that the Emulex UD33 is guilty of using single ended vs balanced line driving on this controller as was perhaps too cryptically mentioned in the last paragraph of .8. At very least, you may not be able to run your cables as far as you might like. > I say: MSCP is the only way to go when putting foreign disks with different > geometries from DEC disks on the system. Good advice. We are now using the Dilog DQ256 and DU256 MSCP controllers, and really like them. They format packs identically, which we needed, and both do 4 physical drives and can partition a pack into 1 to 4 units. This partitioning and all other geometry info is written on the pack along with the unit numbers to use. On a drive with physical unit number 2, I could mount a pack that is just DUA7:, and later I can mount another pack that is DUA15: + DUA16: + DUA17: + DUA18:, (or that normally on the 11 would be DU17: + DU20: + DU21: + DU22:). If I didn't have an 11 running RSX-11M+, where each DU number means a UCB gen'd in that takes precious space from primary POOL, I would have assigned each of our 9766 removeable packs a UNIQUE hardcoded unit number (or a range of numbers for the partitioned packs). I doubt we EVER would exhaust the 0 - 255 range. This may sound wierd, and presents some strange logistic problems about what DU numbers get WRITTEN on packs, BUT also, once you have coped with all that, gives some features that are hard to beat. The 'hard-coded' unit numbers on the packs ARE easily changable, but only with an on-board stand-alone diagnostic. I can use the same 9766 packs on the Unibus 11, or on the Q bus MV2. A 1.23 gig CDC Saber, now seen as one 'big-mother' disk, is about to be partitioned into 4 smaller ones. And an Eagle+ is currently split into 2 disks. Since the drives are dual ported, I could even access some partitions from one CPU, and the rest from another, but we don't normally do that. The 9766 drives are simply accessable from either CPU and are for backup packs, and Decus disks, and whatever we need to take on/off line randomly. The developer of the new Vms BACkup recently spoke at a local LUG here, and swears that ODS-1 disks problems have been fixed, and that PDP-11 bootability will be preserved. So, when the NEW BACkup comes, I may stop using BRU on the 11, and simply BACkup the 11's 2 main disks (both halves of the EAGLE+) using VMS BACkup from the 2nd port. If you do decide to get one of these Dilog controllers, be sure to order it with the dual port PROMS, as they may otherwise ship the single port version. They are the same price, and there is NO reason not to use the DUAL port version in a single port environment, unless your local service company only stocks the other flavor. Dilog should only sell the DUAL version to eliminate confusion, but their product line manager wants to be protected from DUMB users trying to do SI type clustering. I havn't found anyone else in Dilog who agrees with him, and their east coast distributor (that is the Dec OEM we bought them from) is now only going to order/stock the dual port ones. ================================================================================ Note 228.16 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 16 of 23 EISNER::KILGALLEN "Larry Kilgallen" 2 lines 15-JAN-1989 10:32 -< Does this 9766 correspond to some DEC disk? >- -------------------------------------------------------------------------------- I was under the impression that a Fujitsu Eagle was a fixed platter drive. ================================================================================ Note 228.17 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 17 of 23 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 1 line 15-JAN-1989 11:16 -< THE CDC 9766 MECHANICS IS THE BASIS OF THE DEC RM05 >- -------------------------------------------------------------------------------- > -< Does this 9766 correspond to some DEC disk? >- ================================================================================ Note 228.18 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 18 of 23 EISNER::KENNEDY "Terry Kennedy" 1 line 15-JAN-1989 17:31 -< And conversion between 9766 and RM05 is possible... >- -------------------------------------------------------------------------------- > -< THE CDC 9766 MECHANICS IS THE BASIS OF THE DEC RM05 >- ================================================================================ Note 228.19 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 19 of 23 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 14 lines 16-JAN-1989 12:36 -< why a 9766 answer to an Eagle question >- -------------------------------------------------------------------------------- The Eagle, Eagle+, CDC-9766 (RM05), CDC 1.23gig Sabre, are all SMD (Storage Module Drive) interface type drives. What I was really saying was that any random collection of SMD drives and a flexible MSCP controller gives disks that run under VMS with NO special drivers. I was also trying to point out that Q and U versions exist that format the drives identically, and that other games such as partitioning a drive, and even simultaneous access to the same drive (preferably different partitions) from two controllers is possible. It is certainly more interesting to play with these options that to do battle with custom drivers and disk controllers emulating old old hardware. ================================================================================ Note 228.20 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 20 of 23 EISNER::MAGINNIS "Jim Maginnis" 14 lines 19-JAN-1989 13:41 -< More ... >- -------------------------------------------------------------------------------- 228.0> Emulex has just informed us that Version 5 of VMS does not support > the SC7000 and as a result we need to purchase a UNIBUS Controller. I have posted a notice about our problems concerning the EMULEX SC31. Although the University of Arizona (here in Tucson) was told by sales that their SC750 and SC780 would no longer be supported (just after paying the first $400.00/year for automatic upgrades), the next day they received the DRDRIVER for VMS 5.0. I was told with my conversations that the DRDRIVER would continue to be supported for some time due to the SC7003 still being sold by EMULEX. I have just received source code for the UMDRIVER for VMS 4, anyone else interested??? EMULEX indicated that the source code was to be made available to DECUS. Can anyone tell me how to do that??? ================================================================================ Note 228.21 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 21 of 23 EISNER::LEDERMAN "Bart Z. Lederman" 11 lines 19-JAN-1989 19:45 -< DECUS forms are everywhere. >- -------------------------------------------------------------------------------- > EMULEX indicated that the source code was to be made available > to DECUS. Can anyone tell me how to do that??? A: Try to get a copy of that indication in writing, signed by someone in EMULEX, preferably on a letterhead. B: Include that with a normal submission form, which you can find in symposia packets, at the back of the newsletter, in library catalogs, and various other DECUS publications. Or send mail to me or Bob Hassinger and one of us will mail you one (we're not about to let an offer like this get away). ================================================================================ Note 228.22 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 22 of 23 EISNER::MAGINNIS "Jim Maginnis" 3 lines 20-JAN-1989 13:39 -< SC31 UMDRIVER source... >- -------------------------------------------------------------------------------- Recreating the executable from UMDRIVER source produced a LINK error for a missing G^EXE$BUFFRQUOTA from the SYS.STB file, used in the diagnostics section to check the quota buffer size. ================================================================================ Note 228.23 Emulex Fujitsu Eagle(M2351) w/VMS V5.0 23 of 23 EISNER::KENNEDY "Terry Kennedy" 45 lines 22-JAN-1989 04:43 -< In the Library as VAX-386 >- -------------------------------------------------------------------------------- > EMULEX indicated that the source code was to be made available > to DECUS. Can anyone tell me how to do that??? Attached is the Program Library abstract for it. It was processed into the Library 1/19/89... VAX-386 UH/UM/UT Device Drivers for VMS Version: 4.0, December 1988 Submitted by: Emulex Corporation, Costa Mesa CA Operating System: VAX/VMS V4.X Source Language: MACRO-32 Memory Required: 10K through 15K Bytes per driver Special Hardware Required: Emulex CS11/U, CS21/U, SC21/V, SC31 or TC11 are required. The UH, UM, and UT drivers are VMS device drivers for UNIBUS DH11, RH11, and TM11 controllers. They were written for the Emulex CS11/U, CS21/U, SC21/V, SC31, and TC11 controllers, but will work with other Digital Equipment Corporation or emulating controllers with minor modifications. Emulex does not support these controllers under VAX/VMS V5.X or later. These drivers are unsupported by Emulex, and are provided as examples for customers that may wish to modify the code for their specific applications. Note: Operating system VAX/VMS V4.X is required. Documentation not available. Media (Service Charge Code): 600' Magnetic Tape (MA), Format: VMS/BACKUP, 160O BPI Submission Number: 02717 Keywords: Device Handlers Operating System Index: VAX/VMS  ================================================================================ Note 229.0 VMS5.0:TK50 Drive hand during BACKUP 4 replies EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 5 lines 15-NOV-1988 13:32 -------------------------------------------------------------------------------- Is anyone having problems writing multi-volume save sets to TK50s with VMS 5.0-2? Our backup procedure is hanging rather than mounting the next volume of the save set while we are attempting to do an image save of a RD54(clone) disk drive. We cannot even delete the process. Seton ================================================================================ Note 229.1 VMS5.0:TK50 Drive hand during BACKUP 1 of 4 EISNER::MCCORMICK_B "Bob McCormick" 11 lines 17-NOV-1988 08:01 -< What exactly is it doing? >- -------------------------------------------------------------------------------- We had a problem somewhat similar (the same?) as you describe when we went to V5.0 (and -1 & -2) ... Our canned backup procedure $ INIT the tape, and then did /IMAGE backups to it (mounted foreign). With VMS V5 the label processing caught us; on the second tape. When it got to the second tape (which is rotated, and already had a tape label) it got a bit upset since tape labels didn't match. In batch we had to use REPLY to tell backup to OVERWRITE the tape, or respond locally... What exactly is backup doing upon running into the need for the second tape? ================================================================================ Note 229.2 VMS5.0:TK50 Drive hand during BACKUP 2 of 4 EISNER::DOHERTY "Bob Doherty" 14 lines 17-NOV-1988 10:52 -< Sounds like the TS11 problem... >- -------------------------------------------------------------------------------- This (229.0) sound suspiciously like the problem we (and a suffcient number of others to get TSC's attention) have been having with TS11 emulators. The symptoms occur when verify is specified. The backup fails after verifying the first reel, since it doesn't wait for the second reel to be mounted. We thought that this was specific to the TS11 driver, but maybe it is, in fact, a Backup problem and not a driver problem. BTW, we have also noticed that there are holes in the automatic mount logic in Backup(confirmed by TSC). If a multi reel backup is done to more than one physical drive, although the automatic mount can be used for the first reel, the second reel MUST be explicitly mounted on the second physical drive, or the backup will fail at the start of the second reel. ================================================================================ Note 229.3 VMS5.0:TK50 Drive hand during BACKUP 3 of 4 EISNER::GRANT_D 22 lines 17-DEC-1988 16:18 -< version 5.0 *feature* >- -------------------------------------------------------------------------------- > Is anyone having problems writing multi-volume save sets to TK50s > with VMS 5.0-2? Our backup procedure is hanging rather than mounting > the next volume of the save set while we are attempting to do an > image save of a RD54(clone) disk drive. We cannot even delete the > process. I experienced a problem similar to this on my MVII. When I mounted the second tape, I would get a label processing error, so to avoid this, I added /IGNORE=LABEL_PROCESSING to the backup command. Our operator runs a command file to do the backup from OPA0. We have experienced a problem when backup returns an error, and asks a question, such as "RESTART or QUIT?" However, even thought you are prompted for input, you can't enter anything in answer to the prompt. The only way to get things running again is to do a REPLY from another terminal. The DEC people I have talked to about this problem treat this as a feature, not a bug. My question is, why should they even display a prompt and ask for input if they are not going to allow input? Has anyone else experienced this *feature*? darrell ================================================================================ Note 229.4 VMS5.0:TK50 Drive hand during BACKUP 4 of 4 EISNER::MERRIMAN "George Merriman - CCA/NY" 6 lines 17-DEC-1988 21:36 -< Let's keep this a hardware conference >- -------------------------------------------------------------------------------- This topic seems to be straying into the never-never land of software. Let's keep this conference for hardware questions and use the VMS conference for BACKUP questions. Thanks. George Merriman -- HARDWARE_HELP co-moderator  ================================================================================ Note 230.0 Current TK50/TQK50 rev levels No replies EISNER::MCCORMICK_B "Bob McCormick" 14 lines 18-NOV-1988 07:20 -------------------------------------------------------------------------------- We've had some various strangies with our TK50 subsystem on a MicroVAX II system. This includes problems with mounts and dismounts (we're making distribution kits; ie lotsa use...) and the fantom MUA4224: device (which I logged a call and SPR'ed the first week I had a TK50 _years_ ago :-{ ) Our local DEC Field Service rep has new info regarding the rev levels of both the TK50 drive and controller. Try these out: Drive TK50 Rev E1 or later Control TQK50 M7546 Rev J1 or later (look for 23-331E5 in E28) These seem to help.  ================================================================================ Note 231.0 Need Floppy Cntrl that will support > 2 drives 2 replies EISNER::LOBMEYER "Fred Lobmeyer City of NY" 14 lines 20-NOV-1988 17:02 -------------------------------------------------------------------------------- I am trying to track down information on hooking a minimum of three floppy drives simultaneously to an IBM PC or compatible. I am trying to find a solution that either allows this on one floppy controller or multiple where I don't need to worry about the interrupts or have to change software. The three types of floppies are as follows: 1 1.2MB 5 1/4" 1 360K 5 1/4" 1 1.4MB 3.5" Thanks in advance, Fred ================================================================================ Note 231.1 Need Floppy Cntrl that will support > 2 drives 1 of 2 EISNER::KENNEDY "Terry Kennedy" 28 lines 20-NOV-1988 21:11 -< Possible solution,cautions >- -------------------------------------------------------------------------------- > I am trying to track down information on hooking a minimum of three > floppy drives simultaneously to an IBM PC or compatible. > 1 1.2MB 5 1/4" > 1 360K 5 1/4" > 1 1.4MB 3.5" Do you really mean PC as in old PC or XT, or in the general term to describe an AT-class machine? If you meant old PC or XT, a board called a 'Compaticard' will do what you want. It replaces your current floppy controller (a pain if it is part of a multiple-function I/O card). It supports all the above formats and then some, but requires a device driver (which thy supply). I'll go hunt for the address. On an AT, things are harder. First, the existing hard disk controller is part of the hard disk controller, so it's harder to lose. Next, the configuration memory only has space for defining two drives. However, you could use the Compaticard again, this time as a second controller card. I'd suggest putting the 1.4 drive on it, for reasons given below. There is a catch in installing 3.5" 1.44Mb drives on AT's. 1.44Mb drives use pin 2 of the connector to select between 1.44Mb mode and 720Kb mode. Un- fortunately, almost all AT-style controllers do not provide the proper sig- nal on that pin. There are two solutions, however. First, you can order the 1.44Mb drive with the optional media type switch to automatically detect the high-density media. Second, you could cut pin 2 and install a switch on the front of the drive to select the density. ================================================================================ Note 231.2 Need Floppy Cntrl that will support > 2 drives 2 of 2 EISNER::KENNEDY "Terry Kennedy" 0 lines 20-NOV-1988 21:17 -< Vendor's address posted in note 78.3 >- --------------------------------------------------------------------------------  ================================================================================ Note 232.0 DWBUA rev levels? 3 replies EISNER::BRYANT "Geoff Bryant" 2 lines 22-NOV-1988 09:32 -------------------------------------------------------------------------------- Does anyone know the difference between rev F7 and rev H2 of the BI-UNIBUS adapter DWBUA for a VAX 8550? ================================================================================ Note 232.1 DWBUA rev levels? 1 of 3 EISNER::KENNEDY "Terry Kennedy" 9 lines 22-NOV-1988 20:20 -< Answer >- -------------------------------------------------------------------------------- > Does anyone know the difference between rev F7 and rev H2 of the > BI-UNIBUS adapter DWBUA for a VAX 8550? Rev H fixes a problem where a powerfail restart would bugcheck the system. The FCO kit is EQ-01509-01, implementing FCO DWBUA-I002. The quick check is part 23-329F4-00 at location E67. Note - DEC continues a great tradition by labelling this FCO an 'Improve- ment'! ================================================================================ Note 232.2 DWBUA rev levels? 2 of 3 EISNER::BRYANT "Geoff Bryant" 7 lines 22-NOV-1988 21:24 -< Thanks, how 'bout rev E >- -------------------------------------------------------------------------------- Thanks for the info Terry! Since you seem so well informed, do you know the difference then between rev F7 and E? ? Also, if the machine isn't going through powerfails, is there any other benefit to rev H? ================================================================================ Note 232.3 DWBUA rev levels? 3 of 3 EISNER::KENNEDY "Terry Kennedy" 14 lines 23-NOV-1988 02:43 -< Gosh, I don't even have one of these things... >- -------------------------------------------------------------------------------- > Since you seem so well informed, do you know the difference then > between rev F7 and E? ? o It got passive releases wrong (where a device requested the bus but 'went away' before the CPU came back with an acknowlegement. o Drops address bit 0 on DATO(B) after doing a purge of a buffered datapath. > Also, if the machine isn't going through powerfails, is there any > other benefit to rev H? One wouldn't think so, but the FCO orders the retrofitting of all spares, so who knows...  ================================================================================ Note 233.0 CDROM on non-Qbus systems? 38 replies EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 7 lines 22-NOV-1988 12:35 -------------------------------------------------------------------------------- We would like to install one of these here CD-ROM thingies, but have no Qbus equipment and no plans to acquire any. Can anyone comment favorably on one or another of the Unibus-to-Qbus converters (we have an 11/780 and an 8650, with all Massbus peripherals). Or, has anyone tried using the console subsystems (11/03 in the 780 and 11/24 in the 8650) to handle Qbus devices? ================================================================================ Note 233.1 CDROM on non-Qbus systems? 1 of 38 EISNER::HAHN "DECUServe Vice Chair" 10 lines 22-NOV-1988 13:25 -< Wait a while or use a PC >- -------------------------------------------------------------------------------- From conversation which I had at the Fall88 symp. I would venture to guess that there will be another interface available for the CD-ROM reader before the Spring89 symp that would allow you to use it on all of the above. If you cannot wait I would suggest a PC (IBM/CLONE) with the appropriate software for it to be accessible via E-net. This gives you not only the CD-ROM but allows your users to have a PC. Pierre ================================================================================ Note 233.2 CDROM on non-Qbus systems? 2 of 38 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 12 lines 22-NOV-1988 14:07 -< Cluster CD-ROM controller expected, not Unibus. >- -------------------------------------------------------------------------------- >> From conversation which I had at the Fall88 symp. I would venture to >> guess that there will be another interface available for the CD-ROM >> reader before the Spring89 symp that would allow you to use it >> on all of the above. I wasn't at Anaheim but someone else from our facility was, and he heard that a Unibus controller had definitely been ruled out, and that a cluster (HSC-?? UDA-??) controller was in the works; which is an excellent idea but won't do us any good. PC is also under consideration, but we have never had any luck in moving VMS binary files between VAX and PC environments over DECnet-DOS (MS-DOS binary files move OK but I suspect there aren't many on the VMS SIG tapes). ================================================================================ Note 233.3 CDROM on non-Qbus systems? 3 of 38 EISNER::TINKELMAN "Bob Tinkelman - CCA" 7 lines 22-NOV-1988 23:23 -< A $20K CI interface on a $1K CD-ROM drive? >- -------------------------------------------------------------------------------- > ... he heard > ... that a cluster (HSC-?? UDA-??) controller was in the works; Well, I haven't been party to any relevant nondisclosure agreements. So I can officially say that I do not officially know any DEC secrets in this area. But don't you think that an Ethernet interface would be a lot cheaper and therefore more likely than a CI bus interface? ================================================================================ Note 233.4 CDROM on non-Qbus systems? 4 of 38 EISNER::HASSINGER "Bob Hassinger" 2 lines 23-NOV-1988 11:52 -< Yes, and more logical too! But rumors so far say no. >- -------------------------------------------------------------------------------- > But don't you think that an Ethernet interface would be a > lot cheaper and therefore more likely than a CI bus interface? ================================================================================ Note 233.5 CDROM on non-Qbus systems? 5 of 38 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 4 lines 23-NOV-1988 15:57 -< CDROM MSCP interface, not CI. >- -------------------------------------------------------------------------------- >> But don't you think that an Ethernet interface would be a >> lot cheaper and therefore more likely than a CI bus interface? I thought I understood that the idea was an MSCP interface, not CI. ================================================================================ Note 233.6 CDROM on non-Qbus systems? 6 of 38 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 16 lines 23-NOV-1988 19:22 -< PROTOCOL VS INTERFACE >- -------------------------------------------------------------------------------- > I thought I understood that the idea was an MSCP interface, not CI. MSCP is a protocol and not an interface. DEC sells fours type of disk drives interfaces. SDI interface - UDA50, KDA50, ect. type drive to controller interface KESLI interface - RC25 type drive to controller interface Modified ST506 interface - RD51-RD54, CDROM, ect. type drive to controller interface DSSI interface - the new RF30 type drive to controller interface NOTE: in all four cases the controller talk MSCP. ================================================================================ Note 233.7 CDROM on non-Qbus systems? 7 of 38 EISNER::MCCORMICK_B "Bob McCormick" 9 lines 24-NOV-1988 21:48 -< DSSI >- -------------------------------------------------------------------------------- I'd bet you would see a DSSI CD device some time in the future... The question is how will they connect to things like 62XX that don't have the HSC configuration?! As for the DSSI "storage elements": each storage element has a dedicated controller "...an intelligent controller and a Mass Storage Control Protocol (MSCP) server..." I believe they term the part that supports DSSI on the host CPU module (KA640) an adaptor; so this looks more like an HSC than a RQDXn, KDx50, UDA50, etc. ================================================================================ Note 233.8 CDROM on non-Qbus systems? 8 of 38 EISNER::KENNEDY "Terry Kennedy" 23 lines 24-NOV-1988 22:58 -< Just because it's sensible... >- -------------------------------------------------------------------------------- > I'd bet you would see a DSSI CD device some time in the future... > The question is how will they connect to things like 62XX that don't > have the HSC configuration?! Well, I still think the *sensible* thing to do is an Ethernet interface, since: o It's the least expensive of all methods discussed so far o It is the *only* interface supported on *all* VAXen o It makes it easier to share the device Of course, just because it's sensible doesn't mean they'll do it... > I believe they term the part > that supports DSSI on the host CPU module (KA640) an adaptor; > so this looks more like an HSC than a RQDXn, KDx50, UDA50, etc. Actually, it's more like a KLESI. In fact, with suitable programming and a jumper cable, you could make the DSSI adaptor talk to RC25's and TU81's (not to mention TK50-Zs). Sadly, the reverse (RFxx on KLESI) won't work. ================================================================================ Note 233.10 CDROM on non-Qbus systems? 10 of 38 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 4 lines 25-NOV-1988 12:47 -< Oops. Meant CDROM to DSI. >- -------------------------------------------------------------------------------- >> MSCP is a protocol and not an interface. ... Oops. Meant SDI, to allow connection to xDx50 and HSC5X cards. ================================================================================ Note 233.11 CDROM on non-Qbus systems? 11 of 38 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 3 lines 25-NOV-1988 12:59 -< IT IS TIME FOR TECHNOLOGY CHANGE OVER >- -------------------------------------------------------------------------------- > Oops. Meant SDI, to allow connection to xDx50 and HSC5X cards. My guess (for what is is worth) is you have seen the last SDI disk. ================================================================================ Note 233.12 CDROM on non-Qbus systems? 12 of 38 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 11 lines 25-NOV-1988 14:41 -< Wide usage of CDROM needs other controllers? >- -------------------------------------------------------------------------------- >> My guess (for what is is worth) is you have seen the last SDI disk. Mebbe, but there's still a LOT of xDx50 and HSC5X equipment out there that will continue in use for a while (there's a lot of Unibus, and even Massbus, out there also...) and it would seem that if the intent is to get CDROM readers into user's hands as widely as possible, the object should be to make them usable on as much of the **presently- installed** equipment as possible. Maybe I am missing something, though, like maybe a long-range program to discontinue paper documentation and force everyone who needs documentation (most of us, I guess) to upgrade to a newer computer... ================================================================================ Note 233.13 CDROM on non-Qbus systems? 13 of 38 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 7 lines 25-NOV-1988 23:56 -< CDROM must use something other than ST-506 >- -------------------------------------------------------------------------------- >> Modified ST506 interface - RD51-RD54, CDROM, ect. type drive to >> controller interface I don't believe that the CDROM uses an ST506-type interface. The cable doesn't have nearly enough wires. Alan ================================================================================ Note 233.14 CDROM on non-Qbus systems? 14 of 38 EISNER::MERRIMAN "George Merriman - CCA/NY" 11 lines 28-NOV-1988 20:26 -< Is Ethernet cheapest for small configurations? >- -------------------------------------------------------------------------------- > Well, I still think the *sensible* thing to do is an Ethernet > interface, since: > > It's the least expensive of all methods discussed so far I'm not convinced of this. It seems to me an Ethernet-based CD reader would need a lot of on-board smarts, compared to the other suggested solutions. It would depend on whether or not you could hang the CD reader off an existing controller, etc. and the sharing possibilities which you point out would also have to be considered. ================================================================================ Note 233.15 CDROM on non-Qbus systems? 15 of 38 EISNER::KENNEDY "Terry Kennedy" 22 lines 29-NOV-1988 19:23 -< Not so expensive >- -------------------------------------------------------------------------------- > I'm not convinced of this. It seems to me an Ethernet-based CD reader > would need a lot of on-board smarts, compared to the other suggested > solutions. The Ethernet chipset from various silicon foundries costs about $20 to speak thinwire. Complete IBM PC Ethernet cards, which can be looked at as buffered parallel-interface to Ethernet adaptors, cost about $189 qty. 1 retail (Everex). > It would depend on whether or not you could hang the CD > reader off an existing controller, etc. Well, let's see. On VAXen, MicroVAX I's and early II's didn't have Ethernet, but they're Q-bus, so they're covered. All busless MicroVAXen have Ethernet. It's darn hard to buy a current 8000-series VAX without Ethernet. Ditto 6000. New 3000's have it on the CPU. That leaves mainly 7xx and 86xx machines. Used DEUNAs bring between $200 and 400 if you're selling, and sell for about $1000 if you're buying. Thus, the worst case is a user with an old VAX and no current network. Then again, that customer profile doesn't sound like it wants a CD-ROM reader, either. ================================================================================ Note 233.16 CDROM on non-Qbus systems? 16 of 38 EISNER::STRIEGEL "Alan Striegel" 17 lines 29-NOV-1988 22:07 -< Using RRD50 on PC to VAX now >- -------------------------------------------------------------------------------- I have the RRD50 CD reader attached to my VAXmate and have pulled some of the VAX programs off the SIG library disks with the software Digital supplies (CDCOPY). I can copy files to a VAX's disk space where I am running VMS Services for MS-DOS so the PC sees it as a file serer (not disk server). There have been some problems. For common ASCII text files, every several lines will be polluted with test characters and control characters. I have been able to eliminate some of these problems with a short DOS program I wrote to strip off the junk. Contiguous files have been somewhat better. Some Backup savesets have been delivered intact so that I can use VMS Backup to extract the contents for the VAX. Some .EXE files have come across just fine, but the majority have not. I can't be sure, but I think DOS has problems handling the structure of files with their raw Files-11 structure. Alan ================================================================================ Note 233.17 CDROM on non-Qbus systems? 17 of 38 EISNER::MERRIMAN "George Merriman - CCA/NY" 18 lines 30-NOV-1988 17:00 -< CD file server needs more than Ethernet interface >- -------------------------------------------------------------------------------- >> I'm not convinced of this. It seems to me an Ethernet-based CD reader >> would need a lot of on-board smarts, compared to the other suggested >> solutions. > The Ethernet chipset from various silicon foundries costs about $20 to > speak thinwire. Complete IBM PC Ethernet cards, which can be looked at > as buffered parallel-interface to Ethernet adaptors, cost about $189 > qty. 1 retail (Everex). I don't doubt that Ethernet interface hardware is cheap, but I suspect you would need rather more than the Ethernet interface and the CD drive to make a network-based CD file server. The CD file server would have to look like some kind of node or something on the network so that the computers could find it. Some kind of file server protocol would have to be devised to run on the Ethernet and the CD file server would need a processor to run this protocol. I suspect all this would end up being about as complex as a terminal server. ================================================================================ Note 233.18 CDROM on non-Qbus systems? 18 of 38 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 8 lines 30-NOV-1988 17:31 -< Agree, VMS binary files from MS-DOS are trash. >- -------------------------------------------------------------------------------- >> Re .16 >> >> Some .EXE files have come across just fine, but the majority have not. >> This has been our problem also; whenever we get any kind of VMS binary file on MS-DOS, we are *very* seldom able to put it into VMS and have it behave properly. ================================================================================ Note 233.19 CDROM on non-Qbus systems? 19 of 38 EISNER::KILGALLEN "Larry Kilgallen" 12 lines 30-NOV-1988 18:14 -< DEC already *has* a chip to read CDROMs >- -------------------------------------------------------------------------------- The KA620 cpu (unless I have the number wrong) which DEC already sells at a price considerably below that of VMS-capable CPUs is able to run VAXELN. Using VAXELN one is able to read ODS-2, which is the format of current DEC CDROM disks. The incremental cost to DEC for using KA620's as custom "controller" chips is minimal, and the risk of some third party stripping the chips out is minimal (no incentive, it won't run VMS). I don't know what hardware it takes to connect the CDROM drive to the KA620 chip, but the KA620-Ethernet connection has already been designed. ================================================================================ Note 233.20 CDROM on non-Qbus systems? 20 of 38 EISNER::HARVEY "Jack Harvey" 11 lines 30-NOV-1988 18:50 -< CDROM as DECnet Node >- -------------------------------------------------------------------------------- >The CD file server would have to look like some kind of node or >something on the network so that the computers could find it. I agree with George (.17) - the problem is non-trivial. There must be some way to deal with multiple hosts on the Ethernet trying to use it at the same time. One answer might be to make it look like a DECnet node with all the smarts. You would access the CDROM files as though they were on another real DECnet node. The CDROM would also be SYS$SYSDEVICE with VAXELN. Larry (.19) has the processor. ================================================================================ Note 233.21 CDROM on non-Qbus systems? 21 of 38 EISNER::KENNEDY "Terry Kennedy" 21 lines 1-DEC-1988 02:44 -< You're missing the DEC viewpoint >- -------------------------------------------------------------------------------- > I agree with George (.17) - the problem is non-trivial. There must be > some way to deal with multiple hosts on the Ethernet trying to use it > at the same time. You're looking at it from a non-DEC viewpoint, considering backwards compatibility and all the special cases they don't think of... I think that if they were to do it, they'd: o Implement a new protocol, not use DECnet o Require clustering for multiple systems wishing concurrent access o bundle the software into a new point release of VMS o no support for CPU's other than VAXen If the above were done, the only brains required in the drive would be the ability to recieve a request, position to the desired block, and return the information. [Funny, I seem to recall the drive does that right now...] VMS already knows how to read ODS-2 as well, and that functionality does not need to be distributed to the drive. In fact, doesn't the MSCP server pass references as block numbers, not as ODS-2 file data? ================================================================================ Note 233.22 CDROM on non-Qbus systems? 22 of 38 EISNER::HARVEY "Jack Harvey" 10 lines 1-DEC-1988 19:02 -< LAD - Local Area Disks >- -------------------------------------------------------------------------------- Terry's points are excellent and he may be close to the mark. If the new protocol allows writes as well as reads, it would also be a simple matter to substitute a real disk for the CDROM. DEC could then easily hang more disks on MV2000s or PVAXen, or even Tandy PCs. Then Emulex, System Industries, et al, can reverse engineer the new protocol as was done with LAT and give us all the disks we want on Ethernet. The non-Qbus CDROM might be the start of a new sub-industry. ================================================================================ Note 233.23 CDROM on non-Qbus systems? 23 of 38 EISNER::ERSKINE "Chris Erskine" 4 lines 8-DEC-1988 07:19 -< Local Area File Transport (LAFT)? >- -------------------------------------------------------------------------------- I would bet that DEC is working on a LAT type of protocol for disk drives. This is not DECnet but a disk optimized protocol which would do the above. ================================================================================ Note 233.24 CDROM on non-Qbus systems? 24 of 38 EISNER::HARVEY "Jack Harvey" 9 lines 9-DEC-1988 12:20 -< LAFT/LAST >- -------------------------------------------------------------------------------- I have had many requests (demands!) for a UVAX 2000 with a nine track magtape drive. If DEC expands the new protocol to include conventional tape drives, we can call it Local Area Sequential Transfer (LAST) Seriously, though, what other devices besides disks and tapes might be usefully hung on an Ethernet a la DECservers? ================================================================================ Note 233.25 CDROM on non-Qbus systems? 25 of 38 EISNER::MERRIMAN "George Merriman - CCA/NY" 4 lines 9-DEC-1988 17:03 -< another Ethernet server suggestion >- -------------------------------------------------------------------------------- > Seriously, though, what other devices besides disks and tapes might be > usefully hung on an Ethernet a la DECservers? How about a DECWindows server terminal? ================================================================================ Note 233.26 CDROM on non-Qbus systems? 26 of 38 EISNER::LEDERMAN "Bart Z. Lederman" 18 lines 9-DEC-1988 19:26 -< Printers, etc. >- -------------------------------------------------------------------------------- Well, there are already some gateways (X.25 PAD and SNA Gateway) which hang on the Ethernet: I suppose that could be expanded. I gather that a lot of people would like to put their high speed printers on an Ethernet instead of having them hooked to one particular CPU and depend on cluster Queues and having that particular CPU up and running when you want to print. A serial-to-parallel converter on a DECserver might work for some applications, but not all. How about hooking up your fire and power and water alarms to the Ethernet and have them broadcast warnings when they go off: more than one CPU (or node) would know what was going on. It would make fail-safe warning systems easier to build. If you were going to have a disk server which allowed you to send and receive blocks of data, and extended to stuff like floppies and tapes, then it could just as easily be any data peripheral (paper tape? Magnetic card reader? IEEE-488 Bus converter?) ================================================================================ Note 233.27 CDROM on non-Qbus systems? 27 of 38 EISNER::MAYHEW "Bill Mayhew" 11 lines 9-DEC-1988 20:54 -< Digital had LAST First >- -------------------------------------------------------------------------------- < Note 233.24 by EISNER::HARVEY "Jack Harvey" > > If DEC expands the new protocol > to include conventional tape drives, we can call it > Local Area Sequential Transfer (LAST) Sorry, that nice name is already taken. I think the translation is Local Area Storage Transport... it's the protocol used to implement "virtual disks" (as opposed to virtual files) in DEC's Personal Computer Systems Architecture. ================================================================================ Note 233.28 CDROM on non-Qbus systems? 28 of 38 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 155 lines 23-JAN-1989 10:29 -< RRD40 - NOTE MARKED SECTIONS RE: THIS TOPIC >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. 09-Jan-89 10:42 PM Page: 1 ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! THE RRD40 - 09-January-1989 ******************************************************************************* o Now supports VAXstation 3100, MicroVAX II and MicroVAX 3500/3600 Systems o Replaces the RRD50 o Provides higher performance, higher reliability and a convenient front-load design o Available March 1989 PRODUCT DESCRIPTION The RRD40, is a high-performance compact disk drive for customers using CD distributed products. The drive has an average seek time of less than 500ms and a transfer rate of 150 kb/s. Its capacity of 600 MB makes it ideal for holding a large number of software applications, stored on a single, compact disk. Compact disk is the primary distribution device for VMS, documentation and other software layered products. Users can now update their entire VMS operating system by loading a single disk. Searches through the VMS documentation set will take seconds instead of minutes, and customers will save more than 20 feet of shelf space currently devoted to their documentation manuals. The SCSI RRD40 comes in both a half-height form and a tabletop enclosure. Compact disks are front loaded into the drive using a protective caddie. Customers using multiple compact disks with multiple applications can daisy- chain up to four drives together on their VAXstation 3100 system. The Q-bus RRD40 comes in a tabletop enclosure. It is sold with a controller that fits into the backplane of either a MicroVAX II or MicroVAX 3500/3600 system. Customers can fit two drives on each controller. > Both SCSI and Q-bus versions of the RRD40 can be accessed by other > systems through the Ethernet. PRODUCT POSITIONING The RRD40 has been designed to meet Digital standards for SCSI devices. These meet or exceed industry standards for SCSI. The RRD40 will replace Digital's existing compact disk drive, the RRD50. Compared to the RRD50, it is a higher performance drive. Its seek time is three times faster. In addition, it offers higher reliability and a more compact, convenient front-load design. Current owners of RRD50 can use the RRD40 controller. FEATURES/BENEFITS Feature Benefit . License Management Facility allows . Easy customer purchase and customers to access layered installation of VMS layered products on their compact disks products using a single code . More than 20 feet of VMS . Significant space savings documentation on one disk . Compact disks are easier to handle . Convenience of a compact disk than tapes, require less space, and retain their data integrity under a variety of environmental conditions > . Network accessible . Easy to distribute software > to remote locations over the > network . Use of compact disk for software . More frequent and predictable distribution software updates . The RRD40 is 3 times faster than . Quick availability of software the TK50 at downloading software . Increased user productivity . Extremely low seek error rate . High data integrity (10^-16) resulting in highly accurate data retrieval. Disk protected at all times by caddie . Up to four drives can be used at . Multiple applications can one time with no performance be run at one time degradation PRICING/ORDERING INFORMATION Order No. Description U.S. List Price For VAXstation 3100 Model 40: RRD40-EF Field installable, embedded SCSI drive $1,350 RRD40-EG Factory installed, embedded SCSI drive 1,350 For VAXstation 3100 Model 30: RRD40-FA Tabletop SCSI drive, connecting cables, 1,600 110 V power cord RRD40-F3 Tabletop SCSI drive, connecting cables, * no power cord, for Europe and GIA only For MicroVAX II Systems: ALL OPTIONS $2,500 RRD40-HA RRD40-DC (RRD40 compact disk drive with Q-bus interface, desktop enclosure, test disk, spare loading caddie, 120 V/240 V switch), KRQ50-AA, documentation, BC18R-12, factory installed, 120 V power cord. RRD40-HB Same as RRD40-HA without power cord. RRD40-HC RRD40-DC (RRD40 compact disk drive with Q-bus interface, desktop enclosure, test disk, loading caddie, 120 V/240 V switch), KRQ50-AA, documentation, BC18R-12, field installable, 120 V power cord. For MicroVAX 3500/3600 Systems: ALL OPTIONS $2,500 RRD40-SA RRD40-DC (RRD40 compact disk drive with Q-bus interface for BA213 servers, desktop enclosure, test disk, spare loading caddie, 120 V/240 V switch), KRQ50-SA, documentation, BC16V-12, factory installed, 120 V power cord. RRD40-SB Same as RRD40-SA without power cord. RRD40-SF RRD40-DC (RRD40 compact disk drive with Q-bus interface for BA213 servers, desktop enclosure, test disk, spare loading caddie, 120 V/240 V switch), KRQ50-SF, documentation, BC16V-12, field installable, 120 V power cord. RRD40-SG Sames as RRD40-SF without power cord. SERVICE The RRD40 includes a one-year Return-to-Digital Warranty. Further service is available for customers who require on-site material and labor coverage for the product. This flexibility is the result of Digital's recently announced Expanded Warranty Support Program. AVAILABILITY Available in March 1989 as part of the VAXstation 3100. ================================================================================ Note 233.29 CDROM on non-Qbus systems? 29 of 38 EISNER::MAYHEW "Bill Mayhew" 8 lines 23-JAN-1989 17:16 -< "accessible" doesn't cut the mustard >- -------------------------------------------------------------------------------- Ah, but "network accessible" does not necessarily mean "network attachable". "Network accessible", I believe, simply means that a SCSI RRD40 on a VS3100, or a Q-bus RRD40 on a Q-bus host, can be accessed over the network (unlike, say, a tape drive). It doesn't say anything about the ability to attach the RRD40 in any free-standing way to the Ethernet... which I think is what many of us would like to see! ================================================================================ Note 233.30 CDROM on non-Qbus systems? 30 of 38 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 10 lines 23-JAN-1989 21:14 -< THINK OF THE 3100 AS AN INTERFACE BOX >- -------------------------------------------------------------------------------- > "Network accessible", I believe, simply means that a SCSI RRD40 > on a VS3100, or a Q-bus RRD40 on a Q-bus host, can be accessed over > the network (unlike, say, a tape drive). It doesn't say anything > about the ability to attach the RRD40 in any free-standing way to > the Ethernet... which I think is what many of us would like to see! If you look at the cost of base 3100 model 30 with one of these attached I don't see how you could drive the price down much more (Given the way DEC does hardware markups). ================================================================================ Note 233.31 CDROM on non-Qbus systems? 31 of 38 EISNER::NAGY "Frank J. Nagy, VAX Guru" 2 lines 24-JAN-1989 07:55 -< RRD40 media == RRD50 media? >- -------------------------------------------------------------------------------- Does the RRD40 use the same media as the RRD50 (since we have 4 of the later, this is an interesting question to us)? ================================================================================ Note 233.32 CDROM on non-Qbus systems? 32 of 38 EISNER::KILGALLEN "Larry Kilgallen" 4 lines 24-JAN-1989 09:33 -< Effectively the same media >- -------------------------------------------------------------------------------- The disk itself is the same, but there is a plastic carrier you insert it in to make front-loading possible. The design goal I believe is that all of your commonly used disks would be in carriers, even when they are on the shelf. ================================================================================ Note 233.33 CDROM on non-Qbus systems? 33 of 38 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 11 lines 24-JAN-1989 15:51 -< CD-ROM on non-Qbus and non-SCSI systems? >- -------------------------------------------------------------------------------- To update my original inquiry: We would like to install one of these here CD-ROM thingies, but have no Qbus nor SCSI equipment and no plans to get any. Can anyone comment favorably on one or another of the Unibus-to-Qbus converters (we have an 11/780 and an 8650, with all Massbus peripherals). Or, has anyone tried using the console subsystems (11/03 in the 780 and 11/24 in the 8650) to handle Qbus devices? (We have never been able to move VMS binary files successfully from MS-DOS to VMS; the PC version of the CD-ROM is of limited use.) ================================================================================ Note 233.34 CDROM on non-Qbus systems? 34 of 38 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 22 lines 24-JAN-1989 18:17 -< SCSI controllers should come >- -------------------------------------------------------------------------------- > We would like to install one of these here CD-ROM thingies, > but have no Qbus nor SCSI equipment and no plans to get any. > Can anyone comment favorably on one or another of the > Unibus-to-Qbus converters (we have an 11/780 and an 8650, I would not be suprised to see U->SCSI, and BI->SCSI adapters. Or maybe there will be a way to hang all the new SCSI stuff, CD readers included, on an HSCx. The RRD40 blurb said RRD50 disks would work on the new Q controller, but said nothing about RRD40s on the old Q controller. I think DEC is using an LMS (Laser Magnetic Storage - a joint Philips/CDC venture) CD drive. The Philips CD cartridge one will now keep and use the disks in is unique to Philips. The rest of the industry has standardised on Sony's design. I am itching to find 3rd party drives so I can get SEVERAL (2 or 3, probably) for the price of 1 of DEC's. Hitachi's new SCSI #3650 isn't due out til 2nd Q (calender, NOT DEC's kind of Qs). Does anyone know a good drive that will run on DEC's SCSI controller that is available now? ================================================================================ Note 233.35 CDROM on non-Qbus systems? 35 of 38 EISNER::LEDERMAN "Bart Z. Lederman" 2 lines 24-JAN-1989 20:05 -< Newsletter says RRD40 disks = RRD50 disks >- -------------------------------------------------------------------------------- Digital News says in it's latest issue that the two CD-ROM drives read the same disks. ================================================================================ Note 233.36 CDROM on non-Qbus systems? 36 of 38 EISNER::TERRAZAS "Mike Terrazas" 5 lines 25-JAN-1989 09:39 -< BI or Unibus? >- -------------------------------------------------------------------------------- Yesterday I asked how I could put an RRD40 on my 8200, and my DECrep 200 said, "We can work that out." I'm not sure what that means, but he says he'll have me a quote before the end of the month. MikeT ================================================================================ Note 233.37 CDROM on non-Qbus systems? 37 of 38 EISNER::TINKELMAN "Bob Tinkelman - CCA" 12 lines 25-JAN-1989 14:53 -< RRD40 = LMS CM200 (sort of) >- -------------------------------------------------------------------------------- RE: < Note 233.34 by EISNER::BRUCE "Barton F. Bruce - CCA/MA" > > I think DEC is using an LMS (Laser Magnetic Storage - a joint Philips/CDC > venture) CD drive. According to the 1/23/89 issue of Digital News (allright, take it for what it's worth) on p15, "According to sources close to the project, it [the RRD40] is an OEM version of the CM200 CD-ROM from Laser Magnetic Storage International, Inc. of Colorado Springs, Colo., modified with firmware that incorporates diagnostics and interface protocols for Digital systems." ================================================================================ Note 233.38 CDROM on non-Qbus systems? 38 of 38 EISNER::RICE "Gary Rice" 9 lines 28-JAN-1989 09:47 -------------------------------------------------------------------------------- > Yesterday I asked how I could put an RRD40 on my 8200, and my DECrep 200 > said, "We can work that out." I'm not sure what that means, but he says > he'll have me a quote before the end of the month. Mine said the same thing, and actually delivered the quote. However, it didn't have a Unibus version on it. Querying him about it produced a "Sorry" message. Gary  ================================================================================ Note 234.0 Ethernet Printer Servers? 2 replies EISNER::ZIMMERMAN "John Zimmerman, Southern CT Newsp" 7 lines 25-NOV-1988 12:42 -------------------------------------------------------------------------------- Does anyone know whether DEC has plans to introduce a parallel based printer interface in the form of an ethernet server My interest in this should be obvious for all of you with clusters. How does one run a LP27 (or similar printer) in a cluster while maintining the redundant characteristics which presently exist for tapes, disks, and terminals? ================================================================================ Note 234.1 Ethernet Printer Servers? 1 of 2 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 25 lines 25-NOV-1988 16:53 -< parallel printer on DS200 >- -------------------------------------------------------------------------------- >>Does anyone know whether DEC has plans to introduce a parallel based >>printer interface in the form of an ethernet server There are various enhancements to the DS500 they must be contemplating. In Anaheim a developer asked if we liked the idea of having a spare terminal card preinstalled and being software swappable for a broken one since hot swapping Q cards is out. He also indicated that there were lots of things yet to come for the DS500 (now DS550). Since Emulex's terminal server has a parallel port, Dec has a bit more reason to produce such. >>How does one run a LP27 (or similar printer) ... For now, the easy way is to buy anyones serial to Dataproducts parallel adapter. They are only a few hundred dollars. The Black Box one our NY office uses to run a (yuk) LP05 can be used for Dataproducts or Centronics type interfaces, and I think can do flow control inband (x-on/x-off) or out of band (various modem leads). Note that even the DS200 can also support out of band flow control and some prefer it to getting hung x-offed, depending on the stupidity of the connected hardware. LA180 war stories, anyone?... ================================================================================ Note 234.2 Ethernet Printer Servers? 2 of 2 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 3 lines 28-NOV-1988 09:38 -< Parallel printer driven by a serial DECserver port >- -------------------------------------------------------------------------------- We run an LXY01 with a Dataproducts interface off a DECserver 100 and a serial to parallel converter! The converter cost only about $150.  ================================================================================ Note 235.0 N-Series systems and enclosures No replies EISNER::TINKELMAN "Bob Tinkelman - CCA" 131 lines 28-NOV-1988 22:35 -------------------------------------------------------------------------------- Attached is a description of the "N-Series of industrial systems", downloaded from the DEC Electronic Store.  INTRODUCTION The new N-Series of industrial systems are fully integrated cabinet configurations of the DEC/CIM family of computers. The H9901 cabinets provide protection against temperature extremes and common environmental contaminants in a manufacturing work area and can be placed near machines or sensors to increase real-time performance and H9901 Picture reduce factory cabling requirements. Designed to be used by end users, OEMs, CMPs, SCMPs, and Distributors, they are ideal system solutions for supervisory control, cell control, factory data collection and process control applications. These systems can also be used for applications in an environment where there is no climate controlled computer room and self-contained temperature control of computer equipment is required to maintain system reliability. DESCRIPTION The N-Series provide environmentally controlled system platforms for the 12-slot BA213 family. Currently available are a MicroVAX II system which supports a VMS multi-user and resource sharing operating system, the rtVAX which supports a VAXELN multitasking, real-time operating kernel, and the PDP-11/83 system which support the RSX-11 real-time operating system. The customer can choose the cabinet configuration to meet his environmental needs. A NEMA 2, fan-cooled and lightly sealed H9901, offers protection against falling dirt or dripping liquids and will operate at temperatures up to 35 o C (95 o F). A NEMA 12, air-conditioned, tightly sealed H9901, operates at temperatures up to 55 o C (131 o F) and protects the equipment against extremely fine dust, splashing liquids, and oil and coolant seepage. For European end users who purchase equipment in the USA, these enclosures meet IP22 (NEMA 2) and IP55 (NEMA 12) requirements. FEATURES/BENEFITS Feature Benefit o Fully integrated cabinet configuration o Reduces customer integration with BA213 system and installation time o Thermal interface panel with capability o Protects against damage to connect external warning device caused by temperature extremes o Two solid metal doors; optional o Easy front and back access to Plexiglass (tm) door equipment o Sides and doors are removable and o Complete installation can be rotated 180o to reposition flexibility o Two cable egress kits mountable in o Flexibility in cable any of eight different locations location and management; which are 8 x 8 wire way compatible reduces installation time (four on side and four on bottom) o Top-mounted eyebolts and fork-liftable o Can be moved via crane or base fork lift SPECIFICATIONS ----------------------------------------------------------------------- Enclosure Model Numbers H9901-AA/AB H9901-BA/BB ----------------------------------------------------------------------- Protective Sealing NEMA 2 NEMA 12 Ratings IP22 IP55 ----------------------------------------------------------------------- Cooling Mechanism 500 CFM filtered-fan 8000 BTU air conditioner ----------------------------------------------------------------------- Operating Temperature 10 to 35 deg C 10 to 55 deg C (50 to 95 deg F) (50 to 131 deg F) Relative Humidity 20% to 80% 10% to 95% Maximum Altitude 8,000 Feet 8,000 Feet ----------------------------------------------------------------------- External Dimensions 33" w x 26" d 35" w x 26" d x 76.5" h x 76.5" h ----------------------------------------------------------------------- Current Drawn Enclosure only 1.0 Amps 12.0 Amps Enclosure w/System 5.7 Amps 16.7 Amps ----------------------------------------------------------------------- Weight 400 lbs. 540 lbs ----------------------------------------------------------------------- Acoustics 54 dB 52 dB ----------------------------------------------------------------------- SPECIFICATIONS SHARED BY ALL ENCLOSURES ----------------------------------------------------------------------- Agency Approvals UL/CSA or IEC ----------------------------------------------------------------------- Shock 10g ----------------------------------------------------------------------- Vibration .5g ----------------------------------------------------------------------- Internal Mounting Dimensions available 17.75" wide x 22" deep ----------------------------------------------------------------------- Vertical Mounting 30" w/one system installed (all models) Space available 56" without a system (all models) ----------------------------------------------------------------------- Power Requirements 208/240 V nominal single phase at 30 amps ----------------------------------------------------------------------- Frequency 50-60 Hz ----------------------------------------------------------------------- Equipment w/enclosure Power controller, Thermal Interface Panel (TIP), 2 Cable egress kits, 2 keys ----------------------------------------------------------------------- System Limits Holds 2 BA213 boxes (requires 240 V versions) ----------------------------------------------------------------------- Color DEC standard 217 grey ----------------------------------------------------------------------- SUPPORT SERVICES All N-Series systems and enclosures are supported worldwide by Digital's Field Service organization. For additional information or configuration help, call the CSS Sales Support Team, DTN 264-8990 or the CSS Hotline, 1-800-TEAM-CSS. DOCUMENTATION Industrial Family of Products, Configuration Catalog, ED-31156-70  ================================================================================ Note 236.0 Kennedy 9100/TC-131 VMS 5.0 Problem 1 reply EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sap" 8 lines 29-NOV-1988 06:54 -------------------------------------------------------------------------------- I have another guy in the building where I work that can't get his Kennedy 9100 series tape drive to work under VMS 5.0 . There is an old piece of Macro code he has that is called MTDRIVER.MAR and says in it "LFTDRIVER on a TC-131 controller" but it doesn't compile properly any more. The drive is connected to a VAX 11/750. Any ideas on where to go to get some device driver code ? ================================================================================ Note 236.1 Kennedy 9100/TC-131 VMS 5.0 Problem 1 of 1 EISNER::KILGALLEN "Larry Kilgallen" 5 lines 29-NOV-1988 18:20 -< I suggest first trying to fix it yourself >- -------------------------------------------------------------------------------- An easy first step would be to make the necessary fixes to get the driver to assemble (see release notes for V5 and maybe even V4, since you don't know the VMS version to which the driver corresponds). If you don't require SMP support, the changes are rather "cookbook" in nature, and it might just work.  ================================================================================ Note 237.0 24 Node Star Coupler - CISCE-AA 1 reply EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 7 lines 29-NOV-1988 13:16 -------------------------------------------------------------------------------- In the November 1988 issue of the VAXcluster Systems Quorum (Volume 4, Issue 2), on page 44 is a note about 24 node clusters. A new star coupler expander (Part # CISCE-AA) that allows up to three SC008 Star Coupler to be connected. It uses two "active amplifier and coupler boxes", providing dual paths in case of failure. The note also talks about the "Large VAXcluster System Expansion Planning Process" which looks at I/O across the CI and attempts to say something about loads. ================================================================================ Note 237.1 24 Node Star Coupler - CISCE-AA 1 of 1 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 4 lines 30-NOV-1988 15:44 -------------------------------------------------------------------------------- At the last DECUS, they mentioned that this critter didnt change the number of VAXen supported in a CI cluster. The extra ports are intended for HSCs.  ================================================================================ Note 238.0 11/70 Technical advice 7 replies EISNER::KENNEDY "Terry Kennedy" 4 lines 29-NOV-1988 22:24 -------------------------------------------------------------------------------- This topic will be used for 11/70 techinal advice. Since the cost of contract service on 70's is so high, more and more users are turning to self-maintenance on these systems. The tips presented here are in- tended to help these users. ================================================================================ Note 238.1 11/70 Technical advice 1 of 7 EISNER::KENNEDY "Terry Kennedy" 41 lines 29-NOV-1988 22:37 -< Current module revision levels >- -------------------------------------------------------------------------------- The following list documents the latest revision (CS level) for the boards which comprise the 11/70 (KB11-B and KB11-C) CPU. The list is presented in slot order, right to left, not in module number order. M9312 - B - Boot/terminator M787 - B - Clock M8130 - D - DAP M8131 - * - GRA M8132 - C - IRC M8123 - B - RAC (used only W/ M8138-YC, else M8133) M8134 - E - PDR M8135 - * - TMC M8136 - F - UBC M8139 - D - TIG M8137 - A - SAP M8138 - F - SSR (This is the -YA variation) M8140 - B - SCC M8142 - F - CCB M8143 - B - ADM M8144 - * - DTM M8145 - C - CDP M8141 - A - MAP Massbus controller (RH70) M8150 - F - MDP M8151 - A - CST M8152 - A - AWR M8153 - B - BCT M5904 - F - MBS Memory subsystem (MK11-C*) M8158 - C M8160 - F M8161 - A M8164 - A M8728 - C This information is especially important as the DEC FCO documentation (DEC-O-LOG, MDS) is *incorrect* and *incomplete* for the 11/70. ================================================================================ Note 238.2 11/70 Technical advice 2 of 7 EISNER::KENNEDY "Terry Kennedy" 10 lines 29-NOV-1988 22:45 -< Fans >- -------------------------------------------------------------------------------- One of the common problems in the 11/70 CPU is failure of one or more of the cooling fans. In the H960-cabinet 11/70 there are 32 4" fans. The DEC fan is a Nidec-Torin TA450, model A30122-10. These fans are available from many of the commercial component distributors for far less than the DEC price. An important note is to *not* accept any substitutes for this model. This is the top airflow model in its size from *any* fan manufacturer. Also, do not accept sleeve bearings in the fan. The model number above specifies ball bearings. After you've changed all 32 fans, you'll *never* want to do it again, so spend a bit more for the bearings that last. ================================================================================ Note 238.3 11/70 Technical advice 3 of 7 EISNER::KENNEDY "Terry Kennedy" 22 lines 29-NOV-1988 22:55 -< Power supply lamp upgrades >- -------------------------------------------------------------------------------- The modular power supplies in the 11/70 have an indicator lamp which is used to show that the power supply is functional. In older supplies, this is a normal lamp which has long since burned out. Rather than removing the power supply or probing with a meter to see if it's the bulb or if the supply is *really* dead, you can replace the lamp with a LED and a resistor. This should then last nearly forever. Method: 1) remove power supply from frame (clip connector on top, 2 screws on top, large knurled screw on bottom) 2) unsolder lamp adjacent to voltage adjust trimpot 3) cut etch running from upper lamp pad to power bus by capacitor mounting screw 4) solder in a 470 ohm, 1/4 watt resistor across etch just cut. (This is for 5 volt supplies. For 15 volt supplies, use 2.2K ohm instead). 5) solder in a red LED, observing polarity (look at capacitor for + lead). (This is for 5 volt supplies. For 15 volt supplies, use a green LED). 6) replace power supply in frame and test. LED should light when powered on, and may slowly fade when powered off. Note: If the above sounds like gibberish, it's not for you... ================================================================================ Note 238.4 11/70 Technical advice 4 of 7 EISNER::KENNEDY "Terry Kennedy" 14 lines 29-NOV-1988 23:01 -< Troubleshooting RH70 / Massbus problems >- -------------------------------------------------------------------------------- Many problems which manifest as Massbus problems may in fact be caused by out-of-revision CPU modules. The first thing to check when Massbus problems are suspected are the CPU module revisions, especially the cache modules. If all the CPU modules are up to revision, look at the following items: 1) Try swapping the M5904 modules with each other. 2) Look at power supply voltages (+5). All CPU +5 supplies should be within .5 of each other, and set to between 5.05 and 5.20 volts. 3) Inspect Massbus cable connectors. Swap cables if possible. 4) If problem is only on one device of a single Massbus, try replacing the Massbus transceiver modules. 5) Swap the RH70 with a different one (M8150-M1853 modules). Note that the M8153 module has jumper options to set the interrupt and vector. ================================================================================ Note 238.5 11/70 Technical advice 5 of 7 EISNER::KENNEDY "Terry Kennedy" 11 lines 20-DEC-1988 07:25 -< And now for a question >- -------------------------------------------------------------------------------- Has anyone achieved greater than 3.5 Mb on an 11/70 with 256 Kb memory modules (MK11)? The DEC documentation states that 3.5 Mb is the limit, but the processor technical manual gives switch settings all the way up 4 Mb (obviously bogus, where would you put the I/O page?). My operating system SPD states that the max memory limit on this machine is 3840 Kb. It appears the problem may be the inability to 'ignore' memory in the MK11 box, and if I put in some 64 Kb arrays it might work... P.S. - When I install the 2 additional 256 Kb modules ans set the switches, the system chokes trying to boot... ================================================================================ Note 238.6 11/70 Technical advice 6 of 7 EISNER::KENNEDY "Terry Kennedy" 26 lines 7-FEB-1989 00:20 -< Power-up boot and auto-restart for 11/70 >- -------------------------------------------------------------------------------- This reply will discuss getting power-up boot and powerfail restart to work on 11/70 systems with MK11-C memory boxes. In order to get power-up boot and auto-restart on your 11/70, you'll need: o M9312 boot/terminator module, CS B or newer o 7010329 Backplane ECO 8 (Wrap level J or newer) o M8130 CS C or newer o M8136 CS C1 or F, or newer o M8138 CS C1 or E, or newer o Pushbutton (normally open) with Faston tabs on one end of cable o Cable with 2 conductors, Faston tabs on each end Ensure jumpers W1, W3 and W8 are in on the M8130 (W8 is unlabelled, and is to the left of W7). Connect the pushbutton's Faston tabs to TP1 and TP2 on the M9312. Order does not matter. Use the other cable to connect M9312 TP3 to the MK11-C backplane's J101 tab, and M9312 TP4 to MK11-C J100. Ensure M9312 S1 position 2 is OFF. Your 70 will now trap through 24/26 when power returns if the MK11 battery backup is still valid. Otherwise, it will trap through 773024/773026 and ex- ecute whatever boot options you have set up in the M9312 S1. ================================================================================ Note 238.7 11/70 Technical advice 7 of 7 EISNER::KENNEDY "Terry Kennedy" 12 lines 7-FEB-1989 00:27 -< ASCII console for 11/70 >- -------------------------------------------------------------------------------- You can get an inexpensive electronic console for your 11/70 by ordering the 23-248F1-00 ROM from DECdirect at $27.00. Replace the 616 boot ROM with the 248 (on the M9312) and set switch S1 positions 1 & 7 on, all others off. You now have an ASCII console. Notes: 1) The console ROM only understands upper case 2) The ROM diagnostics don't test the 11/70 cache modules 3) You should implement the pushbutton boot described in the previous reply. You'll need it to get back to the console ROM if you hang the CPU.  ================================================================================ Note 239.0 Preventative Maintenance 15 replies EISNER::NADEL "S. Freud" 28 lines 2-DEC-1988 18:06 -------------------------------------------------------------------------------- Would this be the appropriate conference for PM discussion? We have been having some problems lately with our PMs. We have been dinged about 6-8 time in the last 10 months. When the Pm is over the system is in worse shape than before. Things like: o tape drive heads completely out of alignment after pm o system unbootable after pm o system crashes within an hour of pm o hda crash after pm o misc lights and cabinet hdwr not working We have begun to investigate just what exactly PMs have prevented? I have never had a field engineer from DEC come to me after a pm and say that such-and-such a board is ready to fail, and should be replaced. It seems that we find things out WHEN they fail (sounds like a HAL model 9000). I know that the CEs do more than diagnostics, they also physically clean the stuff, but perhaps with the money I save from our monthly PM contract ($15K a month!), we could hire a new operator, and train him/her to clean them. I am interested in what the others out there in DECUServe Land are experiencing, and what your thoughts are about this. I will be preparing a report to present to a joint meeting with our DEC office re: our own experiences and would appreciate input. Thanks, Ron Nadel ================================================================================ Note 239.1 Preventative Maintenance 1 of 15 EISNER::KOZAM 39 lines 2-DEC-1988 23:39 -< PM probably isn't worth the efffort >- -------------------------------------------------------------------------------- A good PM program should tackle problems where there is a gradual pattern of failure or decrease in performance. Some examples: 1. Thoroughly clean printers. Clean out dirt and ink/cloth debris that might cause problems. Clean/replace impact printer parts that are worn (print bands and print heads). Sort out problems such as laser printers that don't work as well as they used to (spots, lines, light/dark areas). 2. Clean heads on tape drives and floppy disks. 3. Replace/Clean filters on disk/tape drives that have them. Inspect/replace the grounding clips that most disk drives use to eliminate static on the moving disk. 4. Replace fans that are very noisy and appear to be on the way to failing entirely. 5. Explore potential problems manifested as recoverable errors on disks. Is it the media or the drive? If the equipment has tests for reading with increased threshold, use them to identify equipment that is working right now, but isn't really up to snuff. 6. Use an oscilloscope to examine power supplies. Increased ripple in the DC output is a sign of impending failure. This is one instance in which electronics may fail gradually. None of these things are serious enough to necessarily serious enough to merit a trouble call, but they're problems none the less. I'd say leave boards and connectors ALONE. I've never seen or even heard of dust disabling a system. Unnecessary fooling with the boards certainly can cause problems. Obviously, doing all this is time consuming and costly and it's not clear to me that much is gained, since a large percentage of REAL problems (system downtime) are electronic and aren't detected by all this work. PM may make you sleep better, knowing you've done all you could, but I don't think it makes any difference. ================================================================================ Note 239.2 Preventative Maintenance 2 of 15 EISNER::KENNEDY "Terry Kennedy" 24 lines 3-DEC-1988 00:17 -< Useful things to do during a PM >- -------------------------------------------------------------------------------- > A good PM program should tackle problems where there is a > gradual pattern of failure or decrease in performance. Some examples: Note that most of the Field Service folks we've had here wouldn't be caught dead doing any of the things mentioned here! Our experience is that the PM is ideal for: o running destructive read/write diagnostics on any media that happens to be in the vicinity (or in the drive!) o testing device cables to see how much strain they can take (BC11 Uni- bus cables are especially good for such testing) o Gives the service rep a chance to collect some new reading material to take back to the office (such as printsets) o Gives the rep a chance to collect some minor hardware for that *other* customer who's always complaining about missing parts Our current service vendod comes by annually to replace filters. So far in a year, we've had one problem (a backplane failure). With the previous service vendor (DEC) we had problems after each PM. Lest some of you who are lucky and have better experience get to think- ing 'that would never happen with our guy', go and look at your CPU cab- inet. Missing any hardware? Care to put some money on it? ================================================================================ Note 239.3 Preventative Maintenance 3 of 15 EISNER::HARVEY "Jack Harvey" 45 lines 3-DEC-1988 07:26 -< Mechanical Preventive Maintenance OK >- -------------------------------------------------------------------------------- In vacuum tube days, we had _real_ reliability problems and huge efforts were made to project failures and do something about them. We did lots of marginal testing (varying supply voltages) to try to find weak tubes. The only real success was invention of the transistor. But the vacuum tube PM idea didn't die. Early solid state computers had variable power supplies so PM procedures could try to find marginal flip flops. When we thought we had found a weak board, about half the time the new replacement was worse. And of course, we discovered a strong correlation between failures and replacements: install replacement, replacement fails within ten days. So most computers don't have marginal test facilities now. But sometimes a field service organization will be over staffed. When a field service manager sees a bunch of people sitting around doing nothing, you can bet that preventive maintenance will get flogged. Very earnest phone calls will tell you of their new program to Maximize Up Time, or whatever the latest buzz phase is. You are also lead to believe that there is a disaster waiting to happen in your computer and this program will prevent it. The first part of that implication may be true; the second is not. There _may_ be a case for looking at error logs of memory and disks with some form of ECC to see if an unusual number of corrections have occurred. Personally, I have never seen a clear cut case. (Error logging usually just tells you precisely _when_ a disaster occurred.) Computers are far, far more reliable than they were when I wrote my first bug. But that's due to hard work in the development laboratory, not preventive maintenance programs. The VAX-11/750 CPU is very popular with third party FS firms. The only thing they have to do is send you a monthly invoice. I concur fully with .1 and .2, although Terry may be a trifle bitter about "site cleanups". And yes, a dusty computer is a contented computer. Checking power supply ripple can be useful but it must be clearly out of spec. Remember, the new replacement will probably fail within ten days. The person that does the PM probably keeps changing. This results in something different being shook up each time. With your budget, you should seriously consider assigning the very important mechanical PM such as filters, fans, printers, etc., to a permanent in-house face. One question that raises is: does the in-house person know how to do it? I suggest you have far better control of that than answering the same question for the out-house person. ================================================================================ Note 239.4 Preventative Maintenance 4 of 15 EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 14 lines 3-DEC-1988 09:43 -< PMs are good for us >- -------------------------------------------------------------------------------- When we had DEC as our service Vendor I felt I could agree with many of the comments. We hardly ever saw the same person for a problem as for PMs, I watched our PM "expert" put ROMS in the wrong order, etc. (Many people have had better luck, and DEC has made many efforts to improve since then, I have been told). We now have a third party service vendor and I really feel good about them comming in. We get to see them on a regular basis (each week, 1 machine each week) and can get a good feel for what they are up to. They run diagnostics on a single machine at a time. This gives us a convieant time to adjust system parameters and such as well. It also gives us a time to make sure the machine will reboot (real important...). They do all those other things as well, such as stay away from the boards unless they break, clean printers, clean tape drives, etc. ================================================================================ Note 239.5 Preventative Maintenance 5 of 15 EISNER::LEDERMAN "Bart Z. Lederman" 48 lines 3-DEC-1988 12:16 -< PMs have some value if done correctly. >- -------------------------------------------------------------------------------- I think there is value to PMs, but I also agree that the DEC people I've seen lately do more harm than good. It is a really good idea to check power supply voltages regularly, if you can do so without shorting anything out (fortunately, DEC equipment is getting better about having test points to do this). It's also a good idea to replace all of the dust filters, no matter how clean your computer room is. Unfortunately, our "trained" DEC serviceman didn't seem to know where they all were (he forgot or didn't bother with a small filter at the bottom front of the 11/785 for the floppy console system). In addition, I've noticed that the adhesive on the new filters doesn't work too well, so there are large gaps where unfiltered air can get in. The older cabinets on our 11/70s and PDP-15s really held them in place (though we changed from the foam filters DEC uses to Fiberglas filters sprayed with a "sticky" stuff to trap dirt). It's also a good time, as was noted, to go around and listen for funny noises from fans and disk drives (and tape drives). You CAN catch motor failures before they are fatal this way. There is also great benefit in cleaning your tape drive and checking the alignment, and if you use removable media (RP06, RM05, RM02, RK05, etc.) the head alignment really should be checked: in our case, the RP06s have to be aligned every 3 months, the tape drives are checked monthly or in some cases weekly. This does pre-suppose that you have PROPERLY trained personnel: for our critical functions we use in-house people. By the way: our PDP-15s and some other equipment is dependent on air cooling, supplied with banks of "muffin" fans. One of our PM actions is to replace ALL of the fans once every 3 years. When they reach the end of the rated life we don't fool around waiting for them to fail (leading to more serious problems, usually intermittent, due to overheating). And contrary to a message posted elsewhere, we have found that sleeve bearings last a lot longer than ball bearings in fans. This seems expensive when you consider that there are at least 15 fans in a PDP-15 CPU (not counting ANY peripherals at all), but it's cheaper in the long run. Less labor charges also to do it all at once on a predictable schedule. The one thing which really irks me the most is VAXsim. DEC has been making lots of loud noises about how the new VAXsim-Plus can predict failures by examining trends in the error log. Well, if it's really all that good, how come our local field service office won't install it on our system? We've been asking for months for them to do it, and they don't. But then, for a long time they didn't come out for PMs until we pointed out that the contract called for them and we weren't going to pay until they lived up to their side of their contract. ================================================================================ Note 239.7 Preventative Maintenance 7 of 15 EISNER::KENNEDY "Terry Kennedy" 17 lines 4-DEC-1988 04:51 -< Balls vs. sleeves >- -------------------------------------------------------------------------------- > And contrary to a message posted > elsewhere, we have found that sleeve bearings last a lot longer than > ball bearings in fans. Actually, it depends on the fan orientation. Fan horizontal is about a toss-up as the rotor is always pressing on one edge of the housing. In a vertical or random orientation, the rotor tends to 'walk' back and forth, which kills sleeves. Also, some vendors will give you an N-hour guarantee on balls, but not on sleeves. WIth the built-in odometer on the 70, this makes it a good deal. > This seems expensive when you consider that > there are at least 15 fans in a PDP-15 CPU (not counting ANY ^^ ^^^^^^ > peripherals at all)... Does that mean only 11 fans in a PDP-11/70? Or is it 70 fans? 8-) ================================================================================ Note 239.8 Preventative Maintenance 8 of 15 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 9 lines 4-DEC-1988 13:46 -< More fans yet... >- -------------------------------------------------------------------------------- >> > This seems expensive when you consider that >> > there are at least 15 fans in a PDP-15 CPU (not counting ANY >> ^^ ^^^^^^ >> > peripherals at all)... >> >> Does that mean only 11 fans in a PDP-11/70? Or is it 70 fans? 8-) Gee, I *wondered* where all that blower noise was coming from on my VAX-8650. Glad we don't have an 8800... 8-} ================================================================================ Note 239.9 Preventative Maintenance 9 of 15 EISNER::HARVEY "Jack Harvey" 11 lines 4-DEC-1988 17:41 -< Where are PM Statistics? >- -------------------------------------------------------------------------------- Ron has asked for Hardware Help and proposed a good question. I think many of us would benefit from a good answer. We have heard anecdotes and opinions, but no hard statistics. We may have a consensus that PM is a good idea for some mechanical devices. I suspect PM is usually worthless for electronics such as we find in computers. But that's just another opinion. Does anyone know of published studies demonstrating a statistically significant benefit from PM of computer electronics? I would think (hope) that field service companies would be eager to publish such results if they have them. ================================================================================ Note 239.10 Preventative Maintenance 10 of 15 EISNER::KENNEDY "Terry Kennedy" 11 lines 4-DEC-1988 18:23 -------------------------------------------------------------------------------- > We have heard anecdotes and opinions, but no hard statistics. Well, the general opinion seems to be that it is more harm than good, but sites with in-house expertise find it worthwhile. > I would think (hope) > that field service companies would be eager to publish such results > if they have them. The vast silence from service companies on this topic appears to speak for itself... [can you spot the contradiction here?] ================================================================================ Note 239.11 Preventative Maintenance 11 of 15 EISNER::TERRAZAS "Mike Terrazas" 1 line 5-DEC-1988 09:36 -< Inquiring minds want to know... >- -------------------------------------------------------------------------------- Shouldn't this discussion be in SITE_MANAGEMENT? ================================================================================ Note 239.12 Preventative Maintenance 12 of 15 EISNER::CONROY "Alan Conroy" 59 lines 5-DEC-1988 13:52 -< PMs CAN be beneficial >- -------------------------------------------------------------------------------- >< Note 239.11 by EISNER::TERRAZAS "Mike Terrazas" > > -< Inquiring minds want to know... >- > >Shouldn't this discussion be in SITE_MANAGEMENT? I agree. But until this thread is moved there, let me add my 2 cents. First, I never let our maintenance company do a PM without looking over their shoulder to keep them from doing *REALLY* dumb things. I think PMs are good for certain things: 1. Check voltages on power supplies. Those which slowly fall in output will have to be replaced some time - why not while the system is down. 2. Fans are good. If you can replace them yourself, then great. If not, then you NEED someone to do it for you. I suspect that many (if not most) sites fall into the dumb-user/manager catagory. 3. Certainly checking out noises can be useful, we just discovered a spindle that is going out on one of our RM03's (blush - yes we still have some of those around...). Since we found out about it ahead of time, we can plan some downtime to fix it rather than just waiting for catastrophic failure. And that brings me to the next point. 4. The most beneficial thing for us about PMs is that they afford us about the only downtime we can get. We use this time to fix non-critical problems (bad ports on a DH, etc) and to move/upgrade hardware (let's see how much memory we can cram into the MK box...) The users require 100% uptime the rest of the month, so having a regularly scheduled downtime to do all these other things really helps. 5. Running diagnostics certainly doesn't hurt. They might even discover some problem. We only allow CPU diags, but I suppose the danger of some tech. running a format on our system disk is possible (which is why I always write-protect the disks myself). 6. It is useful to do an ANALYZE/DISK (/NOREPAIR) to check for disk problems. This is the only "maintenance" we do on our MicroVAX 2000s. 7. When we had a PDP-11/45, the lamps often went out and PM was a good time to replace them (ditto for our RP06s, which have since gone bye-bye). As far as cleaning tape drives - any idiot SHOULD be able to do that (I mean no offense to anyone who doesn't know how - everyone needs training). However, on the removeable disks, the heads should be inspected for damage/dirt. Not too many of us would feel comfortable doing that. And that brings up another point. Twice a year we have the removable packs cleaned and inspected. Now I know that them's fighting words for some, however to defend myself let me just say that we have a reliable disk cleaning group and we've never had any problems with them in the last five years. In fact, the one time we DID have a head-crash was the one time we DIDN'T clean/inspect the new disks. The DP manager insisted that it would cost too much to have them come in (over my loud protestations), and by the end of that weekend, we had two crashed packs and two dead RM05s. Of course, there are bad disk cleaning companies out there and I do not recommend just anyone (heck, I won't recommend ANYONE). Also, this would not apply to those with no removeables. Then there's the issue of who's responsible for a major screw-up. You can eat it yourself or sue the negligent party. If something is going to go wrong, then I'd rather be in the second position. In summary, I think a well-thought out PM schedule is beneficial. ================================================================================ Note 239.13 Preventative Maintenance 13 of 15 EISNER::BYRNE_C "Charlie Byrne" 1 line 5-DEC-1988 15:12 -< WILL be moved to SITE_MGMT soon, meanwhile keep talking here >- -------------------------------------------------------------------------------- ================================================================================ Note 239.14 Preventative Maintenance 14 of 15 EISNER::LEDERMAN "Bart Z. Lederman" 47 lines 5-DEC-1988 16:36 -< Some additional thoughts >- -------------------------------------------------------------------------------- An additional 2 cents (upping the ante to 4 cents...) > 2. Fans are good. If you can replace them yourself, then great. If not, > then you NEED someone to do it for you. Not just fans. If you're the sort of person who has trouble changing light bulbs and such (and there are some perfectly normal people who just have trouble doing mechanical things) let someone else do it. > running a format on our system disk is possible (which is why I always > write-protect the disks myself). It should be SOP (Standard Operating Procedure) to take all critical disks and tapes off of your system before the service people start to work. If disks are non-removeable, write protect them or even spin them down. If maintenance of the system is important to you, keep a spare pack and tape for the maintenance people to fool with. >7. When we had a PDP-11/45, the lamps often went out and PM was a good time to > replace them (ditto for our RP06s, which have since gone bye-bye). This isn't exactly PM, but while we are on the subject: it's quite easy to replace most incandescent lamps with LEDs. There are companies which put them in packages which are the same size and shape (and voltage) as the lamps in RK05's, RP06s, CPU consoles, RS03/4s, etc. I would call this avoiding PMs by putting in longer lasting parts. In all of the disagreements we've had with DEC about maintaining equipment and the detrimental effects of third-party vendor boards on a system, etc., they have never complained about having LEDs instead of incandescents, and even used their lamp puller to put them in on or RK05. They didn't even mind that we had to clip out the keep-alive resistors on the RP06 when we switched. I highly recommend this. > on the removeable disks, the heads should be inspected for damage/dirt. Not > too many of us would feel comfortable doing that. And that brings up another > point. Twice a year we have the removable packs cleaned and inspected. Now Absolutely, except we do ours 4 times a year. You MUST inspect removable disk packs and heads, even if the pack is constantly in one drive. Our RK05s are done in-house using a device sold by DEC, all others are done by an outside firm specializing in the work. Even though our people can align disk drives, we use specialists to examine and clean them. And yes, we've caught crashes that had started to propagate between systems by examining the disks and heads. ================================================================================ Note 239.15 Preventative Maintenance 15 of 15 EISNER::BYRNE_C "Charlie Byrne" 2 lines 5-DEC-1988 19:27 -< This Topic moved to SITE_MANAGEMENT >- -------------------------------------------------------------------------------- This thread has been set NOWRITE and moved to SITE_MANAGEMENT 32.*.  ================================================================================ Note 240.0 PDP-11/73 CPU CARD VERSIONS 11 replies EISNER::CONROY "Alan Conroy" 6 lines 7-DEC-1988 11:27 -------------------------------------------------------------------------------- Can anyone explain to me the different versions of the 11/73 CPU cards? I've heard that there are two (at least) versions. One is faster, etc. The first card in our CPU is numbered 8180. What does that tell me (if anything)? Thanks in advance. ================================================================================ Note 240.1 PDP-11/73 CPU CARD VERSIONS 1 of 11 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 42 lines 7-DEC-1988 12:39 -< 11/73'S & 11/83'S >- -------------------------------------------------------------------------------- > Can anyone explain to me the different versions of the 11/73 CPU cards? Someone will remember the module numbers but here goes... There are 3 11/73 cards and 2 11/83 cards All 73's and 15MHZ clocks and all 83's have 18MHZ clocks The 73's are as follows... The original dual width card which CAN'T use the FPP chip The new dual width card which CAN use the FPP chip The quad width card which CAN'T use the FPP chip NOTE: The quad width card can be 17-20 percent faster than the dual width card even though they have the same clock rate. If you plug 11/83 PMI memory into a Q/CD system with the quad width 73 card you will get a 32-bit memory bus. All of our quad width 11/73's are shipped with Clearpoint QED-1 4MB memory boards. The 83's are a follows... The original quad width card which CAN'T use the FPP chip The new quad width card which CAN use the FPP chip We find the systems rate as follows 11/73 without FPP & without PMI .7 11/73 with FPP & without PMI .8 11/73 quad without FPP & with PMI 1 11/83 without FPP & with PMI 1.2 11/83 with FPP & with PMI 1.3 ================================================================================ Note 240.2 PDP-11/73 CPU CARD VERSIONS 2 of 11 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 11 lines 7-DEC-1988 16:58 -< CD interconnect for PMI >- -------------------------------------------------------------------------------- >>>>NOTE: The quad width card can be 17-20 percent faster than the dual width >> card even though they have the same clock rate. If you plug >> 11/83 PMI memory into a Q/CD system with the quad width 73 card >> you will get a 32-bit memory bus. All of our quad width 11/73's >> are shipped with Clearpoint QED-1 4MB memory boards. If someone wants to do this, the memory MUST preceed the CPU in the backplane to implement PMI. Putting the memory after, makes it act like just fast Q-Bus memory. Actually You can put up to 16 QED-1s for 64 meg as the QED-1s can bank-select a variable size window. Can you say "BIG RAM DISK". ================================================================================ Note 240.3 PDP-11/73 CPU CARD VERSIONS 3 of 11 EISNER::HARVEY "Jack Harvey" 7 lines 7-DEC-1988 19:02 -< KDJ11-A = M8192 >- -------------------------------------------------------------------------------- >The first card in our CPU is numbered 8180. >What does that tell me (if anything)? For the original dual size 11/73, the module number is M8192, non-FPP compatible. I can't identify 8180. ================================================================================ Note 240.4 PDP-11/73 CPU CARD VERSIONS 4 of 11 EISNER::KOZAM 8 lines 7-DEC-1988 20:39 -< PMI only in Q/CD slots >- -------------------------------------------------------------------------------- One word of caution about PMI memory boards: They MUST be used in Q/CD slots! If you put them into a Q/Q slot, you WILL PERMANENTLY damage the board. Aside: With the Unibus systems, it is trivial to blow up things by putting boards in the wrong slots. This is the only Q-bus example I know of. Any others? ================================================================================ Note 240.5 PDP-11/73 CPU CARD VERSIONS 5 of 11 EISNER::MAYHEW "Bill Mayhew" 9 lines 7-DEC-1988 20:47 -< Quad board M-numbers >- -------------------------------------------------------------------------------- I can't find an M8180 either. The information I have from DEC is that an M8190 is a KDJ11-BC (quad 11/73, non-FPP-compatible), an M8190-AB (look "lower" on the module handle for the suffix if any) is a KDJ11-BB (quad, FPP compatible), and an M8190-AE is a KDJ11-BF (11/83 with FP). Sometimes the numbers are very hard to read; might yours be an M8190, not an M8180? ================================================================================ Note 240.6 PDP-11/73 CPU CARD VERSIONS 6 of 11 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 6 lines 8-DEC-1988 07:16 -< THERE IS ONLY ONE BUG FREE QUAD BOARD >- -------------------------------------------------------------------------------- > handle for the suffix if any) is a KDJ11-BB (quad, FPP compatible), > and an M8190-AE is a KDJ11-BF (11/83 with FP). FYI - I believe there were two fixes to the FPP logic. The KDJ11-BF is the only quad board that works with FPP and all O/S's. ================================================================================ Note 240.7 PDP-11/73 CPU CARD VERSIONS 7 of 11 EISNER::GRANZOW 48 lines 8-DEC-1988 17:33 -< PDP 11/73 CPU BOARDS >- -------------------------------------------------------------------------------- < Note 240.0 by EISNER::CONROY "Alan Conroy" > -< PDP-11/73 Questions >- Hope this helps! KDJ11-AA 11/73 M8192 Dual Height CPU, with DCJ11-AC NO Floating Point compatibility. KDJ11-AB 11/73 M8192-YB Dual Height CPU, with DCJ11-AC, With Floating Point Connector (FPJ11 not installed) KDJ11-AC 11/73 M8192-YC Dual Height CPU, with DCJ11-AC, With Floating Point. KDJ11-BA Q-Bus M8190-AD Quad Height CPU 18 MHZ, with DCJ11-AA With Floating Point compatibility (FPJ11 not installed) KDJ11-BB 11/73 M8190-AB Quad Height CPU 15 MHZ, with DCJ11-AA Floating Point compatibility (FPJ11 not installed) KDJ11-BC 11/73 M8190 Quad Height CPU 15 MHZ, with DCJ11-AA NO Floating Point compatibility. KDJ11-BD 11/73, 74 M8190-AC Quad Height CPU 15 MHZ, with DCJ11-AA Floating Point (FPJ11 INSTALLED) KDJ11-BE 11/73 M8190 Quad Height CPU 18 MHZ, with DCJ11-AE NO Floating Point compatibility. KDJ11-BF 11/83, 84 *M8190-AE Quad Height CPU 18 MHZ, with DCJ11-AA Floating Point (FPJ11 INSTALLED) KDJ11-BG 11/83, 84 M8190 Quad Height CPU 15 MHZ, with DCJ11-AB NO Floating Point compatibilty. * The Micro-PDP 11/83 ,84 system uses the KDJ11-BF CPU module, one of more of the following memories: MSV11-JB 11/84 1 MB ECC MEMORY M8637-BA MSV11-JC 11/84 2 MB ECC MEMORY M8637-CA MSV11-JD 11/83 1 MB ECC MEMORY M3637-DA MSV11-JE 11/83 2 MB ECC MEMORY M3637-EA ================================================================================ Note 240.8 PDP-11/73 CPU CARD VERSIONS 8 of 11 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 26 lines 8-DEC-1988 20:11 -< Early MSV11-Jx NOT BLOCK-MODE DMA >- -------------------------------------------------------------------------------- >>KDJ11-BD 11/73, 74 M8190-AC Quad Height CPU 15 MHZ, with DCJ11-AA >> ^^ Floating Point (FPJ11 INSTALLED) This must be a typo, as the 74 is a multi processor 11/70, NOT a J11 product. >>* The Micro-PDP 11/83 ,84 system uses the KDJ11-BF CPU module, one of more >>of the following memories: >>MSV11-JB 11/84 1 MB ECC MEMORY M8637-BA >>MSV11-JC 11/84 2 MB ECC MEMORY M8637-CA >>MSV11-JD 11/83 1 MB ECC MEMORY M3637-DA >>MSV11-JE 11/83 2 MB ECC MEMORY M3637-EA Yes, they will all work, BUT DON'T use the -JB or -JC with a 73 or 83, as these two cards PRECEEDED their getting block mode DMA working, and lacking it your HOT Q-bus disk or tape controllers will c r a w l. The -JD or -JE or Clearpoints QED-1 support block mode DMA, and are all you should ever buy. A while back DEC's order processing REJECTED the -JD or -JE if ordered as expansion memory on an 11/84 order, and Field Service was threatening to not support it. It sure helped to have the 11/84 Product Line Manager's card from a previous symposium! Thank you DECUS. Now even Dec Direct has finally got it straight. ================================================================================ Note 240.9 PDP-11/73 CPU CARD VERSIONS 9 of 11 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 9 lines 8-DEC-1988 20:19 -< QED-1 can be safe in Q/Q >- -------------------------------------------------------------------------------- >> One word of caution about PMI memory boards: They MUST be >>used in Q/CD slots! If you put them into a Q/Q slot, you WILL >>PERMANENTLY damage the board. True, BUT, if you want to get GOOD memory for your dual 73 in a Q/Q box now with hope of using the memory later in a Q/CD box with a quad card later, GET the CLEARPOINT QED-1, as it has a row of jumpers you can yank to make it safe in Q/Q. Later, stuff the push on jumpers back in. ================================================================================ Note 240.10 PDP-11/73 CPU CARD VERSIONS 10 of 11 EISNER::CONROY "Alan Conroy" 8 lines 9-DEC-1988 11:48 -< Thanks >- -------------------------------------------------------------------------------- > Sometimes the numbers are very hard to read; might yours be an M8190, > not an M8180? You're right. The 11/73 is hidden in the recesses of one of our 11/70 cabinets (it's dark in there!). It is M8190. Thanks all! DECUServe is probably the best thing that ever happened to DECUS!!! I don't know where else I could get all this wonderful data. ================================================================================ Note 240.11 PDP-11/73 CPU CARD VERSIONS 11 of 11 EISNER::HARVEY "Jack Harvey" 3 lines 9-DEC-1988 19:50 -< 73 CPUs - Good Question >- -------------------------------------------------------------------------------- >I don't know where else I could get all this wonderful data. The trick was the question. Many others (including me) got the data too.  ================================================================================ Note 242.0 TU81+ load failures. 6 replies EISNER::SCOPELLITI "Whatsa behind is uva no importa" 51 lines 8-DEC-1988 01:04 -------------------------------------------------------------------------------- We've recently run into a nasty tape problem on a TU81+ tape drive. The problem symptom was that a batch of tapes that had once been used would no longer load. The drive would attempt to load the tape by spooling tape until the Beginning of Tape (BOT) mark passed the detector. Then, it would attempt to stop and reverse to position at BOT. At that point it would fail (error E12). All the failing tapes were 3M Blackwatch tapes - the infamous backcoated tape. All other tapes seemed to work just fine. DEC Field Service adjusted the hell out of the tape drive with no effect and finally stated that the problem was the backcoated tapes (Digital maintains that it does not support the use of 3M Blackwatch tapes). This seemed difficult to fully accept since we've used a mess of Blackwatch tapes with very few problems. As we discussed the issues over the tape drive, RDC called (Field Service had placed a call). They suggested the following procedure: Physically mount the tape, press TEST, then LOAD/REWIND. This causes the entire tape to be spooled onto the take-up reel, then spooled back to the original reel. Surprise, surprise! The tape then properly mounted! Seems that the problem is caused by incorrect take-up tension on the tape. All the failing tapes had been used on ANOTHER VAX whose TU81+ had for a while experienced some seemingly unrelated problems. Apparently, unbeknownst to field service, when they fixed the drive, the incorrect tension problem was also fixed. Operations personnel had seen the problem, and just de-activated the failing tapes. RDC also indicated that the problem can be caused by someone spinning the tape in their hands - sorta like a vertical frisbee (Yo! Alan... no pun intended). This causes the tape to slip and loose proper tension. I occasionally do this while waiting for some tape operation to complete. And ya know, I have seen this problem a couple of times and just passed it off as a random event. An interesting thought is that blackwatch tapes do _NOT_ come with anything to hold the tape in the reel - either a vinyl strip or a foam insert. All DEC tapes have one or the other. Perhaps this led to the accusing finger being pointed at blackwatch tapes in this case. Final comments: This does not necessarily indicate that backcoated tapes are fine to use on any or all Digital tape drives. I'm just posting this as an information aide to other tape users. Field service had never heard of the TEST - LOAD/REWIND function. Score 1 for RDC. ================================================================================ Note 242.1 TU81+ load failures. 1 of 6 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 1 line 8-DEC-1988 20:38 -< check previous BLACKWATCH comments >- -------------------------------------------------------------------------------- This ties in with some of the comments back in 182.* amd 213.* ================================================================================ Note 242.2 TU81+ load failures. 2 of 6 EISNER::HIBBITS 61 lines 9-DEC-1988 13:14 -< Mysteries of TU81s >- -------------------------------------------------------------------------------- -< TU81+ load failures. >- > We've recently run into a nasty tape problem on a TU81+ tape drive. > The problem symptom was that a batch of tapes that had once been > used would no longer load. The drive would attempt to load the > tape by spooling tape until the Beginning of Tape (BOT) mark passed > the detector. Then, it would attempt to stop and reverse to position > at BOT. At that point it would fail (error E12). We've been using a TU81+ for about 18 months, pushing about 50- 100 tapes per day through it. We've also amassed alot of experience in the "sensitivity" of the device to cleaning, media, and alignment. The cleaning and media issues are often hotly debated, so I won't go into what works for us, to each his own system that WORKS :-) Most important is the alignment issues that we've encountered. The drive is a "self-adjusting" drive and is supposed to be aligned at the factory. From then on, most drives are left untouched except for the occaisonal Test 1. What we've found, though, is that the electronics and various physical factors tend to make the drive performance drift from the initial tuning. Therefore, after much playing around on the part of a very wise and dedicated Tape Engineer (remember them?), we hit on a program of tuning that includes the normal cleaning, etc, but also the performance of 13 internal alignment/adjustment/diagnostic tests. Following is the list of the tests: TU81 Adjustment/Alignment Tests Test 37 - Setup velocity correction multiplier Test 82 - Enable load tension mode Test 64 - Automatic write current level adjustment Test 18 - Forward 75 ips continuous (for 10 seconds) Test 31 - Setup/check read amplitudes at 25/75 ips Test 48 - Setup powerdown timer Test 60 - Loop write-to-read through formatter Test 59 - Loop write-to-read through read/write Test 58 - Loop write-to-read through driver interface Test 57 - Interface internal diagnostic Test 56 - Formatter internal diagnostic Test 55 - Drive internal diagnostic Test 62 - Take-up fine tach fault detection These must be performed by a Field Service person, since they require that a jumper be moved with the power on. With our volume of tapes, we require that these be done about every 8-12 weeks. It takes about and hour, and when complete PRESTO, problems invariably evaporate. One other issue: there can be repeated errors of TAPE POSITION LOST, due to a bug in TUDRIVER. A patch to TUDRIVER is available to treat recoverable postion lost errors as non-fatal instead of the shipped treatment as fatal. Non-recoverable errors are still treated as fatal. DSIN has a note on this issue. The patch, to all I've talked to, is simply referred to as the TU81E patch and addresses some MSCP protocol garbaging when the cache gets fouled up. The same patch applies to TK70 tape drives. I apologize for the length of the note, but felt a need to be complete. Jim ================================================================================ Note 242.3 TU81+ load failures. 3 of 6 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 17 lines 9-DEC-1988 16:01 -< what TAPE is it? 700? >- -------------------------------------------------------------------------------- >> I apologize for the length of the note, but felt a need to >>be complete. NONSENSE! That was SUPER! and is the sort of material that makes Decuserve so valuable. THANK YOU! >> We've been using a TU81+ for about 18 months, pushing about 50- >>100 tapes per day through it. We've also amassed alot of experience in >>the "sensitivity" of the device to cleaning, media, and alignment. The >>cleaning and media issues are often hotly debated, so I won't go into what >>works for us, to each his own system that WORKS :-) With that much experience, and a clearly working 'SYSTEM', please finish the story. I am very curious about what tape you do use. Now, does someone know how to super-tune a Kennedy 9400... ================================================================================ Note 242.4 TU81+ load failures. 4 of 6 EISNER::HARVEY "Jack Harvey" 2 lines 9-DEC-1988 21:10 -< Useful Preventive Maintenance >- -------------------------------------------------------------------------------- I have cross-referenced .2 to the SITE_MANAGEMENT thread 32 on Preventive Maintenance Issues. ================================================================================ Note 242.5 TU81+ load failures. 5 of 6 EISNER::BYRNE_C "Charlie Byrne" 4 lines 23-MAR-1989 11:17 -< Dual port tapes? TU81 Q-bus Controller? >- -------------------------------------------------------------------------------- Speaking of TU81: 1) Can they be dual ported? 2) If I add one on to MV3600, what is the controller board? ================================================================================ Note 242.6 TU81+ load failures. 6 of 6 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 36 lines 23-MAR-1989 18:35 -< N-way port CDC 92185s (TU81's guts) >- -------------------------------------------------------------------------------- > 1) Can they be dual ported? No idea, can't afford a CDC 92185 @ $30,000. with DEC's label on it. But a non DEC labeled CDC 92185 is the popular Pertec tape drive interface, and the Q bus controller of choice is the Dilog DQ153 (DQ3153 for SKUNK boxes). All that in a nice tape drive cabinet with cables often sells NEW for $12k, and sometimes a lot less... All Pertec interfaced drive/controller combinations are n-way switchable with other brand switches - a popular application. This includes your old TU80 which uses a Dilog designed but DEC manufactured controller (on the U bus). The switch we use is marketed by Iverson (the hop up your 750 folks), and has ports for 4 drives and 4 cpus, and has 4 knobs. ANY drive can be switched to ANY cpu, and 'drive' can actually be several daisy-chained together. Ours was bought as the BIG box (4x4), but only populated with the cards for 1 tape x 2 cpus to save $s, but to leave expansion room. > 2) If I add one on to MV3600, what is the controller board? Your OLD KLESI RC25 controller is it! That was the BIG secret - you COULD have put a TU81 on a Q bus long before DEC said you could. I don't know what else that controller does other that RC25s and TU81s, but someone else may. Asking for it as an RC25 controller should convince the used hardware dealer you don't intend to spend much $s. A 'SKUNKED' version won't be found used, probably, but some people cheat and mount old cards in the new style boxes and sneak tha cables out somehow. This all sounds so familiar as it has all been said somewhere here before, but I typed it all in rather thansearching as it seemed quicker. Usual not connected to ANY of the above applies, and if you need Iverson's # ask and I will dig it out.  ================================================================================ Note 243.0 LJ250 / IBM PC Experience? No replies EISNER::ROOSE "ROOSE" 4 lines 8-DEC-1988 09:20 -------------------------------------------------------------------------------- Has anyone had experience using an LJ250 printer connected to an IBM PC (or clone) with VT340 emmulation software, such as Reflection-4 ? I would be very interested in your experiences and would appreciate input on this subject. Thank you!  ================================================================================ Note 244.0 VT320 Problems Anyone? 11 replies EISNER::ROECKEL 39 lines 8-DEC-1988 15:01 -------------------------------------------------------------------------------- We have had some ongoing problems with our VT320 terminals and wonder if anyone else is experiencing the same problems. Basically, the problems can be boken down into two areas (a) screen adjustment/alignment and (b) 132 character mode. (a) screen adjustment/alignment Many of the terminals have the picture cocked abit. Others have certain areas on the screen that are 'bent'. (i.e. draw a straight line on the tube, and it looks crooked. And still others have 'wavy' characters (power supply problem?) (b) 132 character mode In 132 column mode, there are a few letters that are not readable. namely the 'm' and the 'w'. History: We received the terminals from an OEM with the purchase of a 'syst'. I called DEC to complain, and they said 'sorry, the OEM must resolve your poblems'. O.K. Fine. But the OEM is saying they are getting the runaround from DEC and can't resolve the problem. So off I go to the Spring DECUS, and I talk with the DEC product manager (or so he said he was.) He stated 'OH !! you must have some of those 'early ship' terminals that have a bad character set prom in 132 column mode, and got through QA with poor screen alignments. Have your OEM contact me.' I thought GREAT !! Problem fixed. Told OEM. They told DEC. DEC says 'who is this guy you spoke with? He does not work for DEC!' They even furthur explain that there was an 'imposter' rooming the floor, claiming to be from DEC that was giving the VT320 a bad name. Now what ? How is everyone else making out with thier VT320's? ================================================================================ Note 244.1 VT320 Problems Anyone? 1 of 11 EISNER::MAYHEW "Bill Mayhew" 46 lines 8-DEC-1988 20:42 -< Some VT320 problems >- -------------------------------------------------------------------------------- I have noticed slight "tilt" on some VT320s that we have installed at customer sites, along with a slight "bow" in the horizontal scan lines, particularly at the bottom of the image. I have found that I find the "tilt" more objectionable than my customers have (we are an OEM). It is very slight (1/2 to 1 degree, I'd say) on the units we've installed. I had a client call about the "bow" shortly after we installed 14 VT320s for them. It was *slightly* evident on all the units, but considerably worse on one. That one was replaced under warranty. After staring a whole lot at the "bow", I couldn't convince myself if it was real or an optical illusion caused by the "flat" screen in combination with the typical viewing angle. I could only see it at some viewing angles. I have seen no incidents of "wavy characters", nor have I noticed any problem with the "m" and "w" in 132-col mode (though both we, and our clients, use 132-col mode quite sparingly). I will say that my first impression of the VT320 was that it was not up to normal DEC engineering standards, except for the keyboard (which of course is not significantly different from the VT220 keyboard). One problem I have found very objectionable is the amount of RFI put out by a North American VT320. I have one at home, and if it is turned on, you essentially can't watch channel 4 (and 2 and 5 are marginal) on any TV in the house using a standard antenna -- lots of little dots. I've never seen a DEC monitor do this. I opened the cover and found that the 320 lacks the copper shielding that is found in a 220 or VR201 (Rainbow/Pro/DECmate monitor). I wonder if an international-model 320 (which is FCC rated Class B vs. Class A) would do better... Since we sell all terminals with standard DEC warranty, which the customers have all upgraded to on-site coverage, I would expect DEC to fix any problems like this that the customer feels are, in fact, problems. Depending on the specific terms and conditions under which your purchase took place, you may or may not have a DEC-provided warranty. If you _do_ have a DEC warranty, or a DEC service contract on the system, then DEC's pointer back to the OEM is inappropriate. However I can fully believe that the OEM is getting the runaround in getting the problem resolved from DEC. If I were you, I would get together with a management-level representative of the OEM and make a joint call (personal or phone) on, say, the District Service Manager for your DEC field service district. The two of you should be able to make it clear to him that there is a problem and that you want a solution found, not finger-pointing. This assumes that there is in fact some reason to believe that DEC is providing a warranty and/or service contract on these terminals. ================================================================================ Note 244.2 VT320 Problems Anyone? 2 of 11 EISNER::KOZAM 13 lines 8-DEC-1988 21:23 -< VT320 downloaded character sets >- -------------------------------------------------------------------------------- One difficulty I have seen with VT320s is in using downloaded character sets. The VT220s used the same downloaded set for both 80 and 132 column modes. VT320s (like VT240s) require two separate sets - one for 80 charcter mode and one for 132 character mode. This wouldn't be a problem, except that the VT320s can only store one character set at a time. Thus, each time you switch your VT320 between 80 and 132 character modes, you must also reload the character set. This makes some word processing packages break (they expect to load all special character sets when they first start up). Has anyone else seen this? Is there a fix? ================================================================================ Note 244.3 VT320 Problems Anyone? 3 of 11 EISNER::LEDERMAN "Bart Z. Lederman" 17 lines 9-DEC-1988 08:23 -< Also not impressed with VT320s >- -------------------------------------------------------------------------------- I have to agree that I am not favorably impressed with the VT320s I've seen so far. We bought a few, and put them into operation in what is normally the "best" mode, which is black letters on a white screen. The "paper white" screen is more like "paper gray" compared with VT220s and even my VT125. I did like the improvements in the keyboard over the VT220 (that stupid compose key location, for example), and the shapes of the characters is good, but the screen just wasn't bright enough. (We gave them to people who were using really old and beat up VT100s, and they were an improvement over that.) By the way, black letters on a white background is what the ergonometric experts now endorse. It's closest to what you are used to working with (how many of you print documents using amber letters on black paper?), so there is less of a strain going from documents to CRT screen to documents, and glare affects the screen less. ================================================================================ Note 244.4 VT320 Problems Anyone? 4 of 11 EISNER::PRIGOT "Jonathan M. Prigot" 8 lines 9-DEC-1988 10:11 -< Wavy trained >- -------------------------------------------------------------------------------- > ... And still others have 'wavy' characters (power supply problem?) It could be a power supply or it could be from any small electric devices in the vicinity. I had a user complain about wavy characters. The user had a small "air cleaner" (one of those fan and charcoal filter jobbies) about two feet from the terminal. When we turned it (the filter) off, the wavies went away. ================================================================================ Note 244.5 VT320 Problems Anyone? 5 of 11 EISNER::RENES "Stuart Renes, DFWLug Chair" 16 lines 9-DEC-1988 16:22 -< Mine bad, too. >- -------------------------------------------------------------------------------- I too have a VT320 that has a misaligned CRT. After removing the cover and attempting to correct the problem, I discovered that it's probably the CRT deflection plates that are out of alignment - a situation that can only be corrected to some degree by the little magnets glued to the side of the yoke assembly. This is the worse case of misalignment I've seen since the VT100 terminal days. The problem was so noticable in reverse-video (black char on white background) that I can't stand to use that mode. Stuart. :-( ================================================================================ Note 244.6 VT320 Problems Anyone? 6 of 11 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 2 lines 9-DEC-1988 16:34 -< You're not alone >- -------------------------------------------------------------------------------- My VT320 also has bent horizontals and a *very* poor font in 132 mode ("m" and "n" and "w" are bad!!!). ================================================================================ Note 244.7 VT320 Problems Anyone? 7 of 11 EISNER::STERN "Seth Stern, (216) 266-6721" 13 lines 15-DEC-1988 08:35 -< Terminal Product Enhancement? >- -------------------------------------------------------------------------------- When the VT300 family was announced I went to the DECUS session where Digital described the terminals and asked why they had changed to such an unreadable character set. I was told that numerous ergonomic studies had been conducted to determine how to add serifs to characters without having the necessary display resolution and mumble, mumble. Besides, they said, you can always download your own character set. Great response. Have you noticed the the Terminal Products Group thinks of their output as hardware almost as much as Educational Products thinks of self-paced training as software? (What do you mean, "What version is EDTCAI?" What's a version?) And lets not forget "U" and "V" in 132-column mode. (But then, who uses "V" in our line of work? ;-) ) ================================================================================ Note 244.8 VT320 Problems Anyone? 8 of 11 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 14 lines 16-DEC-1988 00:47 -< The fault lies not in the stars, but in ourselves >- -------------------------------------------------------------------------------- Actually, DEC conducted several studies at a DECUS past. If you wandered by the terminals booth and talked to the right person, you were asked if you were willing to take part in a one hour survey. If affirmative, you signed up for a given time and were told the location (and given the secret word XYZZY :-} ). Anyway, when you showed up at the appointed time/place, you were handed a survey form and asked to evaluate about 8 or 10 terminal screens for various parameters - readability, clarity, etc. The screens were a collection of most of the popular types, cleverly covered with a black cloth so all you saw was the screen. So, maybe we have ourselves to blame (if there is blame to be placed). ================================================================================ Note 244.9 VT320 Problems Anyone? 9 of 11 EISNER::ROECKEL "Bruce W. Roeckel" 33 lines 14-OCT-1989 10:26 -< 132 Column Mode Fix! >- -------------------------------------------------------------------------------- > (b) 132 character mode > > In 132 column mode, there are a few letters that are not readable. > namely the 'm' and the 'w'. A Fix!! Beleive it or not, after more than a year of working on this problem, DEC field service has a fix! It's real simple -- replace two PROMS in the terminal and cut the W1 jumper. And boy, what a difference! 132 column mode is *GREAT* -- you can actually read the characters. Once the upgrade is done, the version number on the setup screen shows V1.2 (from V1.1). Here is the hitch -- DEC policy is that this modification WILL NOT be incorporated into any existing or NEW VT320's !! Can you beleive that?! Because I screamed loud enough, DEC has agreed to replace all the PROMS in our 36 VT320's free of charge. A simple statement like 'looks like it's time to return these for a refund' appeared to work very nicely for us. Of course, it did take some senior level executives to help push the issue!! The way I understand it is that this fix is available by request and for a nominal fee. This is only for the North American version -- the International version was fixed prior to releasing it (or something like that.) Well, I am probably going to realy catch some grief from DEC for letting the cat out of the bag, but darn it all, I'm prepared to take it because I think thier policy *STINKS*! >Bruce ================================================================================ Note 244.10 VT320 Problems Anyone? 10 of 11 EISNER::DELARISCH "Arnold S. De Larisch" 34 lines 15-OCT-1989 17:09 -< Needs to be moved to Business Practices! >- -------------------------------------------------------------------------------- >> < Note 244.9 by EISNER::ROECKEL "Bruce W. Roeckel" > >> -< 132 Column Mode Fix! >- >> Here is the hitch -- DEC policy is that this modification WILL NOT >> be incorporated into any existing or NEW VT320's !! Can you beleive >> that?! Because I screamed loud enough, DEC has agreed to replace >> all the PROMS in our 36 VT320's free of charge. A simple statement >> like 'looks like it's time to return these for a refund' appeared >> to work very nicely for us. Of course, it did take some senior >> level executives to help push the issue!! >> Well, I am probably going to realy catch some grief from DEC for >> letting the cat out of the bag, but darn it all, I'm prepared to take >> it because I think thier policy *STINKS*! It sounds like this is another issue to be brought forth in BUSINESS PRACTICES. -Arnold P.S. I'm starting to get SICK in TIRED with DEC's attitude of "We poorly designed this product, but we're only gonna 'fix' the problem for 'influential' accounts who have to threaten DEC into the fix. They then have the fix, but tell the customer, Sushhhhhh! Don't tell anyone or we may have to install $5 in components into this mal-designed box." This Attitude is NOT GOOD FOR BUSINESS DEC! I think your stocks are starting to accurately represent customer's dissatisfaction with non-uniform Business Practaces. DEC GROW UP! ================================================================================ Note 244.11 VT320 Problems Anyone? 11 of 11 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 36 lines 16-OCT-1989 00:40 -< DEC could WIN some points if... >- -------------------------------------------------------------------------------- > Here is the hitch -- DEC policy is that this modification WILL NOT > be incorporated into any existing or NEW VT320's !! Can you beleive > The way I understand it is that this fix is available by request > and for a nominal fee. It doesn't sound as though DEC really wants to make money fixing this, but simply wants to avoid having to do it free for everyone. Would DEC let the PROM contents be put on some SIG tape in a format acceptable to most folk's handy prom blaster? Are they some standard PROM? I can't believe they wouldn't be standard, but it is wisest to ask... Are they now a on an erasable (recycleable) Prom? Are the NEW proms the same as the old ones, or is the jumper cut to indicate that larger proms are being installed? Actually, DECUSERVE would be a handy place to have the prom contents posted. Also, those with MANY such terminals can get just one fixed, and then don't need anyone else to tell their prom blaster what to do. Is there an FCO number or some such for this and how many hours minimum for a 5 man team will DEC insist on being paid to fix one terminal? This is the sort of thing that DEC could win big points with their slightly disgusted customers, by compromising nicely. Make a neat little kit of the 2 proms, and an instruction sheet. Offer it for $15 via the electronic store only, and PUBLICALLY announce it is legal to COPY them if desired. At $15, most would just buy them. Lets be practical. $15 is too small, so make the kit have 4 sets for 4 terminals, 1 instruction sheet, and sell it for $50. DEC could also GIVE a prom pair to each attendee at Anaheim in the spirit of 'those needed little parts you can't easily get' that were suggested as replacements for all the junk giveaways that got turned off 'for our good' by our 'leadership'.  ================================================================================ Note 245.0 VS3200 + Emulex QT13 + Gigastore = ??? 4 replies EISNER::OSUDAR "John Osudar" 50 lines 9-DEC-1988 19:26 -------------------------------------------------------------------------------- I'm looking for clues as to where to look next for an answer (or where to place the blame for the problem); here's the situation: We just installed an Emulex QT13 tape controller in a VAXstation 3200. (Yes, I know what's been said about Emulex here and elsewhere -- but we didn't have any real alternatives.) This controller does TMSCP, so it works (claims to work) with the vanilla TUDRIVER. (We're using the controller to drive a DigiData Gigastore VHS tape unit, not a real (reel? :-) tape drive.) Emulex claims that the QT13 will work in a 3200 (unlike several other controllers that work in MicroVAX II's but have serious problems in the 3200.) DigiData claims that the QT13 works with their Gigastore unit, but (as far as I know) they've only tested it in a MicroVAX II. Attempting to use MUA0: with BACKUP usually fails after a varying amount of I/O activity, with two (yes, two) active I/O's and possibly several more pending, with the process waiting for I/O completion. Attempting to read the tape (with a simple multi-buffered read program) also fails after some amount of activity, also with two active I/O's plus some pending. Between one failure and the next, the Gigastore drive may be idle when things hang; or it may be doing some very strange-looking sequences of forward reads, forward skips, and backward skips/rewinds, with all the error indicators flashing (which is supposed to mean that you're reading a blank tape -- only this one isn't blank!) Aborting the program (with control-Y or $FORCEX), or stopping the process, throws the process into RWAST, and the only apparent way of clearing this is to reboot. Once I had the system crash shortly after doing this, apparently at fork IPL somewhere in the TUDRIVER. Whatever *is* being written to the tape appears to be correct, at least to the extent that I can read the tape. (I do notice that some I/O's report completion out of order -- I have a vague recollection that this is a "feature" of the TUDRIVER and not a bug???) What I'd like to know is: (1) Is anybody out there using a QT13? If so, does it work? (2) Has anybody tried it in a 3200 system? (Or its brother, the QT14, in a 3500/3600?) (3) Has anyone seen problems with MUxx: devices hanging with I/O's active/pending, and then going into permanent RWAST when aborted? (4) Can anyone shed any other light on the subject? I'll be attempting to get each of Emulex, DigiData and DEC to come up with some answers for me, but I suspect I'll end up with a "circular blame loop" and not much else. Any help at all will be appreciated. ================================================================================ Note 245.1 VS3200 + Emulex QT13 + Gigastore = ??? 1 of 4 EISNER::KENNEDY "Terry Kennedy" 14 lines 10-DEC-1988 01:57 -< Does this help? >- -------------------------------------------------------------------------------- Well, all I know of is the infamous 'Attn: Non-Digital Device Users' memo, the latest rendition of which is dated 12-Aug-88. It lists three differences (their word, I'd say 'problems') in the 3xxx processor. In summary: o Talking directly to the Q-bus map registers is a no-no o Can't DMA into Q-bus memory (only to CPU private memory) o DATIO results in data corruption According to DEC, any device not doing the above should work in a 3xxx. SInce the TK70 is supported and uses the same driver, I'd rule out the driver. Try getting a loan of somebody else's controller (the Dilog SQ703 (or is it 6? one is disk, one is tape, same board, can never remember which is which) and give it a spin... ================================================================================ Note 245.2 VS3200 + Emulex QT13 + Gigastore = ??? 2 of 4 EISNER::CHADWICK "Keith Chadwick, Fermilab" 15 lines 27-JAN-1989 13:45 -< Are you running V5.0-2 ? >- -------------------------------------------------------------------------------- I have observed a similar behaviour with one of the satellite nodes of the LAVC running V5.0-2 that I manage. The node in question is a VAXstation II with a Digidata Gigastore connected to a non-Dilog controller which emulates a TSV05 and uses a patched TSDRIVER (to lengthen the timeouts). Periodically one of the users of this VAXstation runs a job which does a fair amount of I/O (eith DCL COPY or Fortran reads and writes) to the drive, and occasionally this job causes the Gigastore/ACP combination to go crazy - Gigastore flashing all its indicator lights, MSA0ACP generating 90,000+ I/O's per minute, etc. Needless to say the performence of this node suffers. I have attempted STOP/ID's and FORCEX's and have only managed to obtain RWASTed processes as a result. -Keith. ================================================================================ Note 245.3 VS3200 + Emulex QT13 + Gigastore = ??? 3 of 4 EISNER::OSUDAR "John Osudar" 5 lines 27-JAN-1989 15:19 -< yeah, 5.0-2; here's latest word... >- -------------------------------------------------------------------------------- Yeah, we're running V5.0-2. The latest word on this fiasco is: the QT13 (and QT14, for VS3500/3600's) is due for a firmware upgrade that Emulex is about to ship. That's supposed to fix (at least part of) the problem. I'll post any results once I get my hands on the new firmware... ================================================================================ Note 245.4 VS3200 + Emulex QT13 + Gigastore = ??? 4 of 4 EISNER::OSUDAR "John Osudar" 3 lines 4-FEB-1989 11:03 -< Firmware update fixes problem >- -------------------------------------------------------------------------------- We got the update to the Emulex QT13 firmware this week -- and suddenly our Gigastore works in our VAXstation 3200! Now we need to get DEC to fix the TUDRIVER bugs...  ================================================================================ Note 246.0 HSC tape systems -- third party 3 replies EISNER::DEJORDY "Richard DeJordy" 12 lines 10-DEC-1988 14:47 -------------------------------------------------------------------------------- Hello, I'm a software person by nature, and very ignorant about hardware in general. I have a really simple (define/trans=concealed simple stupid) question. We have a VAXcluster with HSC70s. When I read ads for storage a backup systems, what compatibility do I need to look for. (i.e. many say Q-BUS, UNIBUS, etc.) Thanks. Rich ================================================================================ Note 246.1 HSC tape systems -- third party 1 of 3 EISNER::LEDERMAN "Bart Z. Lederman" 18 lines 10-DEC-1988 16:21 -< Not on an HSC unless it's DEC (I think) >- -------------------------------------------------------------------------------- If I'm wrong someone will soon correct me, but as of now I have never seen a backup (i.e., tape) drive which will connect to an HSC controller other than the TA78 produced by DEC. While there are some third-party disk companies which claim to connect to a DEC MSCP controller, I have never seen a tape drive which will connect to an HSC tape channel. If by backup you also include optical disk: the DEC unit might connect to an HSC, but I think it needs it's own controller. Third party units so far only connect to UNIBUS or Q-BUS. The same holds true for video tape storage units (at least all the ones I've seen). So: if you want to put on some kind of tape drive other than a DEC unit we need to know which CPU you want to connect it to so we'll know what kind of bus (UNIBUS, Q-BUS or BI-BUS) it has. And what kind of backup media you want to use. P.S.: I doubt if there is really such as thing as a stupid question. ================================================================================ Note 246.2 HSC tape systems -- third party 2 of 3 EISNER::NAGY "Frank J. Nagy, VAX Guru" 25 lines 11-DEC-1988 11:14 -< 3rd party HSC tapes ARE available >- -------------------------------------------------------------------------------- >> If I'm wrong someone will soon correct me, but as of now I have >> never seen a backup (i.e., tape) drive which will connect to an >> HSC controller other than the TA78 produced by DEC. Both Systems Industries and MicroTechnology Inc. now sell tape systems which connect to the HSCs. In both cases, they sell their own controller boards to plug into the HSC. MicroTechnology currently sells a 200 ips Fujitsu tape drive and expects to soon have 8mm drives. SI sells several systems from 200 ips down (in 9-track) and 8mm drives. Their ads can be found in either Digital Review or Digital News. >> While there >> are some third-party disk companies which claim to connect to a >> DEC MSCP controller, I have never seen a tape drive which will connect >> to an HSC tape channel. Right, no one seems to connect to the DEC tape controller. Systems Industries, Emulex and EMC2 are among those vendors selling disk systems which connect to DEC MSCP controllers (including the HSC boards). >> If by backup you also include optical disk: the DEC unit might connect >> to an HSC, but I think it needs it's own controller. As I remember, the DEC optical disk (RV20) does NOT connect to the HSC. ================================================================================ Note 246.3 HSC tape systems -- third party 3 of 3 EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sa" 16 lines 12-DEC-1988 05:52 -< Installation and Maintenance >- -------------------------------------------------------------------------------- > When I read > ads for storage a backup systems, what compatibility do I need to > look for. I like to look for a few more things. Installation. Will the company I buy it from come out and install it if I am willing to pay their price ? Maintenance. Will the product be compatible with my existing maintenance contract ? What happens if I call out my field service engineer for a problem and it is caused by that product which isn't under my field service agreement ? Everyone should be able to tell you some horror stories where someone tried to install the hardware themselves and failed and then the CPU wouldn't start back up and they had to call field service and claim they don't why the CPU isn't working any more, they hadn't touched the thing.  ================================================================================ Note 247.0 C.Itoh Terminal Firmware Questin No replies EISNER::DEJORDY "Richard DeJordy" 36 lines 10-DEC-1988 15:09 -------------------------------------------------------------------------------- Help! We recemtly have been seeing a bizaar problem with Ingres and C.Itoh terminals, it is definately at least somewhat related to the hardware. Ingres has a key Ctrl-V which will invoke callable EDT from an application. We also have Emacs in which Ctrl-X Ctrl-V is a frequently used command (open another file). Some of our users have programmed Ctrl-X Ctrl-V dev:[directory] to a UDK. Lately, this sequence is being sent to their Ingres sessions. It's one of three possible things, 1) They are accidentally hitting the function key They emphatically say that's impossible - Also, it just started all of a sudden to at least two users. (Despite that, this is still my opinion) 2) The hardware is getting flakey and sending that key's definition for no real reason - not likely but I suppose it is possible. 3) The application is sending some escape sequence which is triggering the terminal to send the definition. Does anyone know if the C.Itoh 200-clones (226?) can receive a sequence which would cause it to send back the actual characters that a key is defined with. Thanks Rich  ================================================================================ Note 248.0 PDP-11/24 CPU BOX (5-1/4") DISSASEMBLY 4 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 6 lines 10-DEC-1988 22:26 -------------------------------------------------------------------------------- The dumb question of the month.... I just pulled an 11/24 CPU box from a customer site. I have not been able slide the CPU chasis all the way out of the black mounting sleeve. I can't find any catches - any ideas? ================================================================================ Note 248.2 PDP-11/24 CPU BOX (5-1/4") DISSASEMBLY 2 of 4 EISNER::RICE "Gary Rice" 11 lines 11-DEC-1988 08:18 -------------------------------------------------------------------------------- > I just pulled an 11/24 CPU box from a customer site. I have > not been able slide the CPU chasis all the way out of the black mounting > sleeve. I can't find any catches - any ideas? I had a similar problem with one of our systems (don't remember which one now). The solution was to pull the sliding foot at the bottom of the cabinett all the way out. This released the Unibus cabinet lock and the chassis would then slide all the way out as well. Gary ================================================================================ Note 248.4 PDP-11/24 CPU BOX (5-1/4") DISSASEMBLY 4 of 4 EISNER::KENNEDY "Terry Kennedy" 5 lines 11-DEC-1988 18:36 -< Well, here is how for 5 1/4" box >- -------------------------------------------------------------------------------- Well, in the small-box '24 I have here (and all 5 1/4" expanders) there are a pair of catches, one on each side. They're the same round ball-on-a- spring as a lot of other DEC stuff. Remember, you have a ground wire and a console ribbon cable which are not long enough to let you remove the box unless you undo them first.  ================================================================================ Note 249.0 Scholar Modem (DF224-AA) power supply source 1 reply EISNER::TINKELMAN "Bob Tinkelman - CCA" 6 lines 13-DEC-1988 10:34 -------------------------------------------------------------------------------- I have a client with a dead Scholar Modem (DF224-AA). It has a bad power supply (part number 29-25043-00). When they called DEC Direct, they were given (eventually) a quote for over $200! Is there any reasonable second source for this? Or should they just throw out the modem? ================================================================================ Note 249.1 Scholar Modem (DF224-AA) power supply source 1 of 1 EISNER::GRANZOW 12 lines 13-DEC-1988 11:22 -< POWER SUPPLY COST >- -------------------------------------------------------------------------------- < Note 249.0 by EISNER::TINKELMAN "Bob Tinkelman - CCA" > -< How much should a power cord cost? >- - -------------------------------------------------------------------------------- > 1) Once the AC gets into the cabinet, does each drive > have a separate power supply? Yes - they're just normal RA82's in there... > 2) Can any one of the drives be repaired/replaced without > having to bring down the entire array. Unless you have an unusually fumble-fingered service droid, yes. I know of one case where the droid extended the top two drives and the whole SA482 fell over... ================================================================================ Note 250.2 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 2 of 27 EISNER::RENES "Stuart Renes, DFWLug Chair" 6 lines 14-DEC-1988 11:11 -< I concur >- -------------------------------------------------------------------------------- I can concur with Terry (.1). We have an SA482 stack and have had to repair one drive (powered down) while the others are spinning away! Stuart. ================================================================================ Note 250.3 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 3 of 27 EISNER::COY "Dale E. Coy" 4 lines 14-DEC-1988 13:07 -< A second power controller might be nice. >- -------------------------------------------------------------------------------- You could potentially buy a second (DEC or 3rd party) power controller and eliminate that as a single point of failure. I've done that on equipment other than the SA, for the same reasons. ================================================================================ Note 250.4 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 4 of 27 EISNER::MCCORMICK_B "Bob McCormick" 6 lines 14-DEC-1988 19:46 -< Just plug them in... >- -------------------------------------------------------------------------------- I've brought the cables right out of the cab to dedicated outlets per drive (for both TU81 & RA8x) series ... and bypassed the DEC controller... ('cuz I had lots of 20amp isolated ground circuits but didn't have the 30amp twist lock available) ================================================================================ Note 250.5 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 5 of 27 EISNER::BYRNE_C "Charlie Byrne" 18 lines 14-DEC-1988 20:42 -< Power Supply = Power Controller? >- -------------------------------------------------------------------------------- >EISNER::KENNEDY "Terry Kennedy" 11 lines 14-DEC-1988 02:04 >> 1) Once the AC gets into the cabinet, does each drive >> have a separate power supply? > Yes - they're just normal RA82's in there... ----------------------------------------------------------------- > < Note 250.3 by EISNER::COY "Dale E. Coy" > > -< A second power controller might be nice. >- > You could potentially buy a second (DEC or 3rd party) power controller > and eliminate that as a single point of failure. Wait a second - from these replies am I to infer that the AC goes into the cabinet to a "power controller", and then the drives (each with separate integrated power supplies) plug into that? Is the power controller nothing more than a distribution unit (i.e. essentially a passive device) (i.e. something I really don't need to worry about as single point of failure)? AdvThanksance ================================================================================ Note 250.6 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 6 of 27 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 20 lines 14-DEC-1988 22:24 -< Power controller LOW risk >- -------------------------------------------------------------------------------- >> Is the power controller nothing more than a distribution >> unit (i.e. essentially a passive device) (i.e. something I >> really don't need to worry about as single point of failure)? I haven't looked carefully at THE controller in that cabinet, but can assume it is like all the rest of Dec's. It has circuit breakers and outlets to split the power up for 120v loads. There probably is a power control relay with a switch that can be set for LOCAL - OFF - REMOTE, and in the remote position it is controlled by the daisy-chaained power control cable (this is how all 13 expansion cabinets in your 11/70 based monster all turn on together with the front key). In a disk application this switch can be TAPED into the LOCAL position. You still have the breakers if some FE is electrocuting himself and you need to chop power fast. Those switches are a hazard as they are EASY to bump off. The power controller IS a single point of failure, but does so little and is so reliable it can be viewed as an extension of the breaker panel on the wall, and will give you about as little trouble, so I would ignore it (if you tape the damned switch...). ================================================================================ Note 250.7 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 7 of 27 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 11 lines 14-DEC-1988 23:51 -< 3 Phase, but all single phase use >- -------------------------------------------------------------------------------- As I recall (and someone correct me if I'm mistaken), the SA482 hooks up to 3-phase power, though it doesn't use the 3-phase characteristic for anything. Keeping in mind that 208/3phase has 120 V on each leg, the first and fourth drives draw power from the X leg, the other two draw power from the Y and Z legs, respectively. This can be an annoyance in that it is easy to overload the X leg on a computer room power distribution system. Beyond that, as stated previously, they're just 4 RA82s with the front covers removed. ================================================================================ Note 250.8 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 8 of 27 EISNER::MERRIMAN "George Merriman - CCA/NY" 10 lines 15-DEC-1988 11:56 -< Not so sure about the power controller >- -------------------------------------------------------------------------------- > The power controller IS a single point of failure, but does so little > and is so reliable it can be viewed as an extension of the breaker > panel on the wall, and will give you about as little trouble, so I > would ignore it (if you tape the damned switch...). I'm not familiar with the particular power controler used with the SA482, but I've seen a fair share of trouble with this type of beast, particularly with the contactor used to implement the remote control feature. I would prefer to run the drives off separate branch circuits, but then I'm weird. ================================================================================ Note 250.9 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 9 of 27 EISNER::COY "Dale E. Coy" 11 lines 15-DEC-1988 14:16 -< Acceptable risk???? >- -------------------------------------------------------------------------------- The high reliability/availability system I had in mind uses separate power distribution for 2 separate CPUs and their associated peripherals. The main power line coming into town is the "single point of failure". We could get rid of that, but... Anyhow, it's clearly a question of how far back you want to push the risk. Personally, I'm skeptical of a solution which depends on anybody (service-person or whatever) who routinely opens the cabinet being careful to NOT accidentally trip the ONE ganged circuit breaker on the power controller. But, as long as the risk is understood and accepted, that's what's important. ================================================================================ Note 250.10 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 10 of 27 EISNER::FULLER_S "Stuart Fuller" 12 lines 15-DEC-1988 16:19 -< Just try and find that breaker! >- -------------------------------------------------------------------------------- Actually, if you're concerned about someone accidently hitting hte breaker on the SA482 cab, try finding it sometime. I know for a fact that the breaker is half buried under the cabling! As for the layout of the 3 phases used, there's a simple fix for overloading one's X leg. Stagger your breakers in the distribution panel - separate each 3 phase breaker with a 1 or 2 pole breaker. This way, the first breaker sees X-Y-Z, while the next 3 phase breaker see Y-Z-X, and so on. Stu ================================================================================ Note 250.11 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 11 of 27 EISNER::MERRIMAN "George Merriman - CCA/NY" 17 lines 15-DEC-1988 17:17 -< overcurrent protection must assume worst case >- -------------------------------------------------------------------------------- > As for the layout of the 3 phases used, there's a simple fix for > overloading one's X leg. Stagger your breakers in the distribution > panel - separate each 3 phase breaker with a 1 or 2 pole breaker. This > way, the first breaker sees X-Y-Z, while the next 3 phase breaker see > Y-Z-X, and so on. This assumes your electricians always maintain the same wiring convention between the breaker poles and the receptical poles, and the power controller is always wired into the plug the same way. It also assumes that the SA482's are the only significant load on the distribution panel -- if you have one or two pole breakers to install in the panel who is to say they won't unbalance the load somewhere? In any case, the overcurrent protection on the three phase branch circuit will always need to accomodate two RA82's. ================================================================================ Note 250.12 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 12 of 27 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 53 lines 15-DEC-1988 19:58 -< Balancing your Breaker panel >- -------------------------------------------------------------------------------- Balancing the load on the 3 phase legs is a good idea, but worrying about breaker placement or wiring color sequence initially is needless hassle. Just read what is actually happening. The electrician might always do the colors in the same order at the breakers, only to have the outlet end wired randomly by a co-worker. The only time actual phase rotation is relevant is if you have a 3-phase motor and expect it to spin in a particular direction. RP04s cared, but not much other Dec stuff that I've run into does. If you are really lucky, you may have metering on your fancy power conditioning system. Otherwise grab your snap-on Amprobe type meter and read all 3 phase legs feeding the panel when everything is running. Often times they will be randomly close enough. If not, do a more detailed study of the loads. You need to know which breakers are fed from which feeder wire. You might assume the pole positions run ABCABCABC... starting at the top, where A-B-C are the feeder cable lugs left to right, but I would look for a circuit diagram label, or better yet find some empty slots and see which pole feeds which slots and count off from there. Every 3rd row will be the same phase (assuming it is a 3 phase panel). With the Amprobe, read the amps at each breaker pole and make a list noting which are on the same multiwire branch circuit sharing a common neutral. Normally these will be 2 or 3 pole breakers with a common trip handle, BUT if the circuit is feeding only 120v loads, may legally be fed from seperate single pole breakers as long as they are on different phase legs. Find unbalanced multiwire branch circuits that have their unbalance adding to the panels unbalence, and rotate the phase legs such that the circuits unbalance subtracts from the panels. This approach is simplest as phase rotation is maintained, and panel directories still have the same breaker in the same place. If there aren't enough of these, or if all else fails rearrange the panel totally. You don't need anywhere near perfect balance, but it is a little stupid to have 85 amps on one leg, 37 on another and 5 on the third. But if that panel were fed with a 225 amp feed, I would be less concerned than if it were a 100 amp feed, and thus nearer capacity on that leg. MY Amprobe is in my desk 2 feet from here, and I still haven't checked the balance in out new computer room since we moved earlier this year, BUT I KNOW that the panel is HUGH compared to our total load. ================================================================================ Note 250.13 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 13 of 27 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 33 lines 15-DEC-1988 21:01 -< Power controllers CAN fail (!) >- -------------------------------------------------------------------------------- >> The power controller IS a single point of failure, but >> does so little and is so reliable it can be viewed as an extension >> of the breaker panel on the wall, and will give you about as little >> trouble, so I would ignore it (if you tape the damned switch...). In general I would agree with you. However, I had a DEC 861-F power controller fail in a spectacular manner. Inside is a small 24 volt power supply, a reed relay, some transient suppressors and a few other components. As the capacitor in the power supply ages and starts to fail, the output voltage starts to drop. At some point, it drops low enough for the reed relay to start to chatter. The mass of the large contactor (relay) is enough so that it doesn't drop out, but some large voltage spikes are being generated by the constant opening and closing of the circuit. Eventually, one of the transient suppressors is triggered and becomes a short circuit, putting the contacts of the reed relay directly across the AC line. Between the transient suppressor and the reed relay, a large belch of smoke is emitted from the controller. If you are lucky, nobody will have heart failure while watching this. The transient suppressors and capacitor cost just a few bucks, but the DEC custom reed relay was about $30-50. The newer controllers may have been redesigned to eliminate this failure mode, but I have no way of checking. I also have never heard of this happening to anyone else, so it DOES seem to be a risk item. Alan Frisbie ================================================================================ Note 250.14 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 14 of 27 EISNER::MERRIMAN "George Merriman - CCA/NY" 26 lines 15-DEC-1988 21:05 -< an electrician's humble opinion >- -------------------------------------------------------------------------------- > You don't need anywhere near perfect balance, but it is a little stupid > to have 85 amps on one leg, 37 on another and 5 on the third. A three-phase four-wire distribution system MUST be capable of operating at ANY level of imbalance as a normal and continuous mode of operation. Consider such a distribution with only single phase individually switched branch circuits -- a typical small office situation. What is keeping us from turning on only the circuits on one leg, or on only two of the legs? If the distribution were not capable of sustaining such a condition we would all have little NEC placards next to our light switches saying "Don't turn off these lights unless Bill down the hall does likewise". Do you know which leg your lights are on? What about your wall outlets? The only reason to worry about balance is if the load on one feeder phase exceeds its capacity. In a good installation, where the sum of the branch circuit breaker ratings is less than or equal to the feeder circuit rating you don't have to worry about it. If your computer room panel is so stuffed with breakers that you have to keep shuffling them around to keep from tripping the main you probably need a bigger panel! The reason installations are designed to be balanced is to reduce their costs -- a well balanced installation will use less copper. Once the installation is done, however, you can do pretty much what you want short of double-linking fuses. ================================================================================ Note 250.15 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 15 of 27 EISNER::KENNEDY "Terry Kennedy" 21 lines 15-DEC-1988 21:57 -< Power controllers / Phase balancing >- -------------------------------------------------------------------------------- > The transient suppressors and capacitor cost just a few > bucks, but the DEC custom reed relay was about $30-50. Alan, I'm surprised! The whole assembly is DECmailer-able for $20.00 - I should know, I just sent two in... Regarding 3-phase power, as other replies have pointed out, things tend to naturally balance out. For our center, we have a substation for our data which converts the 22,000 volt feed into 3-phase (120/208). The main panel for data has a 3-phase 1500-amp breaker. From there power is routed to 3-phase 200-amp breakers in the same panel. Then it is routed to individual 200-amp panels for each cluster of systems, where there is another 3-phase 200-amp breaker (the 'main' for that panel, and a variety of breakers as required for the individual equipment. Now, just by random installation, each panel is balanced within about 20%. The 1500-amp feed is balanced within 10%. This is all by pure chance. If there is enough current drawn to trip one of the 3-phase breakers, then you would know that you have an imbalance that needs correcting. However, you would still need an Amprobe meter to determine *which* phase, as all gthree poles are required to trip when one or more legs is overloaded. ================================================================================ Note 250.16 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 16 of 27 EISNER::LEDERMAN "Bart Z. Lederman" 12 lines 16-DEC-1988 15:48 -< Imbalance can cost you. >- -------------------------------------------------------------------------------- Where I am, we do take some precautions to balance the loads between phases to a reasonable amount (say 10-20%). One reasons is that everything is on our UPS and our power feeders, and WE have to pay for the extra wire, UPS capacity, etc., if the phases are not balanced. Even if this wasn't the case, I think it's a good idea to take some precautions to get some balance. I also think it's a very good idea to double or triple check the wiring before anything is powered up no matter how good you think your electrician is. I've seen a lot of bad wiring by licenced electricians here. ================================================================================ Note 250.17 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 17 of 27 EISNER::MCCORMICK_B "Bob McCormick" 13 lines 17-DEC-1988 11:32 -< Some benefit from controller's guts >- -------------------------------------------------------------------------------- Power controllers do have some limited power conditioning stuff in addition to duplex outlets and a contactor (depending on the model). I use one in the bottom of a rack cab for a ham radio 70cm repeater on top of a local mountain at a TV transmitter site... where there's high lightning activity. A lot of two way stuff gets blown away ... but my power controller plus some other telephone and coax lightning suppressors have (knock on wood) protected my gear through the years... BTW I have an expensive CMOS based "repeater controller", which as you can guess, wouldn't tollerate the lightning ================================================================================ Note 250.18 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 18 of 27 EISNER::BYRNE_C "Charlie Byrne" 6 lines 19-DEC-1988 22:10 -< Any way to get MTBF info (besides experience)? >- -------------------------------------------------------------------------------- I home now and don't have Sys and Opts catalog handy, but I don't suppose by any chance that there is MTBF type info available for DEC hardware, particularly for the equipment we've been discussing? Hmm no, I didn't think so. Or is there? ================================================================================ Note 250.19 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 19 of 27 EISNER::FULLER_S "Stuart Fuller" 1 line 19-DEC-1988 23:33 -------------------------------------------------------------------------------- DEC doesn't publish MTBF info. ================================================================================ Note 250.20 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 20 of 27 EISNER::HARVEY "Jack Harvey" 18 lines 20-DEC-1988 05:32 -< MTBF isn't very useful >- -------------------------------------------------------------------------------- > DEC doesn't publish MTBF info. For good reason, in my opinion. Computed MTBF (based on design rather than experience) is about as safe as predicting a program is bug-free. With the kinds of MTBF we expect of modern computer devices, by the time there is enough *experience* with a product to establish its MTBF, it's obsolete. Of course, bad products start shipping and failing fast. (Proof of the uselessness of computed MTBF.) Then Field Service starts hurting (too) and screams at Engineering to fix the design. In my opinion, third party (such as CDC or Fujitsu) MTBF claims are just hopes - they can't predict failures like the famous RA81 glue problem. A manufacturer's track record isn't any good either - the engineers who created an excellent track record are now in management and a new set of junior engineers are just learning how not to screw up. You pick a disk, buy and hope. ================================================================================ Note 250.21 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 21 of 27 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 58 lines 4-JAN-1989 01:02 -< A big 'WHY' for phase balance. >- -------------------------------------------------------------------------------- Perhaps a very practical way to look at phase balancing is that it wastes the LEAST amount of energy as heat from the breaker panel's feeder cables. If you think these are sized so large as to waste little, consider the following. A commonly used insulation such as XHHN is rated for 90 degrees C. service, and it will often be installed to run at that temperature if the total voltage drop allowed for feeders is not exceeded. Else a slightly larger wire will be used to keep the voltage drop in line for the longer run, the wire will run cooler, but the total loss for the feeder may be the same as for the short hot one! One 'engineers' to cut installation cost, and the two limits one bumps into are feeder % voltage drop, and conductor temperature. One can assume a 'properly' engineered feeder has substantial loss. At full load you have to be balanced or you will pop your supply breaker (or should...), but at partial load, or even at 1/3 load (where one could get everything onto just one phase leg) you don't have to be balanced. But lets look at the ratios of feeder loss (waste heat) to balance. For simplicity lets use 6 equal 120v resistance heaters that draw an amp each. This way there is NO power factor and they can be divided equally onto 1, 2, or 3 phase legs. The feeder is 4 equal size wires - neutral and 3 phase legs. Watts wasted in any wire is AMPS squared X the wire's resistance (R). Since we are looking at the ratio between difference balance conditions, R will simply drop out. And, conveniently, in the 3 cases chosen, current through any wire will either be zero, or equal to that through any of the others carrying current. So in these 3 simple cases we look at AMPs squared and the number of current carrying wires to get our ratios. The 3 cases are: 1) 6 heaters on 1 phase leg (current obviously equal on that phase and the neutral wire). 2) 3 heaters on each of 2 phase legs (current is equal on the 2 phase legs, and not so obviously is also equal on the neutral! - intuitively if one added 3 MORE heaters to the previously unused leg the neutral current would go to a balanced 0 amps, and all the neutral change would have come from that 3rd leg...) 3) 2 heaters on each of 3 phase legs (current is equal in the 3 phase legs, and 0 in the neutral). Now lets do wires times amps squared for the same 3 cases: 1) 2 wires x 6 squared = 72 2) 3 wires x 3 squared = 27 3) 3 wires x 2 squared = 12 That is 6:1 loss ratio from the worst to the best configuration! And remember that extra heat not only does not power your equipment, but also must be removed by your air-conditioning using yet more power. ================================================================================ Note 250.22 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 22 of 27 EISNER::MERRIMAN "George Merriman - CCA/NY" 24 lines 5-JAN-1989 18:58 -< 5/6th of what? >- -------------------------------------------------------------------------------- > < Note 250.21 by EISNER::BRUCE "Barton F. Bruce - CCA/MA" > > -< A big 'WHY' for phase balance. >- > That is 6:1 loss ratio from the worst to the best configuration! So we can save 5/6th of this power loss by balancing the loads, but, I ask myself, 5/6th of what? > For simplicity lets use 6 equal 120v resistance heaters that draw an > amp each. This way there is NO power factor and they can be divided > equally. An estimate for the actual power loss can be made by taking advantage of the fact that that the power dissipated in the feeders is equal to the product of the current in the circuit and the voltage drop in the feeders. According to the 1987 N. E. C., Article 215-2(b), the voltage drop in feeder circuits should be limited to 3% of the nominal service voltage, or 3.6V in this case. 3.6V * 6 A = 21.6W, so the most we could save is 5/6 * 21.6W = 30W, less than half the name-plate power rating of the VT220 I'm using to write this note. As a more realistic example, the most power that could be saved on a 100A service is 300W, less than half the name-plate power rating of a LN03. ================================================================================ Note 250.23 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 23 of 27 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 10 lines 5-JAN-1989 23:37 -< small peanuts >- -------------------------------------------------------------------------------- >> As a more realistic example, the most power that could be saved >> on a 100A service is 300W, less than half the name-plate power rating >> of a LN03. Sure! but just what does 300 watts cost per year? And lets simply ignore the extra energy necessary to then cool the extra heat in the building, and merely use 10 cents per killowatt hour. Except on leap years, that will cost you $262.80. Enough peanuts make a big pile. ================================================================================ Note 250.24 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 24 of 27 EISNER::MERRIMAN "George Merriman - CCA/NY" 25 lines 6-JAN-1989 16:57 -< it all depends . . . >- -------------------------------------------------------------------------------- > Except on leap years, that will cost you $262.80. Let's see now -- at a burdened labor rate of $75/hr (not unreasonable for technical help) if you spend more than 3 1/2 hours/yr. fooling around with your power you haven't saved anything. If you are in a situation where you have to get a union electrician and a helper in for a day it might take years to break even, assuming your configuration doesn't change and you have to get them back again. And remember, the costs to be saved assume you are able to go from worst case to optimal case by balancing the load on a fully loaded marginally engineered installation -- your actual mileage will probably be much less than this. > And lets simply ignore the extra energy necessary to then cool the > extra heat in the building . . . because it is less than you would need if you hired another operator. I'm not saying don't do it -- I'm only trying to come up with some concrete numbers so people can put it all in perspective. If your computer room runs fine on three legs of 100A, it's probably not worth it. If you are the building engineer in a 40 story hi-rise, you probably worry about things like this from time to time. ================================================================================ Note 250.25 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 25 of 27 EISNER::MERRIMAN "George Merriman - CCA/NY" 6 lines 6-JAN-1989 17:13 -< a correction >- -------------------------------------------------------------------------------- > > And lets simply ignore the extra energy necessary to then cool the > > extra heat in the building . . . > because it is less than you would need if you hired another operator. Sorry, I was confusing BTU/hr with Watts. This is about the same as four extra bodies in your computer room. ================================================================================ Note 250.26 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 26 of 27 EISNER::MCCORMICK_B "Bob McCormick" 11 lines 6-JAN-1989 19:39 -< Towering Inferno + 19,751 MegaWatts >- -------------------------------------------------------------------------------- > If you are the building engineer in a 40 story hi-rise, > you probably worry about things like this from time to time. Gee ... wasn't that the problem they had in the Towering Inferno? Oh, that was 140 stories! Here in the north east we just broke a powre consumption record: 19,751 MegaWatts ... and there are some concerns for the next few months about the power distribution system's ability to meet the demand. The last time they had this problem (during the summer heat) we had to replace all the power supplies in our systems! ================================================================================ Note 250.27 Power Supplies/Controllers- AC Phasing- Cabinets- SA482 27 of 27 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 10 lines 3-MAR-1989 19:30 -< Good reading about Neutral Currents >- -------------------------------------------------------------------------------- There is an excellent article in the current (Winter 1988) issue of Computer Technology Review: "Power System Problems from High Harmonic Neutral Currents" by Thomas M. Gruzs of Liebert Corp. In it, he shows that the neutral feeder wire for 3-phase power should be designed to accomodate neutral currents up to 1.73 times the phase current. This is because of the high harmonic current distortion present in computer power supplies.  ================================================================================ Note 251.0 Common VAX-11/780 problems 15 replies EISNER::CONROY "Alan Conroy" 5 lines 16-DEC-1988 14:43 -------------------------------------------------------------------------------- We are about to upgrade one of our 11/750s to an 11/780. Having never been around an 11/780 before, my question is: what is the most likely cause of system failure? For example, on 11/750s the most common problem is that the translation buffer board goes bad (at least on all the 11/750s I've seen). I just want to know what to expect... ================================================================================ Note 251.1 Common VAX-11/780 problems 1 of 15 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 17 lines 16-DEC-1988 19:00 -< Cache, then memory, then Unibus adaptor. >- -------------------------------------------------------------------------------- We have had an 11/780 in use pretty much continuously since 1982; the things which have caused us the most problems have been the cache boards, the memory cards (terrible with 16K chips, much better with the 64K chips, no problems in over a year since we went to the 256K chip cards), and the Unibus adaptor card set. (The last might not affect you; however, we run several Megatek 7250 color monitor controllers right off the Unibus, and really exercise the pi$$ out of it -- our Field Service group knows enough to bring along several sets of cards whenever we report a Unibus card problem, usually one of three will both pass the diagnostics and run OK with the Megateks. We have the same problem with the Unibus adaptor in our 8650, as well. Megatek does (or did) have a marketing agreement with DEC, so if you have this problem and go high enough, someone who knows about it will eventually tell your DEC service center that it's OK to select card sets -- I dunno if this works for third-party service.) ================================================================================ Note 251.2 Common VAX-11/780 problems 2 of 15 EISNER::KENNEDY "Terry Kennedy" 22 lines 16-DEC-1988 19:06 -< Very reliable >- -------------------------------------------------------------------------------- Well, we bought our 780 used in June '86. It was built in 1983 and had one owner before us (Coleco developed the 'Adam' software on it). We got complete service logs for it, and it had been under DEC maintenance since it was installed. A perusal of those records (and ours) shows: 2 Modular power supplies (in CPU) [just failed] 1 LP11 controller [out of revision, scared VMS] 1 UDA50A controller [dead port when we added third drive] We added floating-point and replaced 4Mb of DEC memory with 16Mb of EMC memory in July '86. Neither have given any trouble. We have had problems with the tape (TU78) and drives (RA81 glue problem), but the CPU has been rock solid. Our lifetime average with it has been about 98.7% uptime. Of course, you can go broke feeding it power and air conditioning, and DEC wants an arm and a leg for maintenance. My advice is to make sure you are getting a 'new' 780 (FCC cabinet, etc) built after '82. Also get a whole spare 780 (only costs about $5000 these days) to use to board-swap problems. If you have the in-house expertise, you can save a *lot* that way. Post a reply here if you want to hear more... ================================================================================ Note 251.3 Common VAX-11/780 problems 3 of 15 EISNER::ROSEBERRY "Bert Roseberry-that Oak among Sa" 14 lines 19-DEC-1988 06:17 -< LSI and Ethernet Controller >- -------------------------------------------------------------------------------- Our main problems have been with the LSI-11 console subsystem in the VAX 11/780. One other time we had a problem with the ethernet controller going bad. Other than that, the system has been pretty reliable. I did find at a previous site with AC units that went down all the time, if you get large temperature fluctuations, the machine will start acting "flacky" and finally Digital field service will have to come out and pull some boards and run a pencil eraser over them and put them back in. ================================================================================ Note 251.4 Common VAX-11/780 problems 4 of 15 EISNER::CONROY "Alan Conroy" 2 lines 19-DEC-1988 12:48 -< Could you elaborate, please >- -------------------------------------------------------------------------------- > Our main problems have been with the LSI-11 console subsystem in > the VAX 11/780. ================================================================================ Note 251.5 Common VAX-11/780 problems 5 of 15 EISNER::CONROY "Alan Conroy" 17 lines 19-DEC-1988 12:57 -< Why FCC? >- -------------------------------------------------------------------------------- >DEC wants an arm and a leg for maintenance. My advice is to make sure you >are getting a 'new' 780 (FCC cabinet, etc) built after '82. Also get a >whole spare 780 (only costs about $5000 these days) to use to board-swap >problems. If you have the in-house expertise, you can save a *lot* that >way. Post a reply here if you want to hear more... Fortunately (or unfortunately?) we don't have DEC maintenance. The local CDC office has purchased a whole-spare '780 - mainly to support us. I may (keeping my fingers crossed) be out of system management and into programming in about a month - and my replacement is almost a hardware moron. I had often toyed with the idea of self-maintenance over the years, but never got up the nerve to suggest it. I am curious as to why the FCC cabinet. We don't care about EMI here. Is there some other reason to need the FCC version? Further, this is probably a one or two year situation, with the other '750 being replaced with a '780 and then both '780s being replaced with '785s when the price comes down on them. I hope that 24 months of non-FCC cabinets won't be a problem! ================================================================================ Note 251.6 Common VAX-11/780 problems 6 of 15 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 8 lines 19-DEC-1988 13:13 -< Insurance, if nothing else >- -------------------------------------------------------------------------------- RE: < Note 251.5 by EISNER::CONROY "Alan Conroy" > -< Why FCC? >- | I am curious as to why the FCC cabinet. We don't care about EMI here. You should. Should any of your neighbors (even if they're far away) complain and the EMI can be traced to you, you will pay the cost of its containment. ================================================================================ Note 251.7 Common VAX-11/780 problems 7 of 15 EISNER::CONROY "Alan Conroy" 0 lines 19-DEC-1988 17:09 -< Suggestion noted (sorry for the pun!) >- -------------------------------------------------------------------------------- ================================================================================ Note 251.8 Common VAX-11/780 problems 8 of 15 EISNER::KENNEDY "Terry Kennedy" 6 lines 19-DEC-1988 17:59 -< Why FCC >- -------------------------------------------------------------------------------- Actually, I prefer containing the complainers... Seriously, it had nothing to do with the FCC-ness of the system - I've just seen too many 780's where the reseller said 'yep, sure, all latest revs' where the system was in an old box. Most vital 780 revs came out before the box change, so FCC = better chance of in-rev CPU parts. ================================================================================ Note 251.9 Common VAX-11/780 problems 9 of 15 EISNER::LEDERMAN "Bart Z. Lederman" 12 lines 19-DEC-1988 20:14 -< Erasers are not a good idea. >- -------------------------------------------------------------------------------- FYI: I've seen DEC and other service personnel use pencil erasers to clean the fingers on PC boards. They are sometimes rather chagrinned when I show them an old DEC service document which states very specifically that you should NEVER use an eraser to clean PC board contacts because it takes off the gold plating and leaves a residue (and this was when the gold plating was several times thicker than it is now). We have 785's, which are basically similar to 780s. The only time we have had any really serious problem is immediately after DEC ECOs the machine, such as replacing CI code PROMS. Other than that, they have been quite reliable. ================================================================================ Note 251.10 Common VAX-11/780 problems 10 of 15 EISNER::KENNEDY "Terry Kennedy" 11 lines 20-DEC-1988 07:07 -< Suggested means to clean gold contacts >- -------------------------------------------------------------------------------- > FYI: I've seen DEC and other service personnel use pencil erasers > to clean the fingers on PC boards. They are sometimes rather > chagrinned when I show them an old DEC service document which states > very specifically that you should NEVER use an eraser to clean... The 'approved' method is "goldwipes" from the Texwipe people (Saddle River, NJ) This part should be stocked in FS toolkits along with the related head-cleaning pads. A side effect of erasing is that the small rubber particles get in the backplane and cause all sorts of havoc... ================================================================================ Note 251.11 Common VAX-11/780 problems 11 of 15 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 7 lines 20-DEC-1988 17:38 -< LSI-11 spares ! >- -------------------------------------------------------------------------------- We had some problems with the LSI-11 on our 780 (after 5 year of service) and found it difficult to get **GOOD** replacement parts for it (the LSI-11). Otherwise it was a GREAT machine, the most reliable in my 12 years of computing ! I agree that they are dinasours that eat power and A/C and I'm certain that you could justify a uVAX II on those cost savings alone ! ================================================================================ Note 251.12 Common VAX-11/780 problems 12 of 15 EISNER::FULLER_S "Stuart Fuller" 9 lines 20-DEC-1988 22:56 -------------------------------------------------------------------------------- re: .0 The bottom line here is that there are no common 11/780 problems. EVERYthing breaks at one time or another, and the 11/780 is no exception. There just isn't one part that tends to break more frequently than any other. Stu ================================================================================ Note 251.13 Common VAX-11/780 problems 13 of 15 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 11 lines 21-DEC-1988 10:25 -< uVAX not always a better deal. >- -------------------------------------------------------------------------------- >> I agree that they are dinasours that eat power and A/C and I'm certain >> that you could justify a uVAX II on those cost savings alone ! Well, it depends. If you already have a computer room environment with spare A/C capacity available, the 780 might well be more economic than the uVAX, depending upon how long your bean-counter group allows you to look for financial comparisons. If you can compare costs out to five years or so, the uVAX might win (depending on how much your power costs); but if you are in a cash bind and must show break-even in six months or a year, I don't see how the uVAX would be economic. ================================================================================ Note 251.14 Common VAX-11/780 problems 14 of 15 EISNER::FULLER_S "From the terminal of Stuart Fuller" 9 lines 21-DEC-1988 17:24 -------------------------------------------------------------------------------- >< Note 251.11 by EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" > > -< LSI-11 spares ! >- > > We had some problems with the LSI-11 on our 780 (after 5 year of > service) and found it difficult to get **GOOD** replacement parts > for it (the LSI-11). Otherwise it was a GREAT machine, the most > reliable in my 12 years of computing ! Reliable, except when either of us went on vacation! ================================================================================ Note 251.15 Common VAX-11/780 problems 15 of 15 EISNER::PROVOST 22 lines 22-DEC-1988 11:12 -< 750 => 780 >- -------------------------------------------------------------------------------- We have had a VAX-11/780 since 1979. It is remarkably reliable. We just replaced a VAX-11/750 with a used VAX-11/780 which had been half of a VAX-11/782. One interesting thing to look at: Pull out the floppy drive and check its power supply. Two large capacitors in the p.s. are not accustomed to lying on their sides. They tend to back out of their connectors and short against the metal cab. They should be tie-wrapped in place. This is particularly true if the 780 was shipped. On cost of maintenance: Biggest cost to us was due to the SVAXHALA part number including DEC memory, RK611 controller, and two RK07 drives. The drives had been replaced with RA disks long ago, and the memory with EMC2. By changing the maintenance contract to cover only the CPU and console system, we reduced the cost by 2/3's. We have CDC maintenance, so this was easy. With DEC you may have to maintain some DEC memory. On conversion from VAX-11/750: Some older UNIBUS devices use DATOB instructions for DMA. These will not work on 780's buffered data paths, although the 750 has no problems.  ================================================================================ Note 252.0 CXA16 configuration info needed 3 replies EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 3 lines 20-DEC-1988 14:07 -------------------------------------------------------------------------------- Can anyone supply the switchpack settings (CSR and vector) for the CXA16? (This is the new 16-port, no-modem-control terminal mux for the 3000 series) ================================================================================ Note 252.1 CXA16 configuration info needed 1 of 3 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 8 lines 25-FEB-1989 18:54 -< It's much like the DHV. >- -------------------------------------------------------------------------------- Well, I figured it out (some time ago, actually). Turns out that the switchpacks are just like those on the good 'ole DHV, there are just a few more switches in one of the packs. The next time I have the board out of the machine (soon) I'll make a diagram and post it here, but in the meantime, if you've got a CXA16 and a DHV book, you can likely figure it out, esp. if you know what the CXA's current configuration is. ================================================================================ Note 252.2 CXA16 configuration info needed 2 of 3 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 69 lines 1-MAR-1989 01:00 -< The answer! >- -------------------------------------------------------------------------------- Some time ago I needed to add some options to a MicroVAX 3600 that included a CXA16 terminal mux. The new options required the CXA16 to be moved to a different CSR and vector address, and it turned out that DEC included no documentation on how to do this with the system. (BOO to DEC for continuing to dilute the technical information we receive with our systems!) It turns out that the CXA16 has a pair of DIP switchpacks for CSR and vector selection in nearly the same positions and with nearly the same functions as those on the DHV11, and with the aid of a DHV11 manual (and knowledge of where the CXA16 was presently configured) I was able to figure it out. Anybody else could do the same, but with the following info you won't have to: Orient your CXA16 (on a properly grounded anti-static mat, please) with the component side facing you, the bus connectors "down" (nearest you), and the serial port connections, card handle/extractor, et al, "up" (away from you). \-------------------------/ <- module handle and connectors to | | distribution panel | | | | | CSR VEC | <- approximate loc. of switchpacks | | |____||____|____||____| <- bus connectors Now look in the lower left region of the board. You will find two DIP switchpacks, one with ten switches and one with eight. We will refer to the one with ten switches (the left-hand one) as the CSR switchpack, and to the righthand one as the VEC switchpack. (No nomenclature for these components is stamped, etched, or otherwise marked on the board -- at least not on my sample -- so I have to make up my own.) For CSR assignments: CSR address is controlled by switches CSR-2 through CSR-10. If all of these switches are "open" or "off", the board's CSR address will be 760000. Set the CSR switches, as shown in the following table, "closed" or "on" to add the indicated values to the CSR address: CXA16 switch: CSR-2 CSR-3 CSR-4 CSR-5 CSR-6 CSR-7 CSR-8 CSR-9 CSR-10 octal value: 10000 04000 02000 01000 00400 00200 00100 00040 00020 DHV11 equivalent switches: E58-1 E58-2 E58-3 E58-4 E58-5 E58-6 E58-7 E58-8 E43-1 For example, if switches CSR-4, CSR-7, and CSR-8 are closed, the CSR address will be 762300. For vector assignments: The vector is controlled by switches VEC-3 through VEC-8. If all of these switches are "open" or "off", the board's interrupt vector will be 000 (in theory; I haven't tried it). Set the VEC switches, as shown in the following table, "closed" or "on" to add the indicated values to the interrupt vector: CXA16 switch: VEC-3 VEC-4 VEC-5 VEC-6 VEC-7 VEC-8 octal value: 400 200 100 040 020 010 DHV11 equivalent switches: E43-3 E43-4 E43-5 E43-6 E43-7 E43-8 For example, if switches VEC-4, VEC-5, and VEC-7 are closed, the board will interrupt at vector 320. I don't know what switches CSR-1, VEC-1, or VEC-2 do; I left them as I found them and everything worked. ================================================================================ Note 252.3 CXA16 configuration info needed 3 of 3 EISNER::NORTON "Bill Norton" 16 lines 3-MAR-1989 12:52 -< ...more answer >- -------------------------------------------------------------------------------- >I don't know what switches CSR-1, VEC-1, or VEC-2 do; I left them as I found >them and everything worked. According to the CXA16 installation manual(EK-CXX16-IN-001): VEC-1 ON enables the onboard oscillator VEC-2 ON enables a manufacturing diagnostic test mode VEC-1 must be ON and VEC-2 must be OFF for correct operation Powerup self-test will fail at other settings. CSR-1 is the DHU/DHV mode selector. OFF = DHU mode, ON = DHV mode. Our DS500 seems to like this in DHU mode. Haven't tried DHV mode. Does anyone know what changing this switch does to operating characteristics?  ================================================================================ Note 253.0 Boot 11/23 over Ethernet? 4 replies EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 6 lines 22-DEC-1988 17:28 -------------------------------------------------------------------------------- Has DEC ever produced a set of boot ROMs for an 11/23 (or /23+) that would allow it to boot off of DECNET/Ethernet ? I know 11/24's can. I've been told that the DEUNA has some "smarts" to help start up the boot. Does DEQNA's have same feature (ie. can boot ROM code from 11/24 be modified to boot an 11/23 ?) ================================================================================ Note 253.1 Boot 11/23 over Ethernet? 1 of 4 EISNER::KENNEDY "Terry Kennedy" 13 lines 22-DEC-1988 19:52 -< Yes, it can be done... >- -------------------------------------------------------------------------------- > Has DEC ever produced a set of boot ROMs for an 11/23 (or /23+) > that would allow it to boot off of DECNET/Ethernet ? You're in luck! The KDF11-B3 kit contains new roms, a handy manual with several internal Field Service hotline numbers (800, of course!), and a complete set of diagnostics, all at a *very* reasonable price. (Last time I looked, about $145.00) The manual gives the following switch settings for DEQNA down-line boot: S1-3, S1-5 closed, S1-1,2,4,6 open. Of course, these roms only work in quad 11/23's (M8189), as they are type 68766 24-pin 8K parts. ================================================================================ Note 253.2 Boot 11/23 over Ethernet? 2 of 4 EISNER::KOZAM 5 lines 23-DEC-1988 00:39 -< Would a BDV11 work here? >- -------------------------------------------------------------------------------- > Of course, these roms only work in quad 11/23's (M8189), as they are > type 68766 24-pin 8K parts. Might these ROMS also work with the BDV11? I seem to recall reading that the BDV11 and KDF11-B boot ROMS were similar. ================================================================================ Note 253.3 Boot 11/23 over Ethernet? 3 of 4 EISNER::KENNEDY "Terry Kennedy" 6 lines 24-DEC-1988 03:53 -< I'm not sure >- -------------------------------------------------------------------------------- > Might these ROMS also work with the BDV11? I seem to recall > reading that the BDV11 and KDF11-B boot ROMS were similar. I don't know - I don't have a BDV11 manual. The standard 11/23 boot roms are 2K parts. I seem to recall that a newer BDV11 rev accepts larger parts, but they might be type 2764... ================================================================================ Note 253.4 Boot 11/23 over Ethernet? 4 of 4 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 12 lines 31-DEC-1988 16:54 -< Even 11/23-A boots OK from Ethernet >- -------------------------------------------------------------------------------- >> I made my own set of DEQNA boot ROMs for an old MXV11-A board. It worked fine after I got DEC to fix the ROM on the DEQNA! The engineer in charge of DEQNA fixes sent me early versions of the ROMs, so I don't know which FCO incorporates the fixes. If you need the code to make your own MXV11 (or whatever) boot ROMs, let me know and I'll try to dig it out. Alan Frisbie -- (213) 256-2575  ================================================================================ Note 254.0 SMP and CIBCI (CI750) don't mix? 18 replies EISNER::ANGLEY "Kevin Angley" 14 lines 30-DEC-1988 16:06 -------------------------------------------------------------------------------- Does anyone have any experience with a SMP system (where there are > 2 processors) that also has a CIBCI (which is nothing more that a CI750 with a BI adapter)? We are having problems that when three CPU's are busy, the maintenance timer on the CI times out, and reaches 50 errors quickly. When all three CPU's are running, but not loaded down, there is no problem. I suspect that if this was a BI CI (whatever that is called), there would be no problem. Comments? ================================================================================ Note 254.1 SMP and CIBCI (CI750) don't mix? 1 of 18 EISNER::KENNEDY "Terry Kennedy" 6 lines 30-DEC-1988 22:53 -< On what model CPU? >- -------------------------------------------------------------------------------- > We are having problems that when three CPU's are busy, the maintenance > timer on the CI times out, and reaches 50 errors quickly. When all > three CPU's are running, but not loaded down, there is no problem. Exactly what sort of CPU mudules are involved? This wouldn't happen to be a homebrew 8360, would it? ================================================================================ Note 254.2 SMP and CIBCI (CI750) don't mix? 2 of 18 EISNER::ANGLEY "Kevin Angley" 6 lines 1-JAN-1989 12:36 -< I would say its a BI based machine of some type >- -------------------------------------------------------------------------------- > happen to be a homebrew 8360, would it? Well ... I can't exactly say. 8360 is not exactly the right answer to your question, but .... well, I can't say ... does that answer your question? ================================================================================ Note 254.3 SMP and CIBCI (CI750) don't mix? 3 of 18 EISNER::KENNEDY "Terry Kennedy" 7 lines 1-JAN-1989 17:46 -< Well, it doesn't work on >2 82x0 CPU modules >- -------------------------------------------------------------------------------- > Well ... I can't exactly say. 8360 is not exactly the right answer to > your question, but .... well, I can't say ... does that answer your > question? Well, to put it another way, if you *are* using multiple (>2) 8200 or 8250 CPU modules, it won't work. Just sometimes, DEC has a valid (tech- nical) reason for not offering certain configurations... ================================================================================ Note 254.4 SMP and CIBCI (CI750) don't mix? 4 of 18 EISNER::BURNET "John Burnet" 10 lines 1-JAN-1989 19:59 -< 83x0 (x > 5) -- why not? >- -------------------------------------------------------------------------------- Re. < Note 254.3 by EISNER::KENNEDY "Terry Kennedy" > > Well, to put it another way, if you *are* using multiple (>2) 8200 or > 8250 CPU modules, it won't work. My curiosity is aroused. Just why won't it work? (One reason I'm curious is that we're about to upgrade an 8200 to an 8350, and we'd like to know if there is a further *in-cabinet* upgrade path available for us in the future -- whether supported by DEC or not.) ================================================================================ Note 254.5 SMP and CIBCI (CI750) don't mix? 5 of 18 EISNER::KENNEDY "Terry Kennedy" 23 lines 1-JAN-1989 23:28 -< Explanation >- -------------------------------------------------------------------------------- > My curiosity is aroused. Just why won't it work? Now that's an interesting question. You're both asking why something that 'doesn't exist' doesn't work... Oh well... On 'bigger' BI machines, the BI is *just* an I/O bus, like the Unibus on the 780. On the KA820-based CPU's, it is *also* the primary system bus. Memory references arbitrate to a higher priority level than I/O requests, thus if you have enough CPUs going to clog the BI with mem- ory requests, I/O devices will end up being ignored for long periods of time. In the case of the CI, it has a sanity timer which expires before the board is serviced, and so it goes... You *might* be able to hack the BI priority of the CI up, (and that of memory down) by fiddling around in the inputs to the BIIC, but you would need the BIIC data sheet (*rather* hard to come by) as well as knowing that you're just moving the problem, and will probably just crash the CPU which was held up for the CI, anyway. I seriously believe DEC tried hard to make this work. After all, they could use the revenus generated by an 8350 --> 8360 upgrade, as well as retaining goodwill by offering an in-cabinet upgrade. Unfortunately, with some hardware it just isn't possible... ================================================================================ Note 254.6 SMP and CIBCI (CI750) don't mix? 6 of 18 EISNER::ANGLEY "Kevin Angley" 21 lines 2-JAN-1989 20:11 -< If it doesn't work, I want a refund on my 8300! >- -------------------------------------------------------------------------------- > I seriously believe DEC tried hard to make this work. After all, As I understand it, DEC DID have it working, and developed SMP using it. I also know for a fact that a clustered 8360 or 8370 did in fact work for them ... but, they had a BI CI and not a CIBCI. So, I am puzzled as to why they could get it working and you are saying that it is technically impossible. Do you speak from experience, Terry, or is your explanation an educated speculation? Also, the 8200 Technical Manual says that it is designed to work with up to three processors (I can quote chapter and verse later). > as well as retaining goodwill by offering an in-cabinet upgrade. Ha, ha, ha. That was a funny one, Terry. Goodwill! Ha, ha, ha. You can't be serious ... we are talking about a decision made by the DEC marketing strategists ... the same ones who scuttled a 8300 to 8350 upgrade, vetoed a MicroVax II to III upgrade, etc. etc., etc., ad nauseum. Goodwill. From DEC marketers. Oh, Terry, you're such a kidder. ================================================================================ Note 254.7 SMP and CIBCI (CI750) don't mix? 7 of 18 EISNER::FULLER_S "From the terminal of Stu Fuller" 6 lines 2-JAN-1989 21:59 -------------------------------------------------------------------------------- Kevin, Before getting too excited, how about giving us a look at the configuration in question? What BI devices and BI node numbers? Did you try putting the CIBCI at a low BI number, and the cpus at a higher BI node number? ================================================================================ Note 254.8 SMP and CIBCI (CI750) don't mix? 8 of 18 EISNER::KENNEDY "Terry Kennedy" 42 lines 3-JAN-1989 01:16 -< One more try... >- -------------------------------------------------------------------------------- > As I understand it, DEC DID have it working, and developed SMP using > it. I also know for a fact that a clustered 8360 or 8370 did in > fact work for them ... but, they had a BI CI and not a CIBCI. Yes, *some* configurations work. However, a marketable system implies that *all* options found on the pre-upgrade machine will work after an additional CPU is added. This is not the case, and that's what I meant when I said that they tried hard to make it work. Anyway, if you're sure about all this, just go get the other CI interface and be happy - you've already proved you don't care about 'supported' con- figurations. > So, I am puzzled as to why they could get it working and you are > saying that it is technically impossible. Do you speak from experience, > Terry, or is your explanation an educated speculation? I have direct experience with >2 KA820's in a box, but not with the CI interface gear. My application was with shared memory to an external device, as well as with a DR11 style interface. > Also, the 8200 Technical Manual says that it is designed to work > with up to three processors (I can quote chapter and verse later). The 11/70 Technical Manual claims to work with up to *four* processors, but you can't buy one of *those* either. You mean you believe what the technical manuals say? > Ha, ha, ha. That was a funny one, Terry. Goodwill! Ha, ha, ha. You > can't be serious ... we are talking about a decision made by the > DEC marketing strategists ... the same ones who scuttled a 8300 > to 8350 upgrade, vetoed a MicroVax II to III upgrade, etc. etc., > etc., ad nauseum. Goodwill. From DEC marketers. Oh, Terry, you're > such a kidder. Well, to rebut your points here: 1) I know of several users who were able to upgrade from 8300 to 8350 systems, for a reasonable price (in DEC terms). 2) uVII to III upgrade - true, not in the price book as such (for assorted technical reasons, which *happen* to be quite valid, if not universally applicable). You could buy the boards as spares and do it yourself, though. [In a similar manner to how you apparently achieved whatever system you're having the problem with.] ================================================================================ Note 254.9 SMP and CIBCI (CI750) don't mix? 9 of 18 EISNER::KENNEDY "Terry Kennedy" 15 lines 3-JAN-1989 01:23 -< "I'm sorry, this is abuse" -- Monty Python >- -------------------------------------------------------------------------------- You ask a technical question, and when others ask for clarification you respone with coy 'well, sort of' non-answers. I do my best to respond to your question, and then you come back and tell me I'm wrong. If you knew the answer in the first place, why ask the question? And then the whole thing degenerates into a SOAPBOX-style poke at DEC's 'lack of caring' (my impression of your comments) about users in general. That is a topic for some other conference, not relevant to the question which you originally asked. Anyway, to make along reply short, COUNT ME OUT OF THIS DISCUSSION!!! ================================================================================ Note 254.10 SMP and CIBCI (CI750) don't mix? 10 of 18 EISNER::ANGLEY "Kevin Angley" 28 lines 3-JAN-1989 13:56 -< Can we kiss and make up ... well, at least make up? >- -------------------------------------------------------------------------------- Terry .. as many others have observed, NOTES sometimes lacks in ability to communicate fully the intent and manner of the other party. I believe such was the case here. If my flame was on at anyone, it was certainly the predicament that DEC has me in now, and not at those like you who are trying to help. But I can see how the temperament in my reply may not have had a specific aim. The predicament that we are in now is something that I am constrained from discussing in too much detail. I also did not want to say anything about the specific model number because of a thousand reasons. Yet, once someone else brought it up, I felt I should confirm it. I was not being coy, just cautious under the circumstances that I cannot explain. I did not tell you that you were wrong. I was just soliciting the kind of useful information that you gave me in .8 (i.e. that it is configuration specific, not a total impossibility). You are right about the SOAPBOX-style comments. But I believe they were in specific rebuttal to your inference that DEC would try to make such an upgrade available if it were technically feasible. I disagree, but any discussion along these lines (from both sides) does not belong in HARDWARE_HELP. I plead guilty. If it makes any difference to you, I very much respect your technical expertise (you helped me quite a bit on DEPCA/VGA questions). If I have discouraged you from helping me with these or other issues, I truly regret it. ================================================================================ Note 254.11 SMP and CIBCI (CI750) don't mix? 11 of 18 EISNER::ANGLEY "Kevin Angley" 36 lines 3-JAN-1989 14:34 -< From the soul of my 8370 ... >- -------------------------------------------------------------------------------- > Before getting too excited, how about giving us a look at the > configuration in question? What BI devices and BI node numbers? > Did you try putting the CIBCI at a low BI number, and the cpus at > a higher BI node number? I've been too excited all my life. Here is the configuration: Node Revision Device Code Type 00 0008 0102 DWBUA 02 2833 0105 KA820 05 0000 010B CIBCI 06 2833 0105 KA820 07 2833 0105 KA820 08 0000 0001 MS820 09 0000 0001 MS820 0A 0000 0001 MS820 0B 0000 0001 MS820 0C 0000 0001 MS820 0D 0000 0001 MS820 I am open to suggestions. Just as an experiment I up'ed PASTIMOUT to 10, 15, then 99 seconds with no apparent effect on the problem. I may have been mislead by the documented dynamic attribute of this parameter. It may be "semi-dynamic" and then changes I made never took effect. Do the node addresses determine bus priority (within the same priority?). Are there some configuration rules for the order of the BI options that I must keep in mind or can I just switch plugs around 'til the cows come home? Oh .. in the config info above, this may not reflect that I did upgrade the microcode to patch level 27 on the KA820's. ================================================================================ Note 254.12 SMP and CIBCI (CI750) don't mix? 12 of 18 EISNER::KENNEDY "Terry Kennedy" 9 lines 3-JAN-1989 19:47 -< All right, I'm back >- -------------------------------------------------------------------------------- Ok - no problem (any more). Now, back to the question at hand - Let me go pull my notes & prints on the CIBCI and collect some revision info as well. By the way, I gather that your disks/tapes/terminals are on the DWBUA. Can you give me a list, in bus order, of what's in there? Also, does the problem happen in 'pure' CPU-bound cases or only with associated Unibus activity? ================================================================================ Note 254.13 SMP and CIBCI (CI750) don't mix? 13 of 18 EISNER::ANGLEY "Kevin Angley" 18 lines 3-JAN-1989 21:24 -< Whew! >- -------------------------------------------------------------------------------- All disk I/O is through an HSC50 on the CI. The problem manifests itself when three completely CPU-bound processes (i.e. an image of a PASCAL DO WHILE TRUE;) are running - practically no other "real" activity on the system (but, of course, there's no guarantee that some symbiont or server process didn't sneeze while all this was happening). On the UNIBUS, there is a DEUNA (I know that is first on the UNIBUS), a TS11 emulator and a DEC DMF32, and then a UNIBUS extender cable into twin Megatek graphics controllers. During the testing, there was little or no activity on any of these (i.e. no tape, graphics, or terminal users). There was a little Ethernet as I was SET HOSTed from another node possibly - but nothing terribly significant.) I doubt it is UNIBUS related (oops .. there I go answering my own question again), but I could try eliminating the UNIBUS and give it a go sometime. I am also waiting for a BI CI to test with, but locating a loaner is going to take me some time. ================================================================================ Note 254.14 SMP and CIBCI (CI750) don't mix? 14 of 18 EISNER::HARVEY "Jack Harvey" 15 lines 5-JAN-1989 19:45 -< Strip to Bare-Bones? >- -------------------------------------------------------------------------------- While Terry works on finding the *right* answer, let me suggest something that may be obvious to you but not to some other readers. Trying stripping your system down to the bare minimum set of modules. Take everything, such as the Ethernet and tape controller, not central to the problem out. Stop any VMS process not required for your tests. I know this is a non-trivial exercise requiring many hours with a dedicated system, but it might illuminate the problem. If you get it down to a bare-bones system that works with 2 CPUs but not >2 CPUs, it's pretty convincing it won't work for you. But if you can get it to work as you want bare-bones, then start replacing the things you removed. When it breaks, you may learn something useful about its configuration dependency. ================================================================================ Note 254.15 SMP and CIBCI (CI750) don't mix? 15 of 18 EISNER::ANGLEY "Kevin Angley" 18 lines 11-JAN-1989 14:56 -< Was the symptom, not the disease >- -------------------------------------------------------------------------------- I think I made a major breakthrough on this when I discovered that the "dynamic" attribute documented for PASTIMOUT was misleading or plain wrong. One of the first things I had tried after getting the maintenance timer expirations was to dynamically up PASTIMOUT, but that seemed to have no effect. Last night I changed my assumption that it was dynamic, and rebooted. The 8370 worked for four hours with a simulated 20 user CPU load. It continued to work normally throughout the night, and with heavy users on this morning. However, as I sat down to type this in, the virtual circuit closed down and hung. I forced a crash. It appears from the evidence thus far that pursuing the CI timeout was barking up the wrong tree. The CI timed out because something else is staying at an elevated interrupt level for an awfully long time. I hope to try it bare-bones this weekend. ================================================================================ Note 254.16 SMP and CIBCI (CI750) don't mix? 16 of 18 EISNER::HARVEY "Jack Harvey" 13 lines 12-JAN-1989 11:36 -< Bare-bones = Death's Head? >- -------------------------------------------------------------------------------- >> I hope to try it bare-bones this weekend. I hope I haven't lead you into trouble, and I wish I could be there to help. Shaking up computers like I proposed is generally a no-no. Only when faced with a problem like yours, would I even mention it. Again insulting your intelligence, take copious notes of the current configuration before dismantling. Photographs might not be a dumb idea. Hard copy of the configuration by SYSGEN. (I wish I knew what I have missed. Anybody?) Good Luck! ================================================================================ Note 254.17 SMP and CIBCI (CI750) don't mix? 17 of 18 EISNER::ANGLEY "Kevin Angley" 12 lines 12-JAN-1989 15:48 -< Now all I need is an 8370 label on the thing >- -------------------------------------------------------------------------------- Barebones may not be necessary ... taking the advice of a DEC insider, I rasied PASTIMOUT and PAPOLL to 16, and also increased non-paged pool somewh (it was the latter which caused the hang yesterday). Although raising these times to 16 seconds may seem like it is just hiding the problem, the insider doesn't think so ... reasoning that 3 CPU's will take I/O interrupts even BETTER than the 2, he didn't seem to think that there was indeed a problem. I have now been running for a full day with heavy usage with no problem. ================================================================================ Note 254.18 SMP and CIBCI (CI750) don't mix? 18 of 18 EISNER::HARVEY "Jack Harvey" 1 line 12-JAN-1989 19:32 -< Dare I say? Whew! >- --------------------------------------------------------------------------------  ================================================================================ Note 255.0 MicroVAX/LAVC hardware problem No replies EISNER::GEIER "Jim Geier, General Instrument" 3 lines 30-DEC-1988 19:54 -------------------------------------------------------------------------------- I have entered note in the MicroVAX_systems conference that looks more and more like a hardware problem. Please see note 74.* for the details and to respond. All suggestions, comments, etc. are welcome, of course.  ================================================================================ Note 256.0 Power Consumption Specs Needed 39 replies EISNER::CONROY "Alan Conroy" 37 lines 3-JAN-1989 14:49 -------------------------------------------------------------------------------- It seemed a toss-up as to where to put this node (HARDWARE_HELP or SITE_ MANAGEMENT), but here goes... I am trying to come up with an ACCURATE representation of the amount of heat and power produced/consumed by the equipment in our computer room. It appears that our air conditioning is close to the limit, but I'm guessing at the BTU output of several pieces of equipment. Anyway, I will list the pieces that I'm not sure of and would appreciate any comments from those who know a) what the power dissipation is (preferrably in BTUs, but I can convert from watts), and b) power draw (watts). Thanks in advance. P.S. please include your sources for the information - I'd like to get copies for myself. DEC TS11 controller board(s) SI 9900 box SI 9400 controller Able DH/DM controller DEC RH750 PDP-11/73 CPU (only!) DEC RD51 DEC RLV11 Emulex CS02 Emulex SC03 Emulex CS21 Fujitsu 2284 disk drive DELUA DEC TM03 controller DEC TUK50 controller DZQ11 TK70 TQK70 TQK50 DEC RM03 Emulex TC11 Cipher 900X tape drive NS70/75 memory board MK11-KE memory board Mostek MK8070 memory board DEC DCL box (RM controller) ================================================================================ Note 256.1 Power Consumption Specs Needed 1 of 39 EISNER::CONROY "Alan Conroy" 2 lines 3-JAN-1989 14:50 -< Oops - forgot one biggie >- -------------------------------------------------------------------------------- Also need specs for PDP-11/70 CPU and FP-11C - bare bones, with no other equipment. Thanks again. ================================================================================ Note 256.2 Power Consumption Specs Needed 2 of 39 EISNER::HAHN "DECUServe Vice Chair" 2 lines 3-JAN-1989 16:20 -< Try a clamp-on ammeter >- -------------------------------------------------------------------------------- Would it not be easier to get a clamp on ammeter ? ================================================================================ Note 256.3 Power Consumption Specs Needed 3 of 39 EISNER::LEDERMAN "Bart Z. Lederman" 9 lines 3-JAN-1989 19:47 -< A watt meter would be better. >- -------------------------------------------------------------------------------- You really need a Watt meter, not an ammeter, because the power factor on computer equipment usually isn't 1.0: otherwise, I agree that a measurement is almost certainly more accurate than the estimates in the book. (It would be really interesting to know if DEC gets it's figures by measurement of power, or by temperature rise in the room, or estimate, or just how they do it). You can pretty well figure that all of the power that goes into the system will end up as heat: the amount of energy that comes out as light or power down the RS-232 lines is negligible. ================================================================================ Note 256.4 Power Consumption Specs Needed 4 of 39 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 65 lines 3-JAN-1989 23:06 -< AMPS -> WATTs -> BTUs -> TONs >- -------------------------------------------------------------------------------- >> -< A watt meter would be better. >- >> >> You really need a Watt meter, not an ammeter, because the power factor >> on computer equipment usually isn't 1.0: otherwise, ... But to just get a quick look at how much heat is being dumped into the room, simply using AMPSxVOLTS will get a usable answer that will only err by being too HIGH for ignoring power factor. Your computer room A/Cs should have enough extra capacity, anyway, so using a slightly high number just gets you more safety margin. An adequate AMPROBE (tm) type meter capable or reading AMPs and, with auxiliary test leads, VOLTS, costs from maybe $35. to $100. depending on features such as: ranges, max amps, ohms, and analog needle vs lcd display. Fancier ones also can capture peak inrush AMPS for as a motor starts that are too fast to get on a plain analog meter. I suspect a good WATT meter costs more. >> ...I agree that >> a measurement is almost certainly more accurate than the estimates >> in the book. The cards are just a piece of the load, the damned power supplies pig a lot of power on their own. If you take a big old DEC power supply and look at what it draws loaded as opposed to unloaded, you get a big supprise. You use a lot of power at idle, including some burned in load resistors that are there simply to present a minimum load. (In recent 11/84 packaging, they have refined this by having minimum load resistors on external cards plugged into bus slots and which are removed as one adds useful cards.) This assumes 3-phase WYE power, but the same directions apply to 120/240 single phase. To compute BTUs, read amps in each HOT (Phase leg) wire seperately, and ignore your neutral wire. Multiply the SUM of the amps read by the hot wire to neutral voltage (should be 110 to 125 volt range) or, since we are already ignoring power factor, assuming 117 volts is close enough. That gets WATTs and WATTS x 3.414 gets you BTUs/hour. If you would rather have TONs, divide BTUs/hour by 12000. And don't forget the lights! A 40 watt flourescent light takes almost twice that when you throw in what the ballast wastes. The newer electronic ballasts are a lot better, but few have them unless they were intentionally installed, as they cost more initially. Then if you have a nominal Y ton air-conditioner, DON'T assume it will do Yx12000 BTUs. Find its real rating at your room temp and at the real outside temp (or cooling tower or chiller water temp - or where ever it has to pump/dump its heat to). Then realise that people A/Cs are often sized with smaller evaporators as they can assume a fair portion of their load is LATENT (moisture to be condensed out), but computers don't sweat or breath, so their load is totally SENSIBLE (hot air) which by itself needs a larger coil for the same total BTUs. In other words, if you have a 10 ton people cooler, don't count on 10 tons of computer cooling. ================================================================================ Note 256.5 Power Consumption Specs Needed 5 of 39 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 13 lines 3-JAN-1989 23:55 -< Another source of heat. >- -------------------------------------------------------------------------------- < The cards are just a piece of the load, the damned power < supplies pig a lot of power on their own. Point of order from a former electronics technician: getting a power supply that runs at 85% efficiency is usually considered fantastic; 80% is usually quite good, even for a switching regulator. And this is in stuff built for military use where cost is usually secondary to performance (you can only stuff a certain size battery into a torpedo, so you can't afford to waste too much in converting to DC or AC). Back to heating/cooling... don't forget people! Humans are warm blooded people ( with the possible exception of certain legal types :-} ).. seems like I recall about 1000 BTU / person. ================================================================================ Note 256.6 Power Consumption Specs Needed 6 of 39 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 13 lines 4-JAN-1989 01:19 -< Old power supplies have large idle load >- -------------------------------------------------------------------------------- >> < The cards are just a piece of the load, the damned power >> < supplies pig a lot of power on their own. >> >> Point of order from a former electronics technician: getting a power >> supply that runs at 85% efficiency is usually considered fantastic; 80% >> is usually quite good, even for a switching regulator. Sure! I don't think we are saying different things. But those %s are at full load, and all I was trying to point out is that an old BA11-F class box (he did say 11/70 - right?) must have vastly LOWER efficiency when only partly loaded than when fully loaded. Or saying it yet another way: If you yank 1/2 the cards out of your backplane you do NOT cut your power consumption in half. ================================================================================ Note 256.7 Power Consumption Specs Needed 7 of 39 EISNER::CONROY "Alan Conroy" 38 lines 4-JAN-1989 15:18 -< Thanks, but I need numbers >- -------------------------------------------------------------------------------- A person should be considered 400 BTU, not 1000. Each flouresent light bulb should be counted as 100 BTU. Also, for every cubic foot of room affected by the air conditioner add 1.6 BTU. A good safety margin for air conditioning equipment is to run at no more than 80% load (some would even say 75%). Using a meter would be great IF we had one. We don't however, and management won't buy one. Now, we could rent/borrow one, but as soon as we changed our configuration, we'd need to measure it again. Our power controller shows how much power (KVA, amps & volts per phase) is going into the room - but that doesn't help me figure out how much a particular system is using. This is important as I'm trying to build a case to keep some older equipment and there are power concerns, etc. We could shut down everything except one system, measure the power controller output, then shut it down and bring another up, etc - but I just can't justify that downtime, especially when things keep changing in major ways. And so long as our company is growing, things will continue to change drastically - I need a way to determine (ahead of time too) what the effects of these changes are. Incidentally, a watt meter won't help me find out what out air conditioning load is. When I compared the power controller output against the figures I came up with on paper, my figures were OVER by 50%. Most of my information is from the manuals - so they must be fairly accurate or else padded up well. Or else, my guesses on some equipment are WAY off. Also, I agree that power supplies use up power in and of themselves. Unfortunately, I have no idea what to even guess for them. If anyone has BTU/power ratings for DEC power controllers then I would gladly like those too! My air conditioning comparisons indicate that I only have another 2000 BTU left - and we're about to install a 11/780! Once again, hopefully some of my figures are high, or we're only going to be able to keep the room at about 75 degrees minimum. Since there is no safety margin, we will probably have to add air conditioning capacity - but how much? All these questions come down to needing a fairly accurate idea of what is going on in the computer room now, plus the ability to predict the effects of future growth. For probably 80% of our equipment I have what I feel to be reasonable statistics. However, the other 20% of the equipment probably consists of 60% of the power and air conditioning load. I've guessed on much of that. If I can get statistics to refine my figures, then I can more accurately determine the environmental needs of the systems and provide safe upgrade plans at a minimum of cost - after all, what is a System Manager for? Any concrete help is appreciated! Thanks. ================================================================================ Note 256.8 Power Consumption Specs Needed 8 of 39 EISNER::CERMAK "Denny Cermak - 3M Company - (612)73" 20 lines 4-JAN-1989 20:23 -< Make use of Field Service? >- -------------------------------------------------------------------------------- My apologies if this sounds simplistic, but for most of the DEC items you listed Digital Field Service has the power and heat load data. I remember seeing a "Site Preparation Guide" once upon a time that had much of this too. Most of the field service offices I've worked with have one or two people who do site prep consulting or at least do checkout of new sites. One name comes to mind...Ed McConnell in the Indianapolis/Anderson office. He'd probably be happy to help you or at least put you on to someone in your area. ...denny P.S. Just opinion but I doubt a meter (watt or amp) will do you much good. You need to size power based on peak consuption. Ever see an ammeter peg when you've spun up a disk? Sizing on max load builds in a nice little safety factor too. ================================================================================ Note 256.9 Power Consumption Specs Needed 9 of 39 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 46 lines 4-JAN-1989 20:43 -< % A/C runtime = % capacity used >- -------------------------------------------------------------------------------- >> Using a meter would be great IF we had one. We don't however, >> and management won't buy one. I bet we all have CHEAP management... My solution, especially for tools I may want elsewhere, is to get my own. MY AMPROBE sits in my desk. MY $500. 'butterfly' tool (for putting 25 pair phone connectors on cables) is down stairs. MY SawZall cuts holes in the computer room floor. My key machine just went home. And so on, but then again I have been here 23 years and with some of the same people many years more. >> Our power controller shows how much power (KVA, amps & volts >> per phase) is going into the room - but that doesn't help me >> figure out how much a particular system is using. ... We could >> shut down everything except one system, measure the power >> controller output, then shut it down and bring another up, etc >> - but I just can't justify that downtime, (I wish we had one of those). If it is really accurate, and if there is little dynamically changing (tapes, etc), you might be able to read everything's total and then shut off just what you want to check and compute difference to get its load. >> Incidentally, a watt meter won't help me find out what out air >> conditioning load is. Aside from people and infiltrating air and radiation and conduction from adjacent HOTTER areas I don't understand why a WATT meter wouldn't be useful. >> All these questions come down to needing a fairly accurate >> idea of what is going on in the computer room now ... Some big A/C units unload some cylinders or use hot gas bypass or other schemes to keep the compressor on but modulate the capacity, or if you are on a central chilled water plant you may have a fully modulating supply valve. But if you have a Vanilla one that simply turns the compressor on full blast and then shuts it off (or if you can temporarily con a smarter system into playing dumb) your % compressor running time is about the % of capacity you are using! If you have several units that aren't too far apart, setting some colder may keep them on forcing others to do all the cycling so you only have to track runtime in fewer places. ================================================================================ Note 256.10 Power Consumption Specs Needed 10 of 39 EISNER::LEDERMAN "Bart Z. Lederman" 27 lines 4-JAN-1989 21:33 -< I think you have a different type of problem >- -------------------------------------------------------------------------------- Given your situation, then probably an ammeter on the line going into the desired pieces of equipment may be the best choice. I've long suspected that the BTU figures in the manuals were estimated high. It makes sense, when you figure that people might use them to size their air conditioning equipment: it's generally better (from the computer manufacturers point of view) to get a little larger air conditioner than one which is too small. I disagree that you will have to worry about start-up surges. You are obviously keeping your equipment on most (all?) of the time, and want to know what the equipment uses in the long term. I also think the estimates for fluorescent lights is inconsistent with modern equipment. With ballasts that light two 40 watt rapid start lamps made in the past 20 years or less consumption is around 100-110 watts. It will be marked on the ballast in true watts, if you REALLY need to know accurately. But most computer rooms have a LOT of lights, and it adds up. If management is worrying about the cost of running older equipment (I run into that a lot too) maintenance usually costs more than power and air conditioning, even here in NY City where power is expensive. But be sure to look into the cost of air conditioning people, all of the office equipment (especially copying machines!), if you really want to show what the cost of operation is as a proportion of the whole. ================================================================================ Note 256.11 Power Consumption Specs Needed 11 of 39 EISNER::KENNEDY "Terry Kennedy" 12 lines 4-JAN-1989 22:16 -< RM05, 11/70 specs >- -------------------------------------------------------------------------------- RM05 (spinning, includes adaptor) 1.2 Kw, 4200 BTU/Hr, .7 power factor RM05 (not spinning, " " ) .4 Kw, 1400 BTU/Hr, .9 power factor (above from EK-0RM05-UG-002) PDP-11/70, FP11, 4 x RH70, 5 x SPC, 128K of MK-11: 3Kw, 10,000 BTU/Hr RH70 - 200 W, 700 BTU/HR FP11-B, C- 200 W, 500 BTU/Hr MK11 (loaded) 500W, 1700 BTU/Hr (above from EK-11070-MM-002) ================================================================================ Note 256.12 Power Consumption Specs Needed 12 of 39 EISNER::MERRIMAN "George Merriman - CCA/NY" 7 lines 4-JAN-1989 23:37 -< A/C must be big enough for worst-case day >- -------------------------------------------------------------------------------- > But if you have a Vanilla one that simply turns the compressor on full > blast and then shuts it off (or if you can temporarily con a smarter > system into playing dumb) your % compressor running time is about the % > of capacity you are using! This depends on the outside air conditions. To determine installed capacity needed, better do it on a hot, muggy day. ================================================================================ Note 256.13 Power Consumption Specs Needed 13 of 39 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 21 lines 5-JAN-1989 00:44 -< Not buying meter seems counter-intuitive >- -------------------------------------------------------------------------------- < Using a meter would be great IF we had one. We don't however, and < management won't buy one. Now, we could rent/borrow one, but as soon < as we changed our configuration, we'd need to measure it again. Our < power controller shows how [NOTE: No sarcasm is intended here - just a running commentary] Let's see... you're looking at buying another air conditioner - an easy $5,000 to $10,000 expense. You're getting an 11/780 - another $??,000. And management doesn't want to spend a couple of hundred to evaluate what they need to buy now, and later? Dunno.. this doesn't seem to make a lot of business sense to me Also.. DEC's listings in the site-prep manuals usually reflect full loads in the cabinet. This could easily be the source of your apparent over-estimations. Startup surges shouldn't be considered in A/C determinations - they are usually very short-term ( 5 to 30 seconds ), and so don't affect much in the long run. ================================================================================ Note 256.14 Power Consumption Specs Needed 14 of 39 EISNER::HASSINGER "Bob Hassinger" 25 lines 5-JAN-1989 11:18 -< Design more conservatively. Get the meter! >- -------------------------------------------------------------------------------- Thoughts: 1) You have already spent a fair fraction of the cost of the meter just calling in here looking for help. This is by far your best bet. Is there any chance you have an in house electrician or one who comes in to do work from time to time who could help you with his meter? 2) How good are the paper figures going to be when/if you get them? I think of the book BTUs about the way I do for "shipping weight" - only rather loosely related to the real numbers and high in practice. 3) The A/C units we use have had more variability in their performance over time due to outside factors than the margin you are trying to work with. That is, cutting it this close you are sure to have trouble when it gets really hot in the summer. We went down last summer because some grass from lawn mowing got into the condensers and cut down air flow a little. 4) If you think you are "getting close" to capacity at this time of year, you are in trouble now. You sure need to worry about adding another 780. From an engineers point of view, given all the uncertainties and variabilities I would have to be be more conservative in sizing the A/C than you are trying for. ================================================================================ Note 256.15 Power Consumption Specs Needed 15 of 39 EISNER::CONROY "Alan Conroy" 22 lines 5-JAN-1989 14:29 -< 11/70 questions >- -------------------------------------------------------------------------------- >RM05 (spinning, includes adaptor) 1.2 Kw, 4200 BTU/Hr, .7 power factor >RM05 (not spinning, " " ) .4 Kw, 1400 BTU/Hr, .9 power factor > >(above from EK-0RM05-UG-002) Thanks! Don't have this one around. I don't suppose DEC sells them any more. I'll try DEC Direct, but if anyone has an extra they would like to sell/trade/give, please let me know. >PDP-11/70, FP11, 4 x RH70, 5 x SPC, 128K of MK-11: 3Kw, 10,000 BTU/Hr ^^^^^^^^ Oh, THAT's what they meant by 4 "high-speed controllers". What about the SPCs? What are they assuming for those? >MK11 (loaded) 500W, 1700 BTU/Hr Meaning a full box? If so, then how much does the box hold? What I'm trying to do is strip out all of the extras that DEC included in the manual you are quoting from (for example, we have 1 RH70, 3MB of non-DEC memory, and non-DEC SPC loads). That leaves me with the CPU/FPP which I can add to my other calculations. ================================================================================ Note 256.16 Power Consumption Specs Needed 16 of 39 EISNER::CONROY "Alan Conroy" 17 lines 5-JAN-1989 14:33 -< Watt meter and air conditioning >- -------------------------------------------------------------------------------- >> Incidentally, a watt meter won't help me find out what out air >> conditioning load is. > >Aside from people and infiltrating air and radiation and >conduction from adjacent HOTTER areas I don't understand why a >WATT meter wouldn't be useful. OK, I'm confused now. How WOULD a watt meter help me determine air conditioning load? >But if you have a Vanilla one that simply turns the compressor >on full blast and then shuts it off (or if you can temporarily >con a smarter system into playing dumb) your % compressor >running time is about the % of capacity you are using! Good point! I think that is the case here. You guys certainly are helpful! :-) ================================================================================ Note 256.17 Power Consumption Specs Needed 17 of 39 EISNER::CONROY "Alan Conroy" 14 lines 5-JAN-1989 14:40 -< Whence meters? >- -------------------------------------------------------------------------------- > Let's see... you're looking at buying another air conditioner - an easy > $5,000 to $10,000 expense. You're getting an 11/780 - another $??,000. > And management doesn't want to spend a couple of hundred to evaluate > what they need to buy now, and later? > > Dunno.. this doesn't seem to make a lot of business sense to me Bingo! That's my management for you. I've worked at some bizzare places in my life, but the management here is pretty strange about such things. I must work within the limits given me. You (and others) seem to indicate that a meter is fairly cheap. I was imagining something >$1,000. If this is not true, please let me know where I can get some reliabale equipment at a resonable cost (and whence). If meters are truely cheap, then my power load woes are over! Thanks. ================================================================================ Note 256.18 Power Consumption Specs Needed 18 of 39 EISNER::LEDERMAN "Bart Z. Lederman" 17 lines 5-JAN-1989 19:03 -< SOme more info. >- -------------------------------------------------------------------------------- A watt meter will help you determine the air conditioning load for your electrical equipment, because it will tell you how much real power is being consumed. You can figure on all of this turning into heat. A watt meter will NOT tell you what additional air conditioning load you may have from non-electrical sources, such as heat infiltration through the walls from the outside or other rooms, people, etc. However, there are books and other sources which will help you estimate that. Clamp on ammeters (a little less accurate than a watt meter, but probably good enough) start, as mentioned, at around $45 outright. Watt meters are more, but can probably be rented. Try the yellow pages, or industry directories for any local components distributor. There are also mail order companies (will mail you addresses if you need them). ================================================================================ Note 256.19 Power Consumption Specs Needed 19 of 39 EISNER::HARVEY "Jack Harvey" 18 lines 5-JAN-1989 22:36 -< Check A/C Load for Office Space >- -------------------------------------------------------------------------------- > OK, I'm confused now. How WOULD a watt meter help me determine air >conditioning load? Bart (.18) is quite right. Watts are directly convertable to BTU. It's not magic, it's just basic physics. *Probably* the power into your computer equipment is the bulk of your total A/C load, but for the first pass, you must evaluate the other factors, such as lights, people, etc., as well. One short cut might be to compare your computer room with conventional office space in your building. See if you can find out how much A/C capacity your building engineer allocates per square foot of office space. Calculate the amount required for a space the size of your computer room and then *add* the computer power load. Somebody earlier gave the formula to convert watts to BTU. ================================================================================ Note 256.20 Power Consumption Specs Needed 20 of 39 EISNER::KENNEDY "Terry Kennedy" 14 lines 5-JAN-1989 22:38 -< Clarification of figures >- -------------------------------------------------------------------------------- |>(above from EK-0RM05-UG-002) > I don't suppose DEC sells them any more. Surprise! Available with 2 week lead time (1/2 normal), cost $45.00 > What about the SPCs? What are they assuming for those? Unknown. The figures in the manual for SPC's are blank except for items with external 'pods' (such as the A/D and GT40 terminal). |>MK11 (loaded) 500W, 1700 BTU/Hr > Meaning a full box? If so, then how much does the box hold? Yes, full box. 3.5 Mb (14 memory modules + controller) ================================================================================ Note 256.21 Power Consumption Specs Needed 21 of 39 EISNER::MERRIMAN "George Merriman - CCA/NY" 6 lines 5-JAN-1989 23:19 -< watt meter hardly worth the trouble >- -------------------------------------------------------------------------------- > Clamp on ammeters (a little less accurate than a watt meter, but > probably good enough) start, as mentioned, at around $45 outright. Watt > meters are more, but can probably be rented. Three-phase watt meters are more complicated to use than a clamp-on ammeter and I doubt the improved accuracy is worth the trouble. ================================================================================ Note 256.22 Power Consumption Specs Needed 22 of 39 EISNER::HARVEY "Jack Harvey" 47 lines 6-JAN-1989 14:13 -< Watts to BTU Conversion >- -------------------------------------------------------------------------------- We have been a bit sloppy with use of "BTU", which is a British Thermal Unit, or the amount of heat required to change the temperature of one pound of water one degree Farenheit. In air conditioning, the figure of interest is the *rate* at which heat is soaked up, or BTU per hour. Conventional A/C slang is to omit *per hour*. No problem, as long as it's understood. To convert watts to BTU per hour, simply multiply watts by 3.4134, usually rounded to 3.41. Someone earlier suggested DEC may build a safety factor into their quoted BTU/h. The following table gives watts and BTU/h for typical DEC hardware taken from the site preparation sections of "VAX Systems and Options". I have added a column of BTU/h divided by watts to show how DEC's figures line up with the formula above. Equipment Watts BTU/h Ratio LQP45-AA 58 198 3.41 SA482-AA 2,600 8,873 3.41 RA81-AA 600 2,048 3.41 TU80-AA 500 1,800 3.60 TU79-AB 1,960 6,688 3.41 HSS70-AA 875 2,975 3.41 8250/8350 1,400 4,760 3.40 8600/8650 6,500 22,000 3.38 8820 8,300 28,300 3.41 The departures from 3.41 are probably due to someone rounding the actual figures to make them neat. My guess is that in most cases DEC starts with the watts figure (which is itself usually a round approximate figure), multiplies by 3.41 and then rounds (again) to a nice publishable figure. But no safety factor, unless it's in the watts. Other examples from the site preparations sections (notably the LG01 and LG02 series printers) are *clearly* rough approximations. They all draw 1,000 watts and dissipate 3,000 BTU/h. Impossible. The only electronic devices that depart significantlyA from the 3.41 rule are radio and tv transmitters, where a few percent of the input watts goes up the tower to the antenna. All the rest becomes heat that the A/C must pump out. ================================================================================ Note 256.23 Power Consumption Specs Needed 23 of 39 EISNER::CONROY "Alan Conroy" 14 lines 6-JAN-1989 15:04 -< I am not an EE >- -------------------------------------------------------------------------------- Undoubtably my ignorance will surely show in the following question (but that's never stopped me before!) I would have thought that SOME of the energy going into the computer would perform some work and some of the rest would be converted to heat. What I hear is that 100% of the power is converted to heat at some point (ignoring energy conversion in LEDs, RF energy radiating off the wires, etc). Thus a component that drew 100 watts would put out exactly 341 BTUs. Is this what you are saying? That would certainly be useful, since if you have one (BTUs or power consumption), you can easily determine the other. In fact, for many of the pieces of equipment I have questions about, I already have one or the other of the figures. And yes, technically we should be saying BTU/Hour, but I believe it is common practice to assume /Hour unless otherwise specified. ================================================================================ Note 256.24 Power Consumption Specs Needed 24 of 39 EISNER::HARVEY "Jack Harvey" 23 lines 6-JAN-1989 19:33 -< BTU = Watts x 3.41 >- -------------------------------------------------------------------------------- > -< I am not an EE >- I am, and that means I live in fear and trembly of the other EEs reading this. They will slaughter me if I'm wrong. > Undoubtably my ignorance will surely show in the following question (but >that's never stopped me before!) Terrific, that's what makes DECUServe so great! (We are VERY good at answering dumb questions. *We* usually learn a great deal also.) > Thus a component that drew 100 watts would put out exactly 341 > BTUs. Is this what you are saying? Precisely. (For the purists, I will hedge and say I am being Newtonian; relativity is not considered. Nor do I consider that your modems might dump a few more milliwatts into the telephone lines than they dump into your computer room.) For the practical, I say: > Precisely. ================================================================================ Note 256.25 Power Consumption Specs Needed 25 of 39 EISNER::KENNEDY "Terry Kennedy" 0 lines 6-JAN-1989 19:58 -< Motors convert electricity into motion as well as heat, no? >- -------------------------------------------------------------------------------- ================================================================================ Note 256.26 Power Consumption Specs Needed 26 of 39 EISNER::HARVEY "Jack Harvey" 34 lines 6-JAN-1989 21:27 -< Where the Watts Go >- -------------------------------------------------------------------------------- > -< Motors convert electricity into motion as well as heat, no? >- Good point, Terry. I'm mumbling again. Take computer cooling fans, for example. They blow lots of air around the equipment and room. While the air is moving, the watts used are in the kinetic energy of the air mass, and *not* in heat. But as the air movement out at the walls and other surfaces slows down due to friction, the kinetic energy converts back into heat... Take disk drive motors, for example. While spinning up the disk, watts are being stored in the kinetic energy (think flywheel) of the spinning disk. When at normal speed, the watts from the motor exactly equal the watts lost in the bearings and air drag. When you spin down the disk, you stop watts into the motor. But the kinetic energy you stored in the disk (think flywheel) comes back to keep it spinning for a while. This eventually dissipates in bearing and air drag friction as heat... Take LEDs, for example. Usually, the photons emitted dissipate as heat on the walls or other surfaces they strike. Of course, if there is a glass window, they may escape and the A/C doesn't have to pump their heat out. On the other hand, the sun might shine upon thee through that window, so the A/C is much worse off. Radio frequency energy generated by all the electronic uproar in a computer can indeed be radiated through the walls of your computer room and *not* load the A/C. But the figure is below the limit I claim is of practical interest. If it were not, a computer would jam every TV and radio receiver within miles. ================================================================================ Note 256.27 Power Consumption Specs Needed 27 of 39 EISNER::KILGALLEN "Larry Kilgallen" 1 line 7-JAN-1989 17:05 -< It's those LED photons... >- -------------------------------------------------------------------------------- that's where my calculations were off! ================================================================================ Note 256.28 Power Consumption Specs Needed 28 of 39 EISNER::HASSINGER "Bob Hassinger" 15 lines 7-JAN-1989 18:09 -< The simple ideas are the best - Entropy... >- -------------------------------------------------------------------------------- I try to just remember the simple rules and derive everything else: Entropy - In my simple terms: for practical purposes all energy eventually ends up in its lowest form, heat. That has been the reason for the emphasis on measuring the electrical energy into the room. That, plus the heat the humans bring in, are pretty much the total input and it all has to go someplace. It spins or bounces or reflects or whatever until it has all been reduced to heat. Again, for practical purposes we never create or destroy energy. We can only change it's form. A very simple idea really. That, Ohms law and a couple others got me through EE. ================================================================================ Note 256.29 Power Consumption Specs Needed 29 of 39 EISNER::BYRNE_C "Charlie Byrne" 21 lines 8-JAN-1989 01:26 -< An aside into Physics and E= MC squared >- -------------------------------------------------------------------------------- > < Note 256.28 by EISNER::HASSINGER "Bob Hassinger" > > -< The simple ideas are the best - Entropy... >- > Entropy - In my simple terms: for practical purposes all energy > eventually ends up in its lowest form, heat. I'm straying off the topic a bit, and also showing how much I have forgotten (or never learned) about physics, but if: Energy ends up as heat, and E = Mc² doesn't that mean that there is a transformation for Heat = Mc²? Is it that the Energy that is converted into Heat makes those atoms (Mass) move around faster (speed)? As an aside, let me paraphrase Stephen Hawkings from "A Brief History of Time" (Bantam 1988, required reading ;-)): "Someone told me that each equation I included in the book would halve the sales. I therefore resolved not to have any equations at all. I the end, however, I *did* put in one equation, Einstein's famous equation E= Mc². I hope this will not scare off half of my potential readers." ================================================================================ Note 256.30 Power Consumption Specs Needed 30 of 39 EISNER::HARVEY "Jack Harvey" 42 lines 8-JAN-1989 14:08 -< Safety Considerations >- -------------------------------------------------------------------------------- >> "Someone told me that each equation I included in the book would >> halve the sales. There's gotta be a sophomore at every beer bust. Charlie, you just halved our audience. I will *not* drag in thermo-damn-namics. This thread is *practical*. :-) Speaking of which, I suspect we may be guilty of urging English majors and Business School graduates to open up their power distribution panels and start poking things in there. That not only isn't practical, it's hazardous to health. Not many people get electrocuted with the supply voltages found in most computer rooms. They just lie down quietly for an hour or so until the buzzing stops. Of course, there are some who never get up. Electrocution isn't the only hazard. Fooling around in there with a well insulated screwdriver will protect you from electrocution, but if you cause a short circuit with the conducting shaft or tip of the screwdriver, you will discover a new meaning of Electric Power. Before that 3,000 amp circuit breaker back upstream trips, the screwdriver shaft and parts of the power panel can become molten droplets sprayed into your face by the force of the explosion. And then there is the embarrassment caused by not only killing your computer room, but perhaps cutting power to the whole building. Electricians seem to relish such an event and take their time fixing the damage so they can savor it more. Here's some rules of thumb I propose. First, you must be mechanically adept enough to get the cover off the panel without tripping any circuit breakers. Then you must be able to identify and know the function of all the silvery or reddish objects and the black and colored hoses. You must understand the significance in this world of the great big hoses with their ends clamped by silvery or reddish objects. If you can't, better get a techie to help. ================================================================================ Note 256.31 Power Consumption Specs Needed 31 of 39 EISNER::HARVEY "Jack Harvey" 184 lines 8-JAN-1989 18:14 -< Safe Power Measurements >- -------------------------------------------------------------------------------- Note .30 never once mentioned A/C or BTU and hence could be considered quite off the track. But because of the impression of "it's easy, go ahead and do it" conveyed by early notes in this topic, I felt the caution was justified. I will now try to get back on track. Measuring the total power into the computer room is neat because it's the bottom line on your *current* A/C load. But it doesn't help much to plan for future growth. It can very often be useful to know what individual pieces of equipment are doing to you so you can estimate the effect if you add more of them, or perhaps replace old technology with new. I am assuming you have acquired a clamp-on AC ammeter as suggested by several previous notes. (If you decide on the wattmeter route, what I am going to suggest is still useful, but you'll have to do a bit more to connect the voltage probes. More later.) The following is my opinion of a reasonable safe procedure for the average biology major to use your dandy new toy. Next, go to a hardware store and get one of those heavy duty power extension cords, usually yellow. They have a conventional three prong male (two flat blades and one round pin) on one end. You can plug it in the same place your terminal or PC connects to power. The other end should be a single female connector that matches. The wire of this extension cord should be approximately flat, not round. You should see two small grooves on each flat face. The cord contains three conductors and the grooves mark the space between them. Get the shortest cord they have, no more than eight feet. *Don't plug it in for the following:* Using a sharp knife or razor blade, very carefully make a longitudinal cut a couple of inches long right in each of the small grooves. The cut should be at about the middle of your length of extension cord. Try to do it so the blade comes out in the small groove on the other flat face. What you are trying to do is separate the three conductors *without* exposing them. They should remain covered with the plastic insulation. If you do it right, the cord will be split into three individual yellow plastic insulated wires. (If you accidentally break through and expose the copper, just cover it up with a bit of black tape.) Having gotten the technique right, now extend the cuts until they are about six inches long. You should be able to spread the three insulated wires well apart so you can easily clamp the ammeter jaws around each wire individually. (If you can't find the flat variety extension cord described above, you can use the round type. This has three individually insulated wires and some jute or cotton cord to fill the gaps. It's covered by an outside sheath of plastic. Very carefully cut through this outer sheath and remove it. It's only about a millimeter thick and you don't want to cut into the insulation of the wires inside. Remove about six or eight inches of the sheath.) Now unplug your terminal or PC from its power outlet and plug the extension cord in its place. (Bye!) Plug your terminal or PC into the female at the other end of the extension cord. (Hello again!) Every thing is safe and cool. Your slashed up extension cord is mechanically weaker and will drive the fire inspector up the wall, but you can't get shocked or cause short circuits if you were careful to cover up with black tape any copper you might have exposed. Just don't use it for a permanent extension cord. You're now ready to learn how much power your PC draws. Clap the ammeter jaws around each of the three wires individually and note the current in each. One of them, the ground wire, has zero current. The other two show current and are essentially equal. They are the actual power conductors. Put *both* of the power conductors through the ammeter jaws. Surprise! The current reading doesn't double, it drops to zero. This is because, at any instant, the current in one conductor is going the opposite direction from the other and the magnetic fields (which the jaws sense) cancel each other. That's why we had to go to the agony of splitting the individual conductors up. So, measure the current in one of the power conductors. It doesn't matter which one since they should be equal. (If they're not very close, you have some leakage and that's another kettle of worms.) The simple answer to the power utilization is just to multiply the amperes by 120. For example, if you find your PC (or whatever) is drawing 1.25 amperes (amps), you multiply 1.25 by 120 to get 150 watts or 500 BTU. Now, before the purity patrol stomps on me, let me admit to two errors in the above. Your line voltage may not be exactly 120 volts at the time you measure the current. And even if it is, a law of nature says that volts times amps ain't necessarily exactly watts with alternating current. EEs like to call this "power factor" because it sounds real heavy and confuses the troops. Power factor is a number which always lies between zero and one. It's the number you multiply the volts times amps answer by to get the right answer. If you use a wattmeter, it will take care of these two errors automatically. But note that since power factor is never greater than 1.0, the simple volts times amps formula gives the *maximum* power the current could represent (assuming 120 volts). Extending the example above, let's assume your PC has a power factor of 0.8 which is kinda realistic. We multiply the 150 watts by 0.8 to get 120 watts or about 400 BTU. Most computer rooms these days have pretty good power systems feeding them. If your line voltage departs much from 120 volts, you probably know it from other bad things that usually accompany this. But it might be a good idea to look for some kind of clamp-on ammeter that includes a voltmeter function so you can eliminate this uncertainty. Now, having discovered the power used by our new TRASHmate, let's do something useful in the computer room. Down in the bottom of the DEC cabinets is a box with a black hose about 3/4 inch in diameter that disappears under the floor. The box has a circuit breaker. (Watch it! You'll bring down the whole cabinet if you brush a knuckle against it.) You will also find a bunch of dandy outlets that you can plug a PC (or our tricky extension cord) into. There will be several cords plugged in already. If you follow these up through the cabinet, you will find they go to interesting boxes, like disks, tape drives, etc. Now, find that third party disk drive whose manufacturer can't seem to find where they filed its power usage. Spin down the disk, follow its power cord to the controller box in the bottom of the cabinet (or wherever) and unplug it. (Don't bother notifying the users, you're the S.M. after all.) Put your trick extension cord in place, reconnect the disk through it and spin up. Do your number with the clamp-on ammeter and, presto!, you know more about the power requirements than the manufacturer, and a lot more than you did yesterday. ------------------------- Now if you want to use the voltmeter or the wattmeter to reduce the errors, you need to expose the copper of the two power conductors. I'm not going to give a cookbook procedure for this because it depends on your test leads, etc. Let me just suggest that a reasonably safe way is to strip about an inch of insulation from each of the two power conductors in the trick extension cord. The stripped sections should not be close to each other. Instead, put them at opposite ends of your section where the conductors are separated. (Safety in isolation.) Attach your voltmeter probes (or the voltage probes if using a wattmeter) to the bare copper wire in whatever way seems mechanically sound. If they are clips, fine, just clip them on. If they are simply short little rods, then wind the bare copper wire around around them tightly. Finally, cover all of the exposed metal of the clips, rods, probes (or whatever) as well as the power conductors with tightly wrapped black electrician's tape. Use at least two layers and be sure it overlaps the original conductor insulation by at least a half inch. --------------- You can extend this trick extension cord idea to measuring the power drawn by whole cabinets. Just get your electrician (or whoever understands about those giant power connectors on the ends of the cabinet power cable) to make up a male/female short extension cord with the appropriate connectors. Demonstrate your trick extension cord that we built above, and the idea will get across. If you have a cabinet that uses three phase power, it's a tad more (well about three times as) complicated. ================================================================================ Note 256.32 Power Consumption Specs Needed 32 of 39 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 6 lines 8-JAN-1989 18:32 -< Now if we could just select who gets the job... >- -------------------------------------------------------------------------------- >> Not many people get electrocuted with the supply voltages >> found in most computer rooms. They just lie down quietly >> for an hour or so until the buzzing stops. Of course, >> there are some who never get up. Think of it as Evolution in Action! ================================================================================ Note 256.33 Power Consumption Specs Needed 33 of 39 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 71 lines 9-JAN-1989 00:12 -< Ammeter Purchasing suggentions >- -------------------------------------------------------------------------------- >> I am assuming you have acquired a clamp-on AC ammeter as >> suggested by several previous notes. But if you havn't yet, this may give you more ideas about which models to buy, and, if you are stuck for a source, where to try. >> Put *both* of the power conductors through the ammeter >> jaws. Surprise! The current reading doesn't double, it >> drops to zero. But if you twist just one of the conductors so it goes through the jaws the opposite direction, the reading will double. But these tricks with more than one conductor work this simply only on this single phase example. But the doubling example leads to an even better idea. Take just one of the conductors (you may need to slit a bit more for this) and coil several turns into a donut sized coil whose hole is large enough for your meter's jaws. If you bought a one range meter, this can amplify the current on a small device to get it to where you can get a valid reading. Then just divide by the the number of wires the jaws encircle to get what the device actually draws. Meters have jaw size limits, too. The Monster jaws are great for big cables but are a pain the rest of the time. The meter I use (an AMPROBE RS3) can do a conductor up to 1" in diameter, and has the following ranges: 0-6/15/40/100/300 amps, 0-150/300/600 volts, and 0-1000 ohms. Its bigger brother the RS1007 does 2" , 0-15/40/100/300/1000 amps, 0-150/300/750 volts and same ohms. It only costs a few $ more ($95 vs. $72 for the RS3), but the big jaws make it a little harder to stuff into your pocket. Even on a multi-range meter, the coiled wire multiplier is useful, and AMPROBE makes a neat little thing that replaces the split cord above. It is maybe 6"x1.5"x.65" and has male and female ends. In the middle are 3 openings for the meter jaws labeled X1, X5, X10. If you have a really small load, you set the meter to 6 amps full scale, and chomp onto the X10 hole and then divide the reading by 10. This little adapter also has 2 little holes to poke your volt meter leads into for a safe reading. Well worth getting, but I can't seem to find a model # or price - sorry. For the big Twist-Lock connectored cords, the slit-it yourself extention cord is what you will need, but if you find an X5 reading handy, you may want to make the 'donut' windings more permanent by taping them up. Clamp on ammeters are readily available at any large electrical supply house, but there are national supply houses that have branches all over that make it even easier. Here are 3: I am looking at a W.W.Grainger catalog. It is NOT a computer or data supply house, but has ALMOST anything else. They list such meters by: AMPROBE, SPERRY, FLUKE (has a current clamp accesory that feeds into a regular multi-meter), GE, and TIF. There is also a clamp-on true WATTMETER (by TIF) for about $375. GrayBar has their phone/data/lan division, but they are also a regular electrical supply chain, so they will have them. Anixter (DEC's only DECCONNECT distributor) is also a good source. The catalog I am looking at doesn't show them, but I know AMPROBE sells nice 'KITS' that provide a carrying case for the meter and includes some accessories such as the adapter mentioned above that splits the cord and also gives a X5 and X10 signal multiplier. Such a 'kit' gets you the pieces, and provides a place to keep them all together safely. Needless to say, I have NO affiliation with any of the above companies. ================================================================================ Note 256.34 Power Consumption Specs Needed 34 of 39 EISNER::LEDERMAN "Bart Z. Lederman" 8 lines 9-JAN-1989 07:38 -< You may want to buy the split-out extension cord. >- -------------------------------------------------------------------------------- I'm fairly certain that the companies which sell clamp-on ammeters also sell power extension coards which have the hot lead split out for you to clamp on without having to buy or make your own. Just in case you don't feel safe digging into a power coard with a knife. For twist lock connections I think you can also buy such coards, or worst case you can have an electrical supply house make one up. You should not have to pay to have an electrician come on-site just to make up such a cord. ================================================================================ Note 256.35 Power Consumption Specs Needed 35 of 39 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 57 lines 9-JAN-1989 12:57 -< Twist-Locks + Fat Wire >- -------------------------------------------------------------------------------- >> You should not have to pay to have an electrician come on-site just >> to make up such a cord. (Especially in Local 3's territory where Bart is, the "should not have to" translates as: "could not afford to"). Here is a brief set of notes on cables and things to go on their ends for those inclined to roll their own. The heavy twist lock caps (male plugs) and receptacles (female outlets) that will most often be encountered will be 20 or 30 amp models, and will be 3 , 4, or 5 wires. Years ago these were horrible to wire, but are now quite easy. The code now requires a 'dead front' design where there is no longer a little piece of cardboard covering undersized screws with over sized wires wedged underneath with strands poking out every which way itching to zap or short anything near. The new designs generally offer a generous hole one simply pokes the stripped conductor into, and then a substantial screw will close an internal clamp to capture the wire. A strip length guage is generally moulded into the plastic to make the job even easier. The 2 'special' wires are the green one for the frame ground, and the white for the neutral. The screws will be color coded or labeled, and the other wires can be in any order but connect the same color to the corresponding terminal at each end. Because the rotation of the respective terminals will be opposite for the male and female devices, and because the cables colored wire rotation will reverse at opposite ends, it may be desirable to select the cord end to use to get the colors in the sequence needed for a particular gender end to eliminate crisscrossing wires in it. The other end will then be ok, too. Round cable is made in 2 general grades. The lighter duty 'junior' SJ cable rated @300 volts, and the heavy duty 'service' type S rated @600 volts. The letter T may be there for Thermoplastic (plastic vs rubber), and O may be there for Oilproof (in rubber, that would probably be Neoprene). Typical types would be: S, SO, ST, STO, SJT, SJTO, etc. There are many more, but this is typical. The wire guage for 20 amp cables should be #12, and for 30 amps, #10. If you need a 4 wire 20 amp cable, you could ask for 12/4 SO. Twist lock connectors have changed over the years, and there are now standardised blade configurations that define a particular voltage as well as AMP rating. It used to be that 3 prongs 30 amps could be any voltage and might or might NOT include ground and/or neutral. Unless locally modified, all your DEC stuff will have a NEMA standard number for the blade configuration that will be listed in the documentation. If DEC's specified NEMA numbers doesn't fit your installed connections, consider getting help. ================================================================================ Note 256.36 Power Consumption Specs Needed 36 of 39 EISNER::HARVEY "Jack Harvey" 20 lines 2-FEB-1989 04:59 -< Watts to BTU Redux >- -------------------------------------------------------------------------------- The February 1989 issue of Scientific American has an amusing ad on page 87 headlined "THE SIX INCH, FIVE POUND FURNACE". It describes "The Heat Machine PLUS" and says, "This ceramic heater not only uses minimal amounts of electricity, it is unique in its [...] tremendous heating ability." [Sounds like a unique solution to the energy crisis.] Several hundred words later, it gets to the point: "EFFICIENCY: At full capacity of 1500 watts, The Heat Machine Plus puts out 5120 BTU's per hour." [Wow!] ------------------------------------------------------------------------- Back in .22 of this thread, I gave a formula: > To convert watts to BTU per hour, simply multiply watts by > 3.4134, usually rounded to 3.41. 1500 watts x 3.4134 = 5120.1 BTU per hour. Now I'm trying to figure out where it hides that missing 0.1 BTU per hour. ================================================================================ Note 256.37 Power Consumption Specs Needed 37 of 39 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 19 lines 2-FEB-1989 06:14 -< I've heard this one before... >- -------------------------------------------------------------------------------- > > To convert watts to BTU per hour, simply multiply watts by > > 3.4134, usually rounded to 3.41. > > 1500 watts x 3.4134 = 5120.1 BTU per hour. > > Now I'm trying to figure out where it hides that missing > 0.1 BTU per hour. I saw an ad like that. As I recall, the device was a sort of a rounded-off box. Obviously, the missing 0.1 BTU/hour is due to roundoff error! (This reminds me of one of the questions on a joke physics exam: 23. Calculate the efficiency of an electric heater. No numbers, no nothing. But obviously, sinceall inefficiencies show up as heat, if what you WANT is heat, you get 100%! Even the light from the coils turns into heat when it's absorbed by whatever it hits...) ================================================================================ Note 256.38 Power Consumption Specs Needed 38 of 39 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 5 lines 2-FEB-1989 18:46 -< Except for the windows... >- -------------------------------------------------------------------------------- >> ... Even the light from the >> coils turns into heat when it's absorbed by whatever it hits...) Yes, except for the missing 0.1 BTU/hr represented by the light from those coils escaping through the room's windows... ================================================================================ Note 256.39 Power Consumption Specs Needed 39 of 39 EISNER::BYRNE_C "Charlie Byrne" 3 lines 2-FEB-1989 20:07 -< Back to the topic >- -------------------------------------------------------------------------------- Back to the subject please... your local friendly moderator.  ================================================================================ Note 257.0 DMZ-32 DMA problems and FCO information 8 replies EISNER::MERRIMAN "George Merriman - CCA/NY" 4 lines 3-JAN-1989 16:24 -------------------------------------------------------------------------------- I have copied the following notes concerning DMZ32 DMA problems and relevent FCO information from the VMS conference. George Merriman -- HARDWARE_HELP co-moderator ================================================================================ Note 257.1 DMZ-32 DMA problems and FCO information 1 of 8 EISNER::MERRIMAN "George Merriman - CCA/NY" 22 lines 3-JAN-1989 16:25 -< software workaround >- -------------------------------------------------------------------------------- <<< EISNER::DUA0:[NOTES$LIBRARY]VMS.NOTE;1 >>> -< VMS and bundled utilities >- ================================================================================ Note 370.0 dmz-32 and TTY_DMASIZE 11 replies EISNER::ESBENSEN 15 lines 27-DEC-1988 21:41 -------------------------------------------------------------------------------- We just talked to DEC and discovered something they DEC said is "well known". We have been using VMS for 6 years and seemed to have missed it...so I thought I might share it here. If you have a VAX that has a DMZ-32, you MUST disable all DMA transfers...or you will get disk errors whenever you have high DMZ and high DISK activity at the same time. The DMA transfering is disabled by: setting TTY_DMASIZE = 20000 (or some other large number) So, if you have been having "occasional" disk errors throughout the years...and have a DMZ...this might help!! ================================================================================ Note 257.2 DMZ-32 DMA problems and FCO information 2 of 8 EISNER::MERRIMAN "George Merriman - CCA/NY" 22 lines 3-JAN-1989 16:28 -------------------------------------------------------------------------------- ================================================================================ Note 370.4 dmz-32 and TTY_DMASIZE 4 of 11 EISNER::TERRAZAS "Mike Terrazas" 16 lines 28-DEC-1988 09:41 -< Don't do it!! >- -------------------------------------------------------------------------------- > If you have a VAX that has a DMZ-32, you MUST disable > all DMA transfers...or you will get disk errors whenever > you have high DMZ and high DISK activity at the same time. DEC told us to do this here, too and one of us dutifully did. The other (me) said, "No way, this is broken, fix it." Well, to make a short story long, they came in and replaced components and said "There!" Then they came back and replace components (I got a free upgrade to a DELUA out of it) and they came back and replaced backplanes, and they came back and replaced everything at once (except the rack) and it finally works (sort of, we still have some trouble intermittently with noise characters through the DMZs.) Whatever you do, don't lobotomize your DMZ, DEC is required to make it work (That's what our lawyers and DEC agreed to.) ================================================================================ Note 257.3 DMZ-32 DMA problems and FCO information 3 of 8 EISNER::MERRIMAN "George Merriman - CCA/NY" 104 lines 3-JAN-1989 16:34 -< FCO information >- -------------------------------------------------------------------------------- ================================================================================ Note 370.5 dmz-32 and TTY_DMASIZE 5 of 11 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 16 lines 28-DEC-1988 17:58 -< additional info >- -------------------------------------------------------------------------------- There was a FCO out about a year ago (maybe more) that corrected this situation. The problem surfaced as disk corruption and the apparent noise. I remember reading on this system or another that this was escalated by calling it a security problem. The noise that you are seeing is one of the other ports buffers, thus the security problem. I seem to remember that the problem was even more prevalent on systems that the unibus was connected to a BI. When correcting the microcode they found a possibility that the DMZ32 could corrupt the information on the BI and cause crashes although no one had reported any occurrences. I'm home but in the morning I'll look for this FCO and if found I'll post the additional info. ================================================================================ Note 370.11 dmz-32 and TTY_DMASIZE 11 of 11 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 75 lines 3-JAN-1989 13:12 -< FCO for DMZ32 as you requested! >- -------------------------------------------------------------------------------- > Well, we asked Colorado Springs about if there was any FCO where > the DMZ-32 problem was corrected. They specifically said "No!" > If anyone knows what rev level has this problem fixed, we > would like to know...so we can pass the information on to DEC! Found FCO paper work here goes:  FCO Release Date: 18 May 1987 Parts Availability: June 1987 FCO Number: DMZ32-I001 Level of Urgency "I" Applicability: Replace Revision E1 or earlier M8398 module with a revision F1 or later to incorporate new UBI microcode version 16 and TIU microcode version 1C when problem/symptoms are evident. Phase Revision "F1" module into spares by 14 March 1987 as per ECO M8398-MK003. Purge Stockroom 17 and all Field Service Stockroom by 1 April 1987. This FCO implements ECO M8398-MK003. Problem/Symptoms: Problem: 1. Flush bit not clearing DMA correctly. Symptom: File corruption on unibus disk controller drives and some users getting other users data. System errorlogger logs "IMR" invalid map register errors. This bug requires that "DMA" mode be turned off on all DMZ32 lines on the system. This is a VAX/VMS specific problem. Problem: 2. SSYNC timeout counter starts before NPG is received [NXM error]. Symptom: This causes Ultrix and UNIX systems to hang and or crash based on operating system environment at the time of the timeout. Problem: 3. When H3014 is powered off/on, DTR is not resorted by microcode. Symptom: Terminal lines set up as modem will not be able to connect to system, as well as users logged in through modem connections will be dropped due to no "DTR on" status. System must be rebooted to reenable modem lines. Problem: 4. Unsupported baud rates cause microcode to hang. Symptom: If un-suspecting user tries to set terminal to one of the two unsupported baud rates, DMZ32 microcode will hang and the device will be unusable to the system. A total system reboot is required to clear the condition. Problem: 5. NXM error sets the wrong bit in Octet CSR. Symptom: No error reports from field, but if NXM were to occur, the YC-driver would not be notified of it's occurrence. This could have serious implications on 8XXX system with DWBUA [unibus adapter] Problem: 6. Self test code constants enabled at wrong time. Symptom: None reported from field, but was discovered during internal testing of DMZ32 microcode. Problem: 7. Foreign terminals sending abutted characters lose trailing bits. Symptom: Files get invalid escape sequence data while editing using "arrow" cursor positioning keys. Order yours by FCO Kit # EQ-01457-01 Part #: M8398 Desc: Module at Revision F1 FA-04743-01 Desc: FCO Document F.S. Microfiche Libraries: EP-FSNVX-LB VAX Hope this helps ! If you would like your own copy of this drop me a mail message with your FAX # or address and I'll have it sent right out. ================================================================================ Note 257.4 DMZ-32 DMA problems and FCO information 4 of 8 EISNER::MERRIMAN "George Merriman - CCA/NY" 13 lines 3-JAN-1989 16:35 -< software workaround -- more info >- -------------------------------------------------------------------------------- ================================================================================ Note 370.6 dmz-32 and TTY_DMASIZE 6 of 11 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 7 lines 28-DEC-1988 18:10 -< Dynamically make your DMZ32 = DZ11 >- -------------------------------------------------------------------------------- > setting TTY_DMASIZE = 20000 (or some other large number) The dynamic command to clear this problem before the FCO is: SET TERMINAL/PERMANENT/NODMA TXnn: ================================================================================ Note 257.5 DMZ-32 DMA problems and FCO information 5 of 8 EISNER::PROVOST 90 lines 3-MAY-1989 15:02 -< Unibus loading rules >- -------------------------------------------------------------------------------- Below are some Unibus configuration rules taken from the PDP11bus Handbook, which is no longer available. Basically, if you have too many devices the system won't work. If you're close, there are tricks you can play by adding repeaters, adding excess cable, changing cable sizes, etc. Symptoms of loading problems are unibus hangs, lost interrupts, etc. Unibus Segment - A Unibus segment is a portion of a Unibus sys- tem between and including two terminators. A bus segment con- sists of a terminator, a 120-ohm transmission path (cable) with options having drivers and receivers, and another terminator (in that order). A single bus system has one bus segment. A multiple bus system contains more than one bus segment, usually separ- ated by bus repeaters or bus switches. For example, a DT03 bus switch contains bus repeaters. Lumped Load - A lumped load is a group of Unibus elements (other than cables or jumpers) that are interconnected via the Unibus (backplane wire or PC etch). The group is not a lumped load if it uses a cable to interconnect the elements, or if the ele- ments are separated by a bus repeater. Semi-Lumped Load - A semi-lumped load is a group of lumped loads interconnected by 91.44 cm (3 ft) or less of cable and not separated by a bus repeater. This cable may be one of the fol- lowing: Cable Length BC11-A-OF 15.24cm (0.5 ft) BC11-02 60.96cm (2 ft) BC11A-03 91.44cm (3 ft) M9202 60.96cm (2 ft) AC Bus Load - AC loading is the amount of capacitance a module presents to a bus signal line. AC loading is expressed in terms of ac loads. A unibus segment can support up to 20 ac loads. For more than 20 ac loads, an M9202 (jumper with cable) is used to separate a large load into two Unibus segments. A number is assigned to each module indicating its ac load. The number is derived using reflection techniques in the time domain. The ac load causes a reflection to occur on the bus line shortly after a signal is asserted. This reflection shows up on an oscillo- scope as a spike shortly after the asserting edge of the signal. The ac load rating is usually based on worst case reflections on BBSY, SSYN, and MSYN signal lines. DC Bus Load - DC bus loading is the amount of leakage current a module presents to a bus signal line. DC loading is expressed in terms of dc loads; one dc load is approximately 105uA. This rep- resents 89uA for the receiver plus 25uA for the transmitter (driver). One dc bus load can also be thought of as one bus driver and one bus receiver (or one transceiver). Most modules present 1 dc bus load to the Unibus. A Unibus seg- ment can support a maximum of 20 dc bus loads. This limit is set to maintain a sufficient noise margin. For more than 20 dc bus loads, a Unibus repeater option (DB11-A) or equivalent is used. (For example, a DT03 bus switch also contains a bus repeater.) The DB11-A imposes 1 dc bus load on the first Unibus section, but then it can drive an additional 19 dc bus loads and 50 feet of cable. DC loading is more important on SSYN than on data lines. Unibus Configuration Rules 1. Maximum cable length of a Unibus segment should not exceed 15.24m (50 ft). 2. Maximum dc loading should not exceed 20. 3. Maximum ac loading should not exceed 20 for any lumped load (unless the entire segment consists of one lumped load). 4. Different cable lengths should be selected when: a. a lumped load requires cables longer than 2.59m (8.5 ft) for its bus-in and bus-out connections. b. the sum of the ac loads in the two lumped loads connected by the cables exceeds 18. c. the sum of the ac loads in the two semi-lumped loads con- nected by the cables exceeds 36. The length of cables selected should differ by 1.52m (5 ft) or more with the longer cable connected to the semi-lumped load with the greater ac load if possible without increasing the total cable length more than 1.52m (5 ft). 5. Voltage margin tests should be completed for system accep- tance: a. when the system is originally configured, and b. when any Unibus element is added, deleted, or swapped (including a defective module or backplane). ================================================================================ Note 257.6 DMZ-32 DMA problems and FCO information 6 of 8 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 8 lines 3-MAY-1989 17:04 -< OK for the theory, how about in practice? >- -------------------------------------------------------------------------------- >> 5. Voltage margin tests should be completed for system accep- >> tance: >> a. when the system is originally configured, and >> b. when any Unibus element is added, deleted, or swapped >> (including a defective module or backplane). How does one do this operation? Is is also necessary/desirable for the Q-Bus? ================================================================================ Note 257.7 DMZ-32 DMA problems and FCO information 7 of 8 EISNER::HARVEY "Jack Harvey" 8 lines 4-MAY-1989 07:52 -< Margin Testing was Marginal >- -------------------------------------------------------------------------------- Tom's information is really old. The laws of physics haven't changed, so all the loading rules still apply, but voltage margin testing hardware disappeared long ago. The reason it disappeared is that it rarely ever did much good and cost quite a bit. And we never did have it for the Q-Bus equipment. ================================================================================ Note 257.8 DMZ-32 DMA problems and FCO information 8 of 8 EISNER::PROVOST 4 lines 4-MAY-1989 09:34 -< Misplaced NOTE >- -------------------------------------------------------------------------------- Not only was that old information, but it was mis-placed. I meant to reply to NOTE 357, but replied instead to 257. You will find more comments in the 357 stream.  ================================================================================ Note 258.0 RX33 -- RQDX ROM compatability 7 replies EISNER::MERRIMAN "George Merriman - CCA/NY" 4 lines 3-JAN-1989 16:39 -------------------------------------------------------------------------------- I have copied the following notes concerning RX33 compatability and RQDX ROM sets from the VMS conference. George Merriman -- HARDWARE_HELP co-moderator ================================================================================ Note 258.1 RX33 -- RQDX ROM compatability 1 of 7 EISNER::MERRIMAN "George Merriman - CCA/NY" 15 lines 3-JAN-1989 16:40 -------------------------------------------------------------------------------- ================================================================================ Note 371.0 RX33s on RQDX3 controller: 1.2 MB floppies no work 2 replies EISNER::NAGY "Frank J. Nagy, VAX Guru" 10 lines 3-JAN-1989 08:53 -------------------------------------------------------------------------------- We recently replaced the TK50 on a VAXStation-II in our LAVC with a pair of RX33 diskette drives connected to the RQDX3 controller (DUA0 on this controller is the RD53). Problem is that the system identifies the RX33 drives as RX50's (this from SHOW DEVICE/FULL). Along with this, we have not been able to read/write/mount 1.2 MB floppy diskettes. I'm assuming this is a software problem (which is why the question is here in VMS) and that I'm missing a SET command or a qualifier that I've not been able to find in the manuals or HELP. Anyone have an answer or an idea? ================================================================================ Note 258.2 RX33 -- RQDX ROM compatability 2 of 7 EISNER::MERRIMAN "George Merriman - CCA/NY" 28 lines 3-JAN-1989 16:42 -< ROM information >- -------------------------------------------------------------------------------- =============================================================================== Note 371.1 RX33s on RQDX3 controller: 1.2 MB floppies no work 1 of 2 EISNER::MAYHEW "Bill Mayhew" 3 lines 3-JAN-1989 11:35 -< Check RQDX3 ROM rev level >- -------------------------------------------------------------------------------- Only the newest (well... new_er_) ROMs on an RQDX3 will support an RX33. I dunno what ROM level is required, but this is a path worthy of investigation. ================================================================================ Note 371.2 RX33s on RQDX3 controller: 1.2 MB floppies no work 2 of 2 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 11 lines 3-JAN-1989 14:04 -< RQDX3 ROMs that will support RX33 drives >- -------------------------------------------------------------------------------- I have a pair of RX33 drives on my RQDX3 and they work fine except for not being able to format them. The RQDX3 was replaced on 16-Jan-87 and the ROMs are: DEC86 DEC86 LM8709 LM8705 244E5 243E5 I don't know what revision level this corresponds to. Alan ================================================================================ Note 258.3 RX33 -- RQDX ROM compatability 3 of 7 EISNER::KOZAM 7 lines 4-JAN-1989 22:53 -< ROM Upgrade hard to get >- -------------------------------------------------------------------------------- I tried to purchase these ROMS from DEC, but was unsuccessful (Yes, I tried unlisted parts, local field service). DEC's only answer was to purchase a new RQDX3 or DECMailer the old one. If you've got a field service contract, you shouldn't have trouble. By the way, I think that there are newer ROMS now (I don't know the numbers or why any new ones are needed). ================================================================================ Note 258.4 RX33 -- RQDX ROM compatability 4 of 7 EISNER::NAGY "Frank J. Nagy, VAX Guru" 2 lines 20-JAN-1989 08:16 -< Solution: Newer rev-level RQDX3 >- -------------------------------------------------------------------------------- Our problem has been solved and we have RX33s online (as RX33s) at last. The solution was to replace the RQDX3 with a newer revision level board. ================================================================================ Note 258.5 RX33 -- RQDX ROM compatability 5 of 7 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 4 lines 2-FEB-1989 00:00 -< Is this bug fixed? >- -------------------------------------------------------------------------------- >> Our problem has been solved and we have RX33s online (as RX33s) at last. >> The solution was to replace the RQDX3 with a newer revision level board. Can you format them under VMS v5.0? ================================================================================ Note 258.6 RX33 -- RQDX ROM compatability 6 of 7 EISNER::KOZAM 5 lines 6-FEB-1989 01:14 -< RQDX3 ROMS from Trimarchi >- -------------------------------------------------------------------------------- Trimarchi, Inc. has the new firmware for the RQDX3, priced at $ 75/set. I ordered a set a few days ago. It wasn't clear if their ROMs are an EXACT duplicate of DEC's or are of Trimarchi's own design. I'll know more in a few days. ================================================================================ Note 258.7 RX33 -- RQDX ROM compatability 7 of 7 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 2 lines 6-FEB-1989 08:19 -< THEY ARE SUPPOSE TO BE COPIES (HAPPY CUSTOMER) >- -------------------------------------------------------------------------------- > APPEND DMSLIB:DMSat $ 75/set. I ordered a set a few days ago. It wasn't clear if their > ROMs are an EXACT duplicate of DEC's or are of Trimarchi's own design.  ================================================================================ Note 259.0 POWER LINE MONITORS 1 reply EISNER::MERRIMAN "George Merriman - CCA/NY" 5 lines 3-JAN-1989 19:18 -------------------------------------------------------------------------------- I have extracted this note from POWER TOPICS to start a separate discussion on power monitors. This is to support the new title search facility. George Merriman -- HARDWARE_HELP co-moderator ================================================================================ Note 259.1 POWER LINE MONITORS 1 of 1 EISNER::MERRIMAN "George Merriman - CCA/NY" 14 lines 3-JAN-1989 19:19 -------------------------------------------------------------------------------- ================================================================================ Note 27.1 POWER TOPICS 1 of 39 EISNER::PROVOST 7 lines 21-JUL-1987 14:59 -< LINE MONITOR >- -------------------------------------------------------------------------------- We have some interesting power problems here at the Bates Linear Accelerator Center. Many years ago we had DIGITAL Field Service monitor our power. They used a DRANETZ line monitor, the output of which was quite useful. Unfortunately it suicided overnight. DIGITAL declined to loan us another at the time. We now have need for a similar device for similar reasons. Does anybody have experience or suggestions in this area?  ================================================================================ Note 260.0 computer room power contitioners and UPSs 11 replies EISNER::MERRIMAN "George Merriman - CCA/NY" 4 lines 3-JAN-1989 19:22 -------------------------------------------------------------------------------- I have extracted the following notes from POWER TOPICS in order to start a separate discussion on computer room power conditioners and uninterruptable power supplies. This is to support the new title search facility. ================================================================================ Note 260.1 computer room power contitioners and UPSs 1 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 16 lines 3-JAN-1989 19:24 -------------------------------------------------------------------------------- ================================================================================ Note 27.2 POWER TOPICS 2 of 39 EISNER::PROVOST 10 lines 21-JUL-1987 15:04 -< POWER GATE >- -------------------------------------------------------------------------------- We have found that preceding and following many power failures the power often fluctuates violently. It has been our practice to shut down all computer systems at the first hint of impending power fail, and to start them only after the power is smooth for some period of time. I remember hearing of a device built by a Florida company which would do this automatically. It provided the ability to set the time interval over which power must be bad before shutdown, the time interval over which power must be good before turn-on, and the definitions of good and bad power. Can anyone remind me who made the device and what it was called? Has anyone had experience with similar devices? ================================================================================ Note 260.2 computer room power contitioners and UPSs 2 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 23 lines 3-JAN-1989 19:26 -------------------------------------------------------------------------------- ================================================================================ Note 27.3 POWER TOPICS 3 of 39 EISNER::PROVOST 17 lines 21-JUL-1987 15:12 -< POWER CONDITIONING DEVICES >- -------------------------------------------------------------------------------- Knowing we were going to have power problems in a newly constructed computer area, we purchased and implemented a TOPAZ 35KVA power controller. After a period of lightning storms several devices would fail inexplicably and repeatedly in the area. When replaced, they would fail again. Eventually the TOPAZ would fail. The TOPAZ would then be bypassed until repair could be arranged (usually several months.) Once by-passed, everything would work. This process repeated every year around lightning season. Eventually we put in a switch enabling us to by-pass the power controller easily. Still later we stopped repairing it. I heard later that high voltage spikes had been detected in the output of such devices in their failure mode. This would explain our experience. Now the point: Has anyone successful experience with later models of this device, or with some other similar device? tjp ================================================================================ Note 260.3 computer room power contitioners and UPSs 3 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 26 lines 3-JAN-1989 19:27 -------------------------------------------------------------------------------- ================================================================================ Note 27.4 POWER TOPICS 4 of 39 EISNER::KASPER "Beverly T. Kasper" 20 lines 21-JUL-1987 16:50 -< Liebert products have served us well >- -------------------------------------------------------------------------------- When we built a new facility about 1-1/2 years ago, we installed a pair of Liebert Datawave Power Conditioners. They've been quite reliable, and aside from the fact that we need a UPS (Uninterrupted Power Supply), they meet our needs admirably. THey're controlled from a common console. The Liebert air conditioners in the room hook up to the same console, which has a piercing alarm horn. The power conditioners can be configured to restart if power comes up, or to stay down until manually reset (the option we chose, since we often get fluctuations here). The air units don't go through the power conditioners, so they come right back on when power comes up. This is useful, since it sometimes warms up a lot during the outage. This whole mess is also hooked up to the Halon system, which shuts off the air and power if Halon is discharged. What I think would be ideal for us is exactly this configuration plus a 5-10 minute UPS. This would allow us to ride brownouts, and to bring the systems down gracefully if it looked like we were in for a wait. ================================================================================ Note 260.4 computer room power contitioners and UPSs 4 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 16 lines 3-JAN-1989 19:32 -------------------------------------------------------------------------------- ================================================================================ Note 27.6 POWER TOPICS 6 of 39 EISNER::TABOR "Bill Tabor" 9 lines 21-JUL-1987 21:17 -< Motor generator >- -------------------------------------------------------------------------------- Living in thunderstorm alley, getting power from Florida Plunder and loot, and having been the manager of a site where my DEC field service people wanted to know if Planet Earth was grounded here, the solution to our power problem was a motor generator and power distribution system from a company called EPE in calf. After we installed this unit we went for 6 months with out a system crash. We used to average 2 to 3 weeks between major hardware failures. ================================================================================ Note 260.5 computer room power contitioners and UPSs 5 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 29 lines 3-JAN-1989 19:33 -------------------------------------------------------------------------------- ================================================================================ Note 27.7 POWER TOPICS 7 of 39 EISNER::GARDNER "Tim Gardner" 23 lines 22-JUL-1987 08:06 -< Funny you should ask... >- -------------------------------------------------------------------------------- > After a period of lightning storms several > devices would fail inexplicably and repeatedly in the area. When > replaced, they would fail again. Eventually the TOPAZ would fail. Gee, there are times when some corroboration is nice. Unfortunately, this isn't one of them. We have had numerous problems with our 50KVA Topaz unit, several of which sound similar to yours. Most recently, the RA81 failed twice and the RA60 shortly after that, all with failures in the drive power supplies. Typical you say. Random failures you say. Infant mortality you say. Well just for fun, consider that each of the failures were detected when we powered up the systems after the Topaz shut down. Coincedence? I don't think so. We have had other failures, mostly involving Topaz shutdowns for no apparent reason (no lightning in the area, other power conditioners in the building don't have any trouble.) I had previously believed that we simply had a lemon flavored Topaz. After reading .3 I'm not so sure anymore. T. ================================================================================ Note 260.6 computer room power contitioners and UPSs 6 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 29 lines 3-JAN-1989 19:36 -------------------------------------------------------------------------------- ================================================================================ Note 27.8 POWER TOPICS 8 of 39 EISNER::CETRON 22 lines 30-JUL-1987 02:05 -< dec cvc's are wonderfull... >- -------------------------------------------------------------------------------- We have been using dec's CVC's for 4-6 years now and have yet to blow anything due to passed thru power problems (and next door in CS, they have gone through 2 eagles and an ra80...) we did have some problems with very low voltages upping the currents due to the 'regulation' nature of the cvc and having nominal 20amp machines draw 30+ amps....this has been fixed (don't electricians have any pride in their work??)) also, the new dranetz's that digital uses are fully digital and do everything but monitor the atmospheric charge density of mars ...the last time we had one on, we had a state wide power failure and I have a beautiful log of the power slowly sagging to 30 volts and then surging to 700 volts on a 120 volt line....nothing was seen on the other side of the cvc.....I laugh everytime we have a less than 1 sec dropout and my computers are the only ones which stay up..... (out of almost 100 around campus...) -ed ================================================================================ Note 260.7 computer room power contitioners and UPSs 7 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 38 lines 3-JAN-1989 19:37 -------------------------------------------------------------------------------- ================================================================================ Note 27.9 POWER TOPICS 9 of 39 EISNER::LEDERMAN "Bart Z. Lederman" 31 lines 30-JUL-1987 19:13 -< If you really want reliable power.. >- -------------------------------------------------------------------------------- > (don't electricians have any pride in their work??)) No, they don't. Whenever we had new wiring done we had to check it ourselves to see that they didn't put the circuit breaker in the neutral line, or wire the hot lead to the ground pin of the power plug. (I think the electricians we have now are better). We have to operate 24 hours a day, every day, no matter what. Con Ed power goes through transformers and is rectified to DC, where it "floats" over banks of 48 volt lead-acid batteries. It then goes through inverters which convert it back to 120/208 Volt AC power for the computers, modems, fire alarms, and other essential equipment. Even if Con Ed dies completely, everything keeps running until the gas turbine generators come on-line. (In less critical applications, it might be enough to just have 10 minutes of battery capacity without the generators, if you want to ride out short outages, or just have enough time to shut down cleanly after saving to disk/tape, etc.) No matter what comes in on the power line this system can handle it. If there is a long outage we can switch in some lights, and even feed power to the elevators if we have to (no joke in a 33 story building). Equally important: we run practice drills periodically (on weekends, when there is relatively little demand on the systems) where power is deliberately cut and we see if the system really works. One of our competitors had a nearly identical power system but had never tested it "all the way" as we had, so when a real power failure hit... We also shut down and disconnect 1/2 of each redundant system before these tests just in case there is a transient that damages something. ================================================================================ Note 260.8 computer room power contitioners and UPSs 8 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 33 lines 3-JAN-1989 19:43 -------------------------------------------------------------------------------- ================================================================================ Note 27.10 POWER TOPICS 10 of 39 EISNER::DROPPERS "Seton R Droppers" 26 lines 5-AUG-1987 14:40 -< UPS, New Buildings, elbows, etc. >- -------------------------------------------------------------------------------- Here at PBS we feel it would be highly impolite (embarrassing?) to all of a sudden drop the feed of your favorite PBS show. In response we have installed an UPS (15 minutes) and backup generator (no not gas turbine, simply diesel). All of this worked fine during our "live" tests when we cut power to the technical center. Only problem was when the commercal power went out during one of the areas more impressive thunder boombers the A/C units dropped out. Before we got power back it approached 85 in the computer room. The problem turned out to be that the generator kicked in so quickly that the A/C could not recycle and restart. Repairs are underway to add a delay to the A/C unit startup. We have never had any down time due to comercal power loss, only when a contracters go-fer hit the emergancy power button for the computer room... (At least they didn't try to find out what happened when they pulled the halon release...) Remember: . Cover you switch . Attempt to test not only your technical center/computer room, but, if you can, try shutting the whole building off . Your UPS/generator is only as good as your A/C, (at least in long run) ================================================================================ Note 260.9 computer room power contitioners and UPSs 9 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 25 lines 3-JAN-1989 19:44 -------------------------------------------------------------------------------- ================================================================================ Note 27.24 POWER TOPICS 24 of 39 EISNER::MERRIMAN "George Merriman" 19 lines 8-OCT-1987 20:10 -------------------------------------------------------------------------------- > After a period of lightning storms several devices would fail > inexplicably and repeatedly in the area. When replaced, they > would fail again. Eventually the TOPAZ would fail. > I had previously believed that we simply had a lemon flavored > Topaz. After reading .3 I'm not so sure anymore. I once had a Topaz power unit that went unstable during a brown-out and faded the output line voltage between 0 and 90 volts at a rate of about one second. I'm sure that did the disk drive a lot of good. Speaking of power conditioners, can anyone tell me why it is that I can buy a box to set next to my computer to filter out all the glitches that my computer maker claims will ruin his fancy computer but that same computer maker can't build the same functionality into his equipment's power supplies? ================================================================================ Note 260.10 computer room power contitioners and UPSs 10 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 41 lines 3-JAN-1989 19:45 -------------------------------------------------------------------------------- ================================================================================ Note 27.25 POWER TOPICS 25 of 39 EISNER::KENNEDY "Terry Kennedy" 34 lines 10-OCT-1987 00:00 -< Cost is the main (only) factor >- -------------------------------------------------------------------------------- > Speaking of power conditioners, can anyone tell me why it is that > I can buy a box to set next to my computer to filter out all the > glitches that my computer maker claims will ruin his fancy > computer but that same computer maker can't build the same > functionality into his equipment's power supplies? Well, I used to be one of those computer manufacturers (president of one, even - who let me in?) and this was one I was asked many, many times. So, here is the recorded explanation: Back when computers were very expensive, it made good sense to include a many-thousand dollar piece of power conditioning gear with the system. Many systems of that period required odd power, anyway, so the power con- ditioner could double as the power supply. For example, my old IBM 370/138 had a motor-generator set which eliminated sags/spikes as well as providing the 400 Hz power the system required. Yet, today, computers have reduced in size and cost far more rapidly than useful power protection gear, which is still largely mechanical (witness motor/generators, batteries, etc.) When the vendor is selling a $10,000 computer the $2,500 for a good, solid UPS is 1/4 the price, or maybe all of the profit. People also ask - why $2,500 for a UPS - I can go to Radio Shack or some place like that and get the same protection for $29.95. The answer is, it is nowhere near the same protection. The 30-dollar box will filter out maybe 1/2 to 2/3 of your minor power line spikes, but does nothing for dips or surges/sags. The $2,500 unit will keep your MicroVAX running during a complete power outage, as well! [Note - I no longer sell systems or UPS gear - in fact I now buy them! I do have definite experience with many popular units - if anyone wants to sit through a sermon on what works and what doesn't (and why), MAIL me a note and I'll post another reply here] ================================================================================ Note 260.11 computer room power contitioners and UPSs 11 of 11 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 7 lines 6-APR-1989 17:46 -< pointer to related note 262.8 >- -------------------------------------------------------------------------------- Now that POWER stuff has been split up it isn't always clear where to post things, as spikes or lighting bolts or phoneline spikes can get your computer room as well as your PC, and similar or even identical protective hardware applies. 262.8 perhaps should have been here, so don't skip it just cuz its in a 'PC' thread and you have no PCs.  ================================================================================ Note 261.0 Should terminals be turned off when not in use? 20 replies EISNER::MERRIMAN "George Merriman - CCA/NY" 5 lines 3-JAN-1989 19:48 -------------------------------------------------------------------------------- I have extracted the following notes from POWER TOPICS to start a separate discussion on whether or not to turn of terminals at night. This is to support the new title search facility. George Merriman -- HARDWARE_HELP co-moderator ================================================================================ Note 261.1 Should terminals be turned off when not in use? 1 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 12 lines 3-JAN-1989 19:49 -------------------------------------------------------------------------------- ================================================================================ Note 27.11 POWER TOPICS 11 of 39 EISNER::ERNST_R "Bob Ernst" 5 lines 10-AUG-1987 18:31 -< Terminals & Power >- -------------------------------------------------------------------------------- There seems to be a controversy at most sites about shutting off terminals at night. Some people point to the savings in power and to the avoided damage during lighting storms. Other people point to the increased wear and tear on the machines (and people's patience). Does anyone know of any studies on this topic? ================================================================================ Note 261.2 Should terminals be turned off when not in use? 2 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 26 lines 3-JAN-1989 19:50 -------------------------------------------------------------------------------- ================================================================================ Note 27.12 POWER TOPICS 12 of 39 EISNER::KILLEEN "Jeff Killeen" 19 lines 10-AUG-1987 20:03 -< NOT SCIENTIFIC >- -------------------------------------------------------------------------------- >>> There seems to be a controversy at most sites about shutting off terminals >>> at night. Some people point to the savings in power and to the avoided >>> damage during lighting storms. Other people point to the increased >>> wear and tear on the machines (and people's patience). Does anyone >>> know of any studies on this topic? We manage 78 PDP-11 sites with over 2000 terminals. Those who shut off their terminals at night have had more problems than those who do not - with two execeptions. 1. Terminals without screen savers 2. Wyse 85's - they have heat problem and their cases will turn yellow after a year if left on all the time. There is no question that turning the systems on and off causes many problems. The two sites we had the most problems with were turning the systems on and off. ================================================================================ Note 261.3 Should terminals be turned off when not in use? 3 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 19 lines 3-JAN-1989 19:51 -------------------------------------------------------------------------------- ================================================================================ Note 27.13 POWER TOPICS 13 of 39 EISNER::DELARISCH 12 lines 11-AUG-1987 16:53 -< Leave Them On >- -------------------------------------------------------------------------------- I agree with Jeff. We have found that leaving our terminals have greatly decreased our problems (especially with the VT240 series). As for the lighting, we are in South Florida (Boca Raton) and we have yet (In our Department) had a verifiable lighting strike on our power (unconditioned & w/o UPS). I don't know if we are just lucky or what. Just think of the shock the terminal (or any IC based equipment) receives when it is turned on and off. If the terminal has a screen saver function ... leave it ON! -Arnold ================================================================================ Note 261.4 Should terminals be turned off when not in use? 4 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 16 lines 3-JAN-1989 19:52 -------------------------------------------------------------------------------- ================================================================================ Note 27.14 POWER TOPICS 14 of 39 EISNER::FRISBIE "Alan E. Frisbie" 9 lines 11-AUG-1987 22:12 -< What about color monitors? >- -------------------------------------------------------------------------------- What about the color VAXstation monitors (VR290)? Do the color tubes deteriorate over time with the power on, or should we follow the same rules as for terminals? Personally, I leave my (monochrome) VAXstation and monitor on 24 hours a day unless I will be away for more than a day or two. So far no problems. Alan ================================================================================ Note 261.5 Should terminals be turned off when not in use? 5 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 16 lines 3-JAN-1989 19:53 -------------------------------------------------------------------------------- ================================================================================ Note 27.16 POWER TOPICS 16 of 39 EISNER::NAGY "Frank J. Nagy, VAX Guru" 9 lines 24-AUG-1987 07:55 -< Poof go the VT100s >- -------------------------------------------------------------------------------- When Fermilab installed the Accelerator Control System in 1981, several VT100s were acquired with the initial equipment. Many of these VT100s were in public areas and left on continuously. Luckily, these VT100s were on Digital maintenance, because after about 5 years we discovered that the video boards would die violents (screen would brighten, dim and then shink to vanishing dot - this was accompanied by a "poofing" noise and a fair amount of smoke out the top of the case). Of course, this did not happen to all the VT100s but we did experience 1-2 failures per week for about 2 months. :-) ================================================================================ Note 261.6 Should terminals be turned off when not in use? 6 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 13 lines 3-JAN-1989 19:54 -------------------------------------------------------------------------------- ================================================================================ Note 27.17 POWER TOPICS 17 of 39 EISNER::KASPER "Beverly T. Kasper" 6 lines 2-SEP-1987 15:07 -< Leave 'em on >- -------------------------------------------------------------------------------- I agree with leaving them on -- we've had a lot more failures on terminals which get power cycled. I leave VT1xx's on, too, usually -- just use reset or setup/9/9 to clar the screen. Is there some reason that this is a bad idea? ================================================================================ Note 261.7 Should terminals be turned off when not in use? 7 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 17 lines 3-JAN-1989 19:55 -------------------------------------------------------------------------------- ================================================================================ Note 27.18 POWER TOPICS 18 of 39 EISNER::TILLMAN "Brian Tillman, Lear Siegler, Inc." 10 lines 3-SEP-1987 09:37 -< Turn 'em off. >- -------------------------------------------------------------------------------- RE: < Note 27.17 by EISNER::KASPER "Beverly T. Kasper" > -< Leave 'em on >- | I leave VT1xx's on, too, usually -- just use reset or setup/9/9 to | clear the screen. Is there some reason that this is a bad idea? Other than the fact that it's environmentally unsound, no. Personally, I'd rather save the electricity for my children and their contemporaries than squander our world's resources on keeping a terminal running because I MIGHT experience a little inconvenience. ================================================================================ Note 261.8 Should terminals be turned off when not in use? 8 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 14 lines 3-JAN-1989 19:56 -------------------------------------------------------------------------------- ================================================================================ Note 27.19 POWER TOPICS 19 of 39 EISNER::KASPER "Beverly T. Kasper" 7 lines 3-SEP-1987 11:22 -< Question is: does it hurt the VT1xx to be left on? >- -------------------------------------------------------------------------------- My reason for leaving them on is *Not* to avoid "a little inconvenience." I was treated to a lecture back in college (we had VT52's; no screen save), from a EE friend, on how bad it was for electronic equipment to have power cycled repeatedly. I therefore leave the terminals on because I believe that I am saving my company (and my department, I might add) from the expense of having my terminal repaired. ================================================================================ Note 261.9 Should terminals be turned off when not in use? 9 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 9 lines 3-JAN-1989 19:57 -------------------------------------------------------------------------------- ================================================================================ Note 27.20 POWER TOPICS 20 of 39 EISNER::TILLMAN "Brian Tillman, Lear Siegler, Inc." 2 lines 3-SEP-1987 11:54 -< I don't think the facts support terminal damage >- -------------------------------------------------------------------------------- I've always turned by terminals off every night when I go home and in 12 years I've never had one fail. ================================================================================ Note 261.10 Should terminals be turned off when not in use? 10 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 28 lines 3-JAN-1989 19:58 -------------------------------------------------------------------------------- ================================================================================ Note 27.21 POWER TOPICS 21 of 39 EISNER::LEDERMAN "Bart Z. Lederman" 21 lines 6-SEP-1987 20:29 -< Turn them off. Period. >- -------------------------------------------------------------------------------- As someone who used to make a living fixing TV and other electronic equipment, and who has a Masters Degree in E.E., etc.... If the equipment comes anywhere close to being properly designed, it will not be hurt by being turned on. People who think so are probably being influenced by 30 or more year old thinking having to do with vacuume tubes. Second, there is very definately a problem leaving any CRT equipment on in that the deflection and high voltage circuits have to be running at full power. Screen saving circuitry (which VT100s don't have anyway) don't help. These are the circuits which are most likely to fail, and the longer they are on the more likely they are to fail. Next comes the CRT itself, and it's life is based almost entirely on how long it's powered up. Quite apart from conserving electrical energy (a sentiment I agree with), your CRT's are much more likely to last a long time if they are turned off when they are not to be used for any length of time (nights and weekends). You may not save much if you turn it of for 1/2 hour during lunch. ================================================================================ Note 261.11 Should terminals be turned off when not in use? 11 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 30 lines 3-JAN-1989 19:59 -------------------------------------------------------------------------------- ================================================================================ Note 27.22 POWER TOPICS 22 of 39 EISNER::KENNEDY "Terry Kennedy" 23 lines 6-SEP-1987 22:56 -< A responsible opposing viewpoint... >- -------------------------------------------------------------------------------- > ... thirty or more year old thinking having to do with vacuum tubes But, the CRT display is a vacuum tube - granted, the only one, but... Also, there are two other issues to take into account: First, it is much harder to build/buy a properly designed switch-mode power supply than ti build/buy a badly designed one which works, most of the time. Switchers are very itchy at turn-on time, and if you have the slightest bounce in the power switch, poof goes the supply! Since DEC exhibis the poor tast to buy ASTEC (pronounced spazz-tech) supplies, I wouldn't turn anything on or off more than I have to. For an example of ASTEC in inaction, find a Radio Shack Model 2 (you'll irritate the owner of it less than the owner of a Micro-11 when you blow up the supply). Now, assuming the unit is on, flip the power switch off and then RIGHT back on. Observe the whine and the pop. This is the same thing that will happen if the switch bounces when you power your DEC terminal on/off. Secondly, much RS-232 equipment (including some of DEC's), will inter- pret 0 volts on the signal line as a BREAK keypress. It isn't supposed to happen that way, but it does. If the computer you are connected to stays on, so should the terminal. ================================================================================ Note 261.12 Should terminals be turned off when not in use? 12 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 21 lines 3-JAN-1989 20:00 -------------------------------------------------------------------------------- ================================================================================ Note 27.23 POWER TOPICS 23 of 39 EISNER::ROCHELEAU "Rod Rocheleau" 14 lines 8-SEP-1987 11:21 -< Leave 220's on ... power switch will fail >- -------------------------------------------------------------------------------- >> >> Secondly, much RS-232 equipment (including some of DEC's), will inter- >> pret 0 volts on the signal line as a BREAK keypress. It isn't supposed >> to happen that way, but it does. If the computer you are connected to >> stays on, so should the terminal. >> Hear hear ... This is most true. I have a data switch that see's a dead terminal as a break. It put their drop out of service. This cure's the problem of people turning off their terminal. One other comment. If you have VT220 you would be advised to leave the power switch alone. It is very prone to failure ... in the off state. ================================================================================ Note 261.13 Should terminals be turned off when not in use? 13 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 45 lines 3-JAN-1989 20:02 -------------------------------------------------------------------------------- ================================================================================ Note 27.27 POWER TOPICS 27 of 39 EISNER::CONROY 38 lines 1-JUL-1988 17:07 -< My 2 cents >- -------------------------------------------------------------------------------- Here's my 2 cents on turning electrical equipment on and off. My company repairs monitors (only when asked, but we've been asked a lot) and what we have discovered is that certain discrete components have a problem with power cycling (turning on and off and ...). Basically, they eventually wear out. Obviously, mechanical parts (like switches) get wear and tear, but it has been our experience that they last longer than many other things. Also, the longer a semiconductor is running, the sooner it will die. But, again, this can take longer than the life of the equipment. In fact, a well designed and constructed terminal will last a LONG time. So much for well designed and constructed devices. Now to DEC terminals. The VT100s had an apparent manufacturing flaw which caused a cold-solder joint. After enough cycles this would break down and cause a resistor to smoke. Hence the problem with "puffs of smoke" and other problems mentioned in a previous message. The fix is simple: resolder the connection. This has fixed more VT100s than anything else I've seen. Of course, if you don't know how, you can take it to DEC for a "nominal" fee. My opinion would be, find some third-party company or buy a new unit. Some of the early VR201 monitors (used on Rainbows, PROs, VT220s, etc) were built with a coil that was marginal enough that it would break after about two-three years of normal use ("normal" being one cycle per day). This would cause a resistor to burn and the screen to go out. The replacement part is, I believe, an improvement. However, none of our repaired units has been in the field long enough to know for a fact. The parts cost about $5 and can be replaced in about 1/2 hour. OR you can have DEC fix it for $120.00 - or the price of a new monitor. I have mentioned some specifics because that is the only safe thing to do. Based on my experiences I personally prefer to leave equipment on constantly (less cycles=less problems). But that can lead to problems (witness power spikes), although turning terminals off my not prevent such things. There are a lot of variables invovled, such as the type of equipment, the rev level of the equipment, how clean your local power is, how clean your building power is, the weather, and the phase of the moon :-} - (just kidding about that!) The temptation is, of course, to generalize from one's own experience, which may or may not be valid for others. So, I'll leave my conclusions at this: For people with VR201 monitors and/or VT100s who are in the Seattle metro area - you'd probably come out ahead if you leave your equipment on over-night (assuming you reduce the brightness on the VT100s as much as possible when not in use). ================================================================================ Note 261.14 Should terminals be turned off when not in use? 14 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 12 lines 3-JAN-1989 20:03 -------------------------------------------------------------------------------- ================================================================================ Note 27.28 POWER TOPICS 28 of 39 EISNER::MERRIMAN "George Merriman - CCA/NY" 5 lines 1-JUL-1988 17:23 -< Semiconductors don't like power cycles either >- -------------------------------------------------------------------------------- > Also, the longer a semiconductor is running, the sooner it will die. Some semiconductor devices, particularly "power" devices, are very susceptible to temperature cycling problems and this is often what sets their lifetime. ================================================================================ Note 261.15 Should terminals be turned off when not in use? 15 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 16 lines 3-JAN-1989 20:04 -------------------------------------------------------------------------------- ================================================================================ Note 27.29 POWER TOPICS 29 of 39 EISNER::SHUMAKER 9 lines 7-JUL-1988 16:42 -< Equipment on and unattended m be a problem... >- -------------------------------------------------------------------------------- Re - Leaving equipment on. If you do leave equipment on and unattended, do go through your fire insurance policy carefully for possible requirements for smoke and/or temperature detectors which ring alarms at a guard station, or fire station, or otherwise attract the attention of humans. Mark Shumaker ================================================================================ Note 261.16 Should terminals be turned off when not in use? 16 of 20 EISNER::MCCORMICK_B "Bob McCormick" 19 lines 3-JAN-1989 20:29 -< Turn 'em off... >- -------------------------------------------------------------------------------- I bet your terminal is only in use 8 to 10 hours a day ... =1/3 or you waste 66% power X how many tty's left on in the country... power to the people right on! I'm employed by a DEC Distributor ... installed base of over 60 VAX systems, and many thousand terminals ... all turned off at night ('cept the console tty: video on low end; hardcopy otherwise) Only known problem was with a bunch of Wyse terminals - replaced by VT320's ... Suggest turn on and off once a day ... If you have them under warranty or support why not? As for the issue about the Break character (as in what happens to your VAXstation when you connect the LN03 to the console, and power IT up and down) unless the terminal is connected to the console or to something funny (like a port selector, etc) that gets upset at random breaks, who cares? Try using DECconnect stuff with DS200/DL and VT320's ... plug and unplug causes breaks... ================================================================================ Note 261.17 Should terminals be turned off when not in use? 17 of 20 EISNER::CONROY "Alan Conroy" 17 lines 4-JAN-1989 14:44 -< A different perspective, Bob >- -------------------------------------------------------------------------------- > I bet your terminal is only in use 8 to 10 hours a day ... =1/3 > or you waste 66% power X how many tty's left on in the country... > power to the people right on! How much power does an average terminal use - maybe 20 watts? You might as well start at home - you'll save a lot more power than that... > Suggest turn on and off once a day ... > If you have them under warranty or support why not? Warranties do not last forever and many (most?) of us do not have our terminals under maintenance - it is cheaper to do depot repair. The cost is further reduced by limiting how often we make use of the repair (by being nice to the equipment). True fact: The only terminal that completely failed here last year was one that was dutifully turned off each night and on in the morning. ================================================================================ Note 261.18 Should terminals be turned off when not in use? 18 of 20 EISNER::MERRIMAN "George Merriman - CCA/NY" 4 lines 4-JAN-1989 17:53 -< Can't write much code on a broken terminal >- -------------------------------------------------------------------------------- > Suggest turn on and off once a day ... > If you have them under warranty or support why not? ^^^^^^^^ Because I'm out of work while waiting for the terminal to get fixed! ================================================================================ Note 261.19 Should terminals be turned off when not in use? 19 of 20 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 10 lines 9-JAN-1989 11:58 -< Power effects on other monitors >- -------------------------------------------------------------------------------- Just a note on other kinds of CRTs and monitors : We have a lab that does calibrations on our image display station monitors once a week. They report that if a monitor is turned off for the night, then turned on again in the morning, it can take up to three hours for the tube to stabilize. This isn't important for your VT100, VT220, etc., but if you have an image display station where it is critical that a pixel in one spot be just as bright as one elsewhere, turning the monitor off a night means you don't get reliable results until after lunch. ================================================================================ Note 261.20 Should terminals be turned off when not in use? 20 of 20 EISNER::DOLAR 13 lines 26-JUN-1989 15:29 -< A passing thought >- -------------------------------------------------------------------------------- Whatever you decide to do, remember this story: At a bank I worked at years ago, a senior VP got an industry study that showed that many $$ were being wasted on devices that were being left powered up at night. Terminals were cited in the study as specific culprits, and charts, etc. showed how much could be saved. An edict came from above in the form of a memo, saying that from that day forward, all terminals would be powered off at night. Most of these terminals had not been powered off in years. Well, that night 1500 IBM-type terminals were powered off. The next morning when they were powered on, several hundred power supplies blew out.  ================================================================================ Note 262.0 PC Power Controller/Conditioner Equipment 12 replies EISNER::MERRIMAN "George Merriman - CCA/NY" 3 lines 3-JAN-1989 20:07 -------------------------------------------------------------------------------- I have extracted the following notes from POWER TOPICS in order to start a separate discussion on PC power controller equipment. This is to support the new title search facility. ================================================================================ Note 262.1 PC Power Controller/Conditioner Equipment 1 of 12 EISNER::MERRIMAN "George Merriman - CCA/NY" 18 lines 3-JAN-1989 20:08 -------------------------------------------------------------------------------- ================================================================================ Note 27.30 POWER TOPICS 30 of 39 EISNER::SIMONS "Paul, Not that CONVEX!" 11 lines 20-DEC-1988 10:47 -< PC power controllers >- -------------------------------------------------------------------------------- We are in the process of installing a number of PC's. It was decided that they should be accompanied by one of those slim boxes that fit between the system unit and the monitor. The ones that have separate switches for the monitor, modem, hard drive, system unit, etc. These boxes have easily generated the most controversy associated with the procurement of these systems. Can anyone provide me with a reasonable specification for these devices? For instance, is UL approval necessary? What is a reasonable figure for "surge suppression"? Models from Radio Shack and Tripp Lite cost about $80.00, whereas Black Box is selling one for $170.00. What's the difference? ================================================================================ Note 262.2 PC Power Controller/Conditioner Equipment 2 of 12 EISNER::MERRIMAN "George Merriman - CCA/NY" 35 lines 3-JAN-1989 20:10 -------------------------------------------------------------------------------- ================================================================================ Note 27.31 POWER TOPICS 31 of 39 EISNER::CONROY "Alan Conroy" 28 lines 20-DEC-1988 12:57 -< Some answers >- -------------------------------------------------------------------------------- > is UL > approval necessary? What is a reasonable figure for "surge suppression"? > Models from Radio Shack and Tripp Lite cost about $80.00, whereas > Black Box is selling one for $170.00. What's the difference? Whether or not UL approval is "necessary" depends upon who okays the purchase. There are no laws that I know of that require UL approval on anything, but I'd check with your local Fire Department/Building Inspectors/etc to be sure. The difference between different power conditioning equipment (besides price) includes such things as noise filtering, power draw through the equipment, speed of surge supressing, etc. What you NEED depends entirely upon your situation. Do you have very clean power coming into your building (some buildings have their own power conditioning)? If so, you need less protection that others might. Of course, will you always have clean power? Do you want something that will keep your PCs up long enough to save your work after a power failure (a la UPS's)? The amount of power that your PCs need will have to be supplied by the equipment you choose. For example, if your CPU takes 2 watts, your disk takes another 2 watts, and your monitor takes another 2 watts, then you need equipment that can handle at least 6 watts continuous and probably twice that for short periods of time. Do you want fuses on the equipment? Breakers? On/off switches? What about other considerations - like handling line voltage drops (which are more common{than spikes)? So, as you can see, there is no simple answer to your question about "reasonable figures". Someone else can probably give more details about things like noise ratios, cut-off time, etc. But to make things simple: assuming all other things are equal, the more expensive boxes generally offer you more protection. Also, having some protection is probably better than having none. Personally, I'd never buy ANYTHING from Radio Shack unless I was desperate. ================================================================================ Note 262.3 PC Power Controller/Conditioner Equipment 3 of 12 EISNER::MERRIMAN "George Merriman - CCA/NY" 17 lines 3-JAN-1989 20:13 -------------------------------------------------------------------------------- ================================================================================ Note 27.32 POWER TOPICS 32 of 39 EISNER::MERRIMAN "George Merriman - CCA/NY" 9 lines 20-DEC-1988 13:56 -< UL approval sometimes needed >- -------------------------------------------------------------------------------- > Whether or not UL approval is "necessary" depends upon who okays the > purchase. There are no laws that I know of that require UL approval on > anything, but I'd check with your local Fire Department/Building > Inspectors/etc to be sure. I think some jurisdictions require UL approval in their fire and building codes -- Los Angeles comes to mind. Your insurance carrier might be interested as well -- the UL approval is granted by an insurance industry body. ================================================================================ Note 262.4 PC Power Controller/Conditioner Equipment 4 of 12 EISNER::MERRIMAN "George Merriman - CCA/NY" 16 lines 3-JAN-1989 20:14 -------------------------------------------------------------------------------- ================================================================================ Note 27.34 POWER TOPICS 34 of 39 EISNER::HARVEY "Jack Harvey" 9 lines 20-DEC-1988 18:42 -< Asking us will only get opinions >- -------------------------------------------------------------------------------- >< Note 27.30 by EISNER::SIMONS "Paul, Not that CONVEX!" > > -< PC power controllers >- > > We are in the process of installing a number of PC's. Why not test? Put "power controllers" on ten percent of the PC's. Then see if you can tell the difference. P.S. If you do, you must report the results back here. ================================================================================ Note 262.5 PC Power Controller/Conditioner Equipment 5 of 12 EISNER::MERRIMAN "George Merriman - CCA/NY" 19 lines 3-JAN-1989 20:15 -------------------------------------------------------------------------------- ================================================================================ Note 27.35 POWER TOPICS 35 of 39 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 12 lines 21-DEC-1988 13:23 -< Get UL approved equipment! >- -------------------------------------------------------------------------------- >> I think some jurisdictions require UL approval in their fire and >> building codes -- Los Angelese comes to mind. My former employer (in Los Angeles) got a visit from the Fire/Building/???? Inspector who found several items of non-UL approved equipment (printers, etc.). It cost a bunch of bucks to have the equipment inspected by the city and get "LA City Approved" stickers for all of them. It isn't worth the problems to get non-UL approved equipment. If we hadn't been super cooperative with the inspector, we could have had our power shut off that day. ================================================================================ Note 262.6 PC Power Controller/Conditioner Equipment 6 of 12 EISNER::MERRIMAN "George Merriman - CCA/NY" 43 lines 3-JAN-1989 20:18 -------------------------------------------------------------------------------- ================================================================================ Note 27.38 POWER TOPICS 38 of 39 EISNER::KENNEDY "Terry Kennedy" 36 lines 28-DEC-1988 02:17 -< I am not impressed by them... >- -------------------------------------------------------------------------------- > Nobody out there uses these things? We had to set up several dozen PC's for people who never used them before (administrative secretaries, etc.). The base configuration was a turbo XT (10 Mhz V20), color display/adapter, 20Mb hard disk, 80-column printer, power controller, and 2 pieces of security hardware. So far, most of the power controllers have failed, the users have demanded mono monitors instead of CGA (which *they* wanted in the first place), and the security hardware has been yanked. So, feel free to use this as a case history. Also, ask them why they want a power controller - the three plugs you have (system unit, monitor, and printer) could just as well be plugged into a switched outlet strip... About the quality (or lack thereof) of these things - the majority of them are made in Taiwan, etc. and don't receive the care of assembly that motherboards, for example, get. Many of the power-related items coming from there have UL stickers - but that's all - they've never been tested and the stickers are fake. 'Buying American' in this case will get you a higher-priced piece of gear, but it's still a glorified on-off switch. I don't think much of the protection afforded when the suppression parts have one leg tied to the ground leg of the cord, but the cord is plugged into a 2-prong wall socket with an ungrounded 'cheater'... Most of your machine problems relating to power will be from things like a photocopier on the same circuit, etc. which *cannot* be helped by such devices. Proper wiring is a must, and any attempt to correct for bad wiring after the wall socket will be less than optimum. For those really concerned with power surges, there are large (20- to 500-amp) surge suppressors designed to be mounted next to the power dis- tribution panel, and can be used to protect one circuit or the whole panel, depending on rating... ================================================================================ Note 262.7 PC Power Controller/Conditioner Equipment 7 of 12 EISNER::MERRIMAN "George Merriman - CCA/NY" 43 lines 3-JAN-1989 20:18 -------------------------------------------------------------------------------- ================================================================================ Note 27.39 POWER TOPICS 39 of 39 EISNER::MCCORMICK_B "Bob McCormick" 35 lines 1-JAN-1989 12:54 -< Two Professional Solutions >- -------------------------------------------------------------------------------- Most of the $20-$50 surge suppressors that you plug into the duplex outlet and then plug your equipment into I stay away from. I've got a couple of them that have failed the test of time ... especially after a lightning strike nearby! I have invested quite a bit of money in protection devices for RF + computer equipment (CMOS) that sits on top of a mountain (a hi-tec amateur radio repeater). Protection for antenna coax, connections to the telephone network, external devices hooked up to a computer controller, and (of course) the power source have all been implemented. For protection of things like RS232 data lines you should look for specs like: 500 joule surge handling capability with a maximum surge current of 25,000 amps 20 volt peak clamping voltage and 100 ma load current ... with a response time of less than 5 ns. For AC power lines look for 5,000 amps maximum surge current, and around 500 joule energy dissipation (for L5-15R 15 amp outlets). Surge eliminators should provide protection for things like HV spikes, over voltages and lightning surges. Many of the generic surge suppressors only use one type of suppression technology: varistor, MOV, gas tubes. Some features that you may want to look for in a surge suppressor: High and low frequency filters, voltage control, and high energy dissipation. A fuse in series with the power which is caused to blow when high energy dissipation is shorted to ground also is worthwhile. Two companies that I have obtained protection devices from, that have passed the test of time, are: PolyPhaser Corporation LEA Dynatech Inc. P O Box 1237 12516 Lakeland Road Gardnerville NV 89410 Santa Fe Springs CA 90670 702.782.2511 213.944.0916 Both orgainzations deal direct as well as through distributors. I don't claim to be an expert in any of these areas ... ================================================================================ Note 262.8 PC Power Controller/Conditioner Equipment 8 of 12 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 148 lines 5-APR-1989 21:51 -< Spike, Surge, Lightning protection info >- -------------------------------------------------------------------------------- Various power, phone, and building suppressor products - some with educational sales literature. _____________________________________________________________________________ I got an interesting catalog from Leviton that is WELL WORTH reading. It is their Transient Voltage Surge Suppression (TVSS) Devices catalog. The actual product descriptions are not what I and refering to. The good portion of the catalog is devoted to educating the reader about the different waveforms, where each type originates, and how to suppress them. They build a network of suppressors that start at your service entrance, and can continue right up to past your outlet. There is even a remote status capability to trigger an alarm if something gets zapped to where it no longer can protect. And, YES, there are DIFFERENCES is power strip TVSSs, now you can learn more details. Leviton is the name I always associated with desk lamp sockets that are broken, and replacement ones that have even less strength than their predecessor. That is their 'home' grade stuff, but something is going on in that company! I have now seen several new catalogs for various groups of products that are well designed good products. Their other catalog worth mentioning here is their Telecom products catalog that covers a very flexible modular and BNC type DATA outlet that you assemble to suit your needs. The styling and approach is different than the vanilla DEC or MODTAP fill the hole in the flat faceplate. Their TVSS phone # in NY is 718-229-4040 and x6384 is for literature. Their Telecom group in WA? is 206-485-4288. _____________________________________________________________________________ For Telco central office phone line protectors, there are many standard companies. Most all make modules that plug into each other's blocks. These range from simple carbon block lightning arrestors (that also create crackles on your line as they get zapped many times) to very sophisticated multiple element devices where very fast triggering devices are backed up by slower ones that can dissipate more total energy. All that is backed up by fuse or heat coil (think slow circuit breaker) for steady overcurrent that could start a fire, but that doesn't have the voltage to fire the other devices. This 'sneak current' could be caused by an accidental power line cross to 120 volts which is less than ringing voltage, so arrestors have to be set way higher, but has to be dealt with before the small guage wire starts cooking. Each plugin unit does a pair, and the blocks they plug into have stub cables, 66 punch blocks, wire-wrap, or 25 pair phone plugs in various combinations for input and output. Porta Systems makes a DELTA (tm) protector that does all you want and plugs into AT&T, COOK, RELIABLE, etc. blocks. You don't have to be a telco CO to use these. Put them right after the DEMARC where the TELCO hands off wires to you. Run ALL your 2 wire phone and 4 wire data lines through them. If your DEMARC is in the basement, and your computer room is on the 25th floor, you need something there, too (but it is different so see their literature). Your modems and phone hardware will thank you. They have their own style protector that only plugs into their blocks, and they make the style that interchanges with all the other brands. They have a square peg in round hole connection that lets them use tin plate where the 'other' design MUST use gold plate pins. The pennies saved make no difference to local site usage, but for a TELCO where each CO has 10s of thousands of these, it does count. A typical 3 electrode gas tube plug in unit is maybe $5.xx and is pennies more that the 2 electrode ones that unfortunately are often used. The "DELTA" (tm) ones with all of the above good features are about $8.xx. Porta Systems has the obvious catalog type stuff, but also ask for 2 other brochures: "The Line Protection Planning Guide" and "When Lightning Strikes...". Porta Systems : 516-364-9300 (sorry if this Porta Systems data is a rehash of a much earlier note I did somewhere - DEJA-VU is scratching my skull) _____________________________________________________________________________ A rather interesting company (that also sells LEA DynaTech - see .7 or below - as well as their own products) makes products that GUARANTEE NO LIGHTNING STRIKES! to your structure. You don't believe it? Try a customer list that starts with: new FBI building in VA, various NASA, various FAA,... They make custom designed (for your building, control tower, radio station tower, ski lift on solid shale mountain, or whatever) dissipation arrays of potentially THOUSANDS of tiny points that go into corona state and bleed the charge away never letting it get high enough for a bolt. The radio tower top type is about a 12' sphere shaped array of zillions of needles. They view lightning rods as archaic and dangerous devices! that INVITE trouble. A dissipation array can be grounded with #6 wire (actually carries only maybe 1/2 amp in storm), rather than 4/0 for a lightning rod. If you want a 'lightning is near' detector for your explosives plant or refueling depot, they make one. It is basically a small dissipation array with an opto-isolated current detector circuit in the ground lead. But if you do want a VERY low resistance ground (for your lightning rod or whatever), and your soil needs help, they make a 'CHEM ROD' that leaks salts into the soil continuously from tiny holes in a 10' long 2.5" round copper electrode. You top it up yearly with more salts through a cap/plug in the top. And these folks are: LEC Corp (Lightning Elimination & Consultants) 303-447-2828 ask for Peter Carpenter _____________________________________________________________________________ And from .7 : >PolyPhaser Corporation LEA Dynatech Inc. >P O Box 1237 12516 Lakeland Road >Gardnerville NV 89410 Santa Fe Springs CA 90670 >702.782.2511 213.944.0916 Both have 800 #s: 800-325-7170 800-654-8087 _____________________________________________________________________________ I have NO connection to any company mentioned. ================================================================================ Note 262.9 PC Power Controller/Conditioner Equipment 9 of 12 EISNER::KILGALLEN "Larry Kilgallen" 3 lines 6-APR-1989 07:20 -< Ring voltage >- -------------------------------------------------------------------------------- > cross to 120 volts which is less than ringing voltage, so But I thought ring voltage was 90 VAC (albeit 20 Hz)! ================================================================================ Note 262.10 PC Power Controller/Conditioner Equipment 10 of 12 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 17 lines 6-APR-1989 16:43 -< 90VAC biased by at least 48VDC >- -------------------------------------------------------------------------------- It is nominally 90, but, especially in rural areas, may be higher, and at other frequencies that 20hz. The RINGING 90v AC is (biased) on top of the 48 volts DC which may itself on long loop rural range extenders be 72 VDC or even 96 VDC. The ringing generator has one end connected to the ungrounded battery lead, and the other generator end and the grounded battery lead are used to ring you. This DC component is needed to pull in the ring trip relay if you go off hook in ring cycle. Many non Bell TELCOs use either of 2 different 'scales' with the lowest 'note' being 16.x and the highest being 66.x. This gives individual ringing on multi party lines. The tuned ringers only clang at their frequency. The vanilla 'straight line' (non-tuned) ringer that is best at 20hz but is often used at 30hz (keysystem ringing generators are often 30hz as they are much less expensive than 20hz ones) actually needs over 150 volts to ring at 66 hz! ================================================================================ Note 262.11 PC Power Controller/Conditioner Equipment 11 of 12 EISNER::BYRNE_C "Charlie Byrne" 4 lines 6-APR-1989 16:47 -< Topic Police arrive shortly after >- -------------------------------------------------------------------------------- Back to PC Power controllers, please, or start new topic on Telephones. Thanks. Your humble moderator, BYRNE_C. ================================================================================ Note 262.12 PC Power Controller/Conditioner Equipment 12 of 12 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 40 lines 6-APR-1989 18:28 -< clamp as low as possible >- -------------------------------------------------------------------------------- >>>< Note 262.11 by EISNER::BYRNE_C "Charlie Byrne" > >>> -< Topic Police arrive shortly after >- >>> Back to PC Power controllers, please, or start new topic on >>> Telephones. Thanks. >>> Your humble moderator, BYRNE_C. While picking spike protection for your *** P C *** you may have to select clamping voltage levels for various devices. The power strip ones are preselected by being in a 120v power strip, but the ones for your breaker panel YOU get to choose. And you DO want to work back and beef up protection all the way back to the service if possible. If your power strip's protection gets blown away, it is because you are asking it to do MORE that it is intended to do. The heavy duty (but hopefully less frequent) spikes MUST be caught earlier in the distribution network in your building. You local breaker panel is most likely 120/something, and 120 suppressors wired from each hot leg to Neutral will most likely be the best choice. But big buildings may be distributing at higher voltages, and possibly delta connected rather wye connected 3 phase will be found before the local stepdown transformer. One manufacturor lists protection modules with the following optional nominal voltage ratings: 120, 220, 240, 277, 346, 480, 600. The corresponding clamp voltages for 300 amp surges and a specified waveform range from 320 - 1550 volts, and with the 3000 amp spike from 450 - 1800 volts. Keep working back applying protection in stages until you get to the service entrance. If your PC is connect to other wires extending out of your building (e.g. phone wires), those wires need suppressors too. There are many different voltages available. The Delta (tm) units noted in .8 are available in 75, 230, and 350 volt ratings. If it is a leased line modem circuit, there will be NO ringing, and the low 75 volt protectors are best. If there is a normal CO line with ringing the 230 volt class should work, but if your CO uses boosted voltages as noted in .-2, you may need to select 350 volt units.  ================================================================================ Note 263.0 HSC configuration and performance 4 replies EISNER::BRYANT "Geoff Bryant" 32 lines 9-JAN-1989 09:56 -------------------------------------------------------------------------------- We currently have the following configuration of disks on our HSCs. Note that all disks are dual-ported to both HSCs. Requestor HSC 1 HSC 2 --------- ----- ----- 3 0 DUA15: 1 DUA0: 2 DUA10: 3 DUA5: 4 0 DUA4: 1 DUA8: 2 DUA1: 3 DUA2: etc... The column I have listed a disk in refers to the HSC that we normally fault a disk over to. The question we have is: would we get better performance by having 2 disks faulted to each requestor of each HSC, or is it better to have one HSC handle all requesets for a given requestor? In other words, would the following be better? Requestor HSC 1 HSC 2 --------- ----- ----- n 0 DUAnn: 1 DUAnn: 2 DUAnn: 3 DUAnn: ================================================================================ Note 263.1 HSC configuration and performance 1 of 4 EISNER::FULLER_S "From the terminal of Stu Fuller" 10 lines 9-JAN-1989 11:41 -< pick door number 2 >- -------------------------------------------------------------------------------- In general, the 2nd config is better. In the 1st config, you have one channel working, and the other one sitting idle on both HSC's, while in the 2nd config, you have both channels working. There can only be one data transfer going on on a particular disk channel, regardless of the number of drives connected to that channel. Channel == requestor, also. ================================================================================ Note 263.2 HSC configuration and performance 2 of 4 EISNER::BRYANT "Geoff Bryant" 16 lines 11-JAN-1989 08:33 -< A little more help? >- -------------------------------------------------------------------------------- Thanks for the response Stu. Forgive me if I sound dense here, but this configuration was recommended by a consultant, and we rewired things to be this way, so I want to be sure and understand what you are saying so I can resolve an apparent discrepency between what you recommend and what the consultant recommended. As I understand it from you: A requestor in an HSC can have a data transfer going on only one disk on the requestor. Two data transfers can happen however if the two disks are faulted so as to be split between the two HSCs. OK, it's obvious to me that I get twice the data moving. But, do I really get twice the data moving or is there an interaction (coordination) between the two HSCs which adds overhead that detracts from this? ================================================================================ Note 263.3 HSC configuration and performance 3 of 4 EISNER::FULLER_S "From the terminal of Stu Fuller" 27 lines 11-JAN-1989 13:43 -------------------------------------------------------------------------------- The HSCs don't talk to each other in any way. They are aware of each other, since they are CI nodes, but not any more than that. Since you're already wired up, what you can do is to simply manipulate the port buttons (A & B) on the drives to force, say, the odd numbered drives to HSC1, and the even numbered drives to HSC0. This way, you have both HSCs working for you, but they can still back each other up in case of a failure of either. Look at it this way: +--------------+ Drive 0 --| | Drive 1 --| disk |----- HSC internal bus Drive 2 --| requestor | Drive 3 --| | +--------------+ The requestor has 4 inputs from 4 drives, but only one output. It doesn't multiplex between drives during a transfer. Once it starts on a data transfer on a drive, it's locked on that drive until a) it's done, or b) it has to move the heads or select a new head, or c) an error occurs. This is the way that all disk controllers have worked, including MASSbus controllers, UDA50, etc. Stu ================================================================================ Note 263.4 HSC configuration and performance 4 of 4 EISNER::BRYANT "Geoff Bryant" 15 lines 12-JAN-1989 10:42 -< We will try faulting the drives the other way >- -------------------------------------------------------------------------------- Thanks for the additional info. We will definitely change the port A/B selections on the drives to get all of the requestors in both HSCs going. I was given more info about why the drives were re-wired to be arranged differently. It seems that the original wiring was pretty random, so that a given disk may have been on requestor 1 in HSC A and on requestor 2 on HSC B. It was rewired so that each disk is now on the same requestor in each HSC. Your statement about there being no interaction between the HSCs explains why this worked. We will also try to pay attention and separate the busiest disks between HSCs and requestors when possible. Thanks for the help!  ================================================================================ Note 264.0 LN03r : software required? 2 replies EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 12 lines 9-JAN-1989 11:46 -------------------------------------------------------------------------------- We recently ordered an LN03s ("LN03-PLUS") and an LN03r ("ScriptPrinter") for our facility. The LN03s I have managed to get working, although I'm still not sure what I have to do to get SGR 15 running in portrait mode. That may be another topic, or further down this one someday. The big question is, what do you need to use an LN03r besides an LN03r and a host? Our Field Service rep sez there's some kind of software that has to be ordered separately, but he hasn't been able to find anything. Our sales rep sez "Oh, no, it's all there, just plug it in." One of the glossies that came with it mentions a software license that's bundled with it, but no media of any kind came with the printer. Does anybody have one of these benighted contraptions? Does anybody know what we should have ordered with the fool thing? ================================================================================ Note 264.1 LN03r : software required? 1 of 2 EISNER::MAYHEW "Bill Mayhew" 17 lines 9-JAN-1989 12:00 -< VAX ScriptPrinter software >- -------------------------------------------------------------------------------- < Note 264.0 by EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" > > Does anybody know what we should have ordered with the fool thing > [ LN03R ]? Sure. A new sales rep... Only _slightly_ more seriously, you need the ScriptPrinter software. Buying the LN03R gives you a license, but you need the media and doc kit. Page 194 of the Summer/Fall '88 DECdirect-Plus catalogue tells you this (2nd column), but does not give the part numbers. (I don't seem to have a Winter/Spring '89 one... does anybody else?) The part number is QA-VF9AA-H*, *=5 for TK50, *=M for magtape. $536. ================================================================================ Note 264.2 LN03r : software required? 2 of 2 EISNER::HASSINGER "Bob Hassinger" 15 lines 9-JAN-1989 12:21 -< Generally true. Some exceptions though... >- -------------------------------------------------------------------------------- > -< VAX ScriptPrinter software >- In general a true answer, assuming you want to use the LN03R on a VMS system and you want to queue output to it (a normal queue is not going to work very well because, in simple terms, PostScript printers talk back and the normal print symbiont does not deal with that very well) and particularly if you want to output ASCII LN03 style files or 4014 or Regis as well as PostScript. It could be that you only want to do PostScript printing and you are happy with allocating the port and connecting directly from a VAX application or maybe using it from a PC, etc. If you have an application that knows how to talk to a PostScript printer directly then you do have a fair chance of getting it to work without the famous ScriptPrinter software. Your best bet is to order the H kit though.  ================================================================================ Note 265.0 System Industries & KDB50's 10 replies EISNER::ROBERSON 13 lines 9-JAN-1989 12:26 -------------------------------------------------------------------------------- Does anyone know of any problems with the KDB50 and System Industries model SI57 disk drives? We are in a rock throwing contest with S.I. who says that the problem we have is DEC's problem with the KDB50. DEC says that the KDB's are at the current rev level. The problem is that whenever the system (an 8820) goes down for whatever reason the chances are that it will not reboot without having to powerfail the S.I. drives. The error is "unable to initialize device". S.I. says that there are 50 known problems with the KDB50. They also say that they have a 15 page document that describes these. They have talked about it for weeks. I haven't seen it yet. We have these drives hanging on HSC's without this particular problem. Any information would be appreciated. ================================================================================ Note 265.1 System Industries & KDB50's 1 of 10 EISNER::CONROY "Alan Conroy" 14 lines 9-JAN-1989 13:29 -< I'd laugh, but it really isn't funny... >- -------------------------------------------------------------------------------- > model SI57 disk drives? We are in a rock throwing contest with S.I. > who says that the problem we have is DEC's problem with the KDB50. DEC > says that the KDB's are at the current rev level. The problem is that Let me get this straight: you bought a 3rd-party disk that was supposedly compatible with DEC's hardware and now the 3rd-party company is telling you that it doesn't work because of a problem with DEC's hardware? What's wrong with this picture? This is only ONE of the many reasons I steer clear of SI. If you haven't already paid them, I suggest holding off until SI gets it to work. If you have paid them: then I'm sorry, you may be up the proverbial creek. Other than legal action (which would likely be more expensive than the cost if the SI gear) you're probably stuck until SI makes their equipment REALLY compatible. ================================================================================ Note 265.2 System Industries & KDB50's 2 of 10 EISNER::ROBERSON "Frank Roberson" 4 lines 9-JAN-1989 17:10 -< right. not funny >- -------------------------------------------------------------------------------- You're correct. It's not funny. I tried to stay clear of SI also. I was unable to control the purchasing agent. The reason for the SI drives is too sad and long for here. But now that I have them I am looking for some truth. ================================================================================ Note 265.3 System Industries & KDB50's 3 of 10 EISNER::KILGALLEN "Larry Kilgallen" 3 lines 9-JAN-1989 17:35 -< A proposed tactic >- -------------------------------------------------------------------------------- Don't talk directly to SI. Send all communications through your purchasing agent. That won't get your drive fixed, but it might make them listen to you on the next purchase. ================================================================================ Note 265.4 System Industries & KDB50's 4 of 10 EISNER::HASSINGER "Bob Hassinger" 1 line 10-JAN-1989 10:14 -< Or maybe send them through the PA's supervisor ;-) >- -------------------------------------------------------------------------------- ================================================================================ Note 265.5 System Industries & KDB50's 5 of 10 EISNER::RICE "Gary Rice" 8 lines 11-JAN-1989 09:05 -------------------------------------------------------------------------------- We had a similar problem with the SI Eagle drives attached to a UDA50 on a 785. After any aclls and several visits by the (later proved imcompetent) FE, SI did fix the problem. Later, SI was SO anxious to "prove" that this sort of problem was "no longer possible", that we have been able to obtain free disks from them. Gary ================================================================================ Note 265.6 System Industries & KDB50's 6 of 10 EISNER::ROBERSON "Frank Roberson" 7 lines 18-JAN-1989 15:53 -< New(?) FCO for KDB50 >- -------------------------------------------------------------------------------- There is an FCO to the KDB50 (KDB50-R002) dated 14 October 1988. I found out about it from S.I.! The memo states that "This FCO is an ABSOLUTE requirement for all ULTRIX based systems." It further states that "This FCO should be performed on all ALL VAXBI systems using the KDB50 DSA Disk Controller." It addresses 20 known problems with the microcode. Our Field Service office did not know about it until I told them. I now have it on order and will see if it fixes our problems. ================================================================================ Note 265.7 System Industries & KDB50's 7 of 10 EISNER::RENES "Stuart Renes, DFWLug Chair" 11 lines 20-JAN-1989 13:33 -< Didn't work for me! >- -------------------------------------------------------------------------------- I just had that KDB50 FCO applied before Christmas and things got WORSE for me. The new KDB50 board (T1002?) would get host timeout failures and do a re-init cycle (like it does during reboot) at the rate of about 10 per day. I finally had DEC replace the set but what I got was a board set that wasn't UP TO REV again! I'm gonna try to get another new set and see if I just had bad luck with the first! Stuart. ================================================================================ Note 265.8 System Industries & KDB50's 8 of 10 EISNER::BRUCKER "Cyberspace Jockey" 18 lines 24-JAN-1989 17:07 -< vms 5.0 patch for host timeout errors >- -------------------------------------------------------------------------------- >The new KDB50 board (T1002?) would get host >timeout failures and do a re-init cycle (like it does during >reboot) at the rate of about 10 per day. We are having the same problem only on a 3600 with an RA82 and an RA70 drive. Both will go into a mount verify timeout status and then come back online. Looking in the errorlog I found that there are numerous re-inits taking place over the course of the day, and they are more frequent on an unloaded system! After chatting with my FE about the problem, he mentioned that a patch had gone out over DEC's e-net that 'fixed' the problem. I contacted Colorado, and they found a patch related to host timeouts on KDA and KDB50 controllers when running under VMS 5.0-x. I received the tape yesterday and have yet to install it, so I don't know if it will indeed work. - Ken ================================================================================ Note 265.9 System Industries & KDB50's 9 of 10 EISNER::RENES "Stuart Renes, DFWLug Chair" 8 lines 25-JAN-1989 09:31 -< Wonderful! >- -------------------------------------------------------------------------------- Hmmm, that's interesting cauz the problem ONLY occurs with the latest FCO'd KDB50 module set! As I indicated previously, replacing the new KDB50 set with an earlier rev set stopped the problem. Please let us know how the patch works! Stuart. ================================================================================ Note 265.10 System Industries & KDB50's 10 of 10 EISNER::BRUCKER "Cyberspace Jockey" 19 lines 26-JAN-1989 13:22 -< patch to DUDRIVER solves host timeouts >- -------------------------------------------------------------------------------- I have had the patch installed for over 40 hours now, and haven't seen any more than the usual 2 errors on PUA0 that occur during the boot. It seems to have solved the problem for me since I was seeing them on the order of 10-15 a day, usually in the evening and early morning. The patch is to the DUDRIVER. It adjusts an internal timer slightly upwards to account for some system latency when sensing controller timeouts. The system would get a timeout on the KDA50 during idle activity and issue an init request. CSC indicated that the patch also applies to the KDB50 controller. To make it easier for CSC to locate the needed patch, the cover letter I received with the patch indicates that this a fix to the DUDRIVER for spurious host timeouts. I'm pretty sure that this only applies to VMS V5 systems. -Ken  ================================================================================ Note 266.0 NAME THAT DEVICE 25 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 11 lines 10-JAN-1989 00:06 -------------------------------------------------------------------------------- RRD40 TZ30 RX23 RZ22 RZ23 RZ55 ================================================================================ Note 266.1 NAME THAT DEVICE 1 of 25 EISNER::MAYHEW "Bill Mayhew" 21 lines 10-JAN-1989 00:19 -< I can name it in 1200 bytes... >- -------------------------------------------------------------------------------- TZ30: Tape cartridge, probably half-height, probably SCSI, possibly a half-height TK50, and definitely $3500. RZ23: 104Mb SCSI disk, available on DECstation 3100 (PMAX) or VAXstation 3100 (PVAX), and definitely $2400. RZ55: 300Mb SCSI disk. $6500 with expansion box for DECstation 3100; or, $5950 (332Mb, RZ55A) for mounting in a BA213 (!!) And here are a few more: PC400 series: DECstation 200 and 300 family, the Tandy-built PCs. PV* series: VAXstation 3100, the PVAX. PM* series: DECstation 3100, the PMAX (MIPS chip set). And here's one I don't know: QL-001AN-V* - VMS licenses for a MicroVAX II, priced at about 60% of the standard prices. I find this V* suffix on MVII and MV2000 variants; maybe elsewhere too. All this courtesy of a late night visit to the New Electronic Store... ================================================================================ Note 266.2 NAME THAT DEVICE 2 of 25 EISNER::TINKELMAN "Bob Tinkelman - CCA" 6 lines 10-JAN-1989 07:02 -< Locking the barn door after ... >- -------------------------------------------------------------------------------- RE: < Note 266.1 by EISNER::MAYHEW "Bill Mayhew" > > All this courtesy of a late night visit to the New Electronic Store... And I was wondering why the Store greated me this morning with a banner saying "closed until 11AM" ================================================================================ Note 266.3 NAME THAT DEVICE 3 of 25 EISNER::SHANNON "Terry C. Shannon, IDC" 1 line 10-JAN-1989 07:34 -< The Electronic Store Has It NOW! >- -------------------------------------------------------------------------------- Holy Hyperium! The Electronic Store does it *again*! ================================================================================ Note 266.4 NAME THAT DEVICE 4 of 25 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 1 line 10-JAN-1989 11:39 -< And now it's closed until 1:00 PM >- -------------------------------------------------------------------------------- ================================================================================ Note 266.5 NAME THAT DEVICE 5 of 25 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 21 lines 10-JAN-1989 17:16 -< More secrets from The Store ! >- -------------------------------------------------------------------------------- Well the Store was open this afternoon (although it was running so slow that I think it was on an 11/730 !!) and I found some interesting things. If you look up RRD40-* (the new SCSI CDROM reader) you will find that one of the options includes a KRQ50-AA. Could this be a DEC Q-Bus SCSI adapter ??!!?? Now that DEC has adopted SCSI as it's new low-end workstation "bus" can someone try and explain to me the (technical) differences between the DSSI (used on the 3300/3400) and SCSI. Rumor has it that the VS3100 (PVAX) has a async SCSI port (at about 1 Mbyte/sec) and the DS3100 (PMAX) has a **SYNC** (at about 4 Mbyte/sec). Why ? And how can the same drive support async and sync (or are there different version not shown in the Electronic Store) ? Digital Review claims that the RZ22, RZ23 and RZ55 are "off the shelf" non-DEC drives (can't remember the vendor). Does anyone have an idea on what the "off the shelf" price is (ie. before DEC markup). ================================================================================ Note 266.6 NAME THAT DEVICE 6 of 25 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 14 lines 10-JAN-1989 19:42 -< DEC had SCSI problems >- -------------------------------------------------------------------------------- > Rumor has it that the VS3100 (PVAX) has a async SCSI port (at about 1 > Mbyte/sec) and the DS3100 (PMAX) has a **SYNC** (at about 4 > Mbyte/sec). I believe they said 1.25 meg for the ASYNC, and that at 4 they were at the upper limit of the spec for SYNC. We just had a LUG meeting with some live hardware, and live DECies. Apparently all this stuff was delayed at least 6 months because DEC found NO chips on the market that did SCSI 'right' and at full speed. A significant thing to note is that DEC intends you to be able to use what you want on it, and the chip makers now know a lot more (thanks to DEC) about how to make suitable SCSI chips. The drivers they are shipping initially target their hardware, but more generic ones should follow. ================================================================================ Note 266.7 NAME THAT DEVICE 7 of 25 EISNER::MAYHEW "Bill Mayhew" 18 lines 10-JAN-1989 20:48 -< We now return you to the original topic... >- -------------------------------------------------------------------------------- Filling in the gaps: RRD40: SCSI CDROM reader, substantially faster than old RRD50 (500 ms seek) RX23: 3.5" diskette, read-only 720Kb, read/write 1.44Mb RZ22: 52Mb 3.5" SCSI hard disk, 33ms avg access, only available as a paging/swapping disk on VS3100s And, while I'm at it, performance specs on the others: TZ30: 1/2 height, TK50-media-compatible, 95Mb, 62.5Kb/sec transfer, SCSI, delayed availability (June?) RZ23: 104Mb, 3.5", 33ms average access, SCSI 1.25MB/sec transfer RZ55: 332Mb, 5.25" full-height, 24ms average access, SCSI 1.25MB/sec transfer (4MB/sec in DECstation 3100 via sync SCSI) ================================================================================ Note 266.8 NAME THAT DEVICE 8 of 25 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 12 lines 11-JAN-1989 02:31 -< CLOSE BUT NO SCSI >- -------------------------------------------------------------------------------- > Now that DEC has adopted SCSI as it's new low-end workstation "bus" can > someone try and explain to me the (technical) differences between the > DSSI (used on the 3300/3400) and SCSI. From the rumor mill but well placed and not non-disclosure. The DEC DSSI disk used a SCSI bus. The SCSI bus DEC is talking about is *NOT* the industry standard SCSI bus. It is electrically a SCSI bus *BUT* the command packets that go back and forth are a unique DEC set of packets. Meaning third party drive will not talk to this until they figure out how to talk the DEC command packet language. ================================================================================ Note 266.9 NAME THAT DEVICE 9 of 25 EISNER::MCCORMICK_B "Bob McCormick" 29 lines 11-JAN-1989 07:54 -< DSSI / RF30 experience >- -------------------------------------------------------------------------------- > The DEC DSSI disk used a SCSI bus. The SCSI bus DEC is talking about > is *NOT* the industry standard SCSI bus. It is electrically a SCSI bus *BUT* > the command packets that go back and forth are a unique DEC set of packets. > Meaning third party drive will not talk to this until they figure out how to > talk the DEC command packet language. I've got the RF30's here on a 3300 system (two ISE's ... remember, they're not disks!) I was shocked when I got this system ... These storage units may have an interface that looks like scuzzy ... but its significantly enhanced! At the KA640's >>> prompt you can say: SET HOST/DUP/DSSI 1 and you will be connected to ISE 1 ... and you can run programs that are stored IN THE DISK ADAPTOR ... the ISE keeps track of how long its been powered up, and how many power cycles... Each ISE has a node name: so these disks on VMS look like on a cluster! Ie R3XYZY$DIA0: and R3ABDF$DIA1: ... better be prepared to handle the new concepts these present on single CPU systems! The performance is GREAT! This dual disk 3300 blows away a MicroVAX II with RD54's... Bottom line: RF series of disks may be using SCSI like bus protocols to communicate on the DSSI but that's about it ... looks more like HSC configuration than anything else! ================================================================================ Note 266.10 NAME THAT DEVICE 10 of 25 EISNER::MCCORMICK_B "Bob McCormick" 10 lines 11-JAN-1989 07:56 -< Runtime VMS?! >- -------------------------------------------------------------------------------- > QL-001AN-V* - VMS licenses for a MicroVAX II, priced at about 60% > of the standard prices. I find this V* suffix on > MVII and MV2000 variants; maybe elsewhere too. Bill: From fairly reliable rumours ... this may be the VMS runtime license (it was proposed to me that this may occur) But what a runtime license means beats the h*l* out of me! Suspect important concept for VAXstations? ================================================================================ Note 266.11 NAME THAT DEVICE 11 of 25 EISNER::RICE "Gary Rice" 14 lines 11-JAN-1989 09:15 -------------------------------------------------------------------------------- > But what a runtime license means beats the h*l* out of me! > Suspect important concept for VAXstations? A non-educated guess would be: Standard VMS less Linker Debugger Macro-32 Librarian STARLET.MLB etc. Gary ================================================================================ Note 266.12 NAME THAT DEVICE 12 of 25 EISNER::KILGALLEN "Larry Kilgallen" 1 line 11-JAN-1989 09:52 -< Is the media different? >- -------------------------------------------------------------------------------- Are you sure it isn't Desktop VMS? ================================================================================ Note 266.13 NAME THAT DEVICE 13 of 25 EISNER::MAYHEW "Bill Mayhew" 10 lines 11-JAN-1989 10:19 -< Media not an issue; -Vx only affects License >- -------------------------------------------------------------------------------- It's not Desktop-VMS; that has a different order number, i.e. doesn't use 001 as the unique product identifier. (0VV or something? don't remember) I am "hoping" it's a runtime VMS since it begins to make the price of VMS acceptable ($600/user vs $1K) for the small commercial system end user. It is ONLY listed on MV II and MV 2000 though. And, this discussion should move... somewhere. VMS? It's not HARDWARE_HELP, and I'm sorry I started it, sidebar or no... Moderator? ================================================================================ Note 266.14 NAME THAT DEVICE 14 of 25 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 7 lines 11-JAN-1989 11:00 -------------------------------------------------------------------------------- > If you look up RRD40-* (the new SCSI CDROM reader) you will find > that one of the options includes a KRQ50-AA. Could this be a DEC > Q-Bus SCSI adapter ??!!?? ... and I wonder how long it will be until DEC announces the BI to SCSI adapter so that we can finally hang CDROMs off of big machines !! ================================================================================ Note 266.15 NAME THAT DEVICE 15 of 25 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 2 lines 11-JAN-1989 11:22 -< Emulex makes Q-Bus SCSI adapter >- -------------------------------------------------------------------------------- Digital News has an article about a Q-bus SCSI adapter on page 17 of the January 9 issue. ================================================================================ Note 266.16 NAME THAT DEVICE 16 of 25 EISNER::KILGALLEN "Larry Kilgallen" 2 lines 11-JAN-1989 11:50 -< Yes to RRD40 on Q-Bus >- -------------------------------------------------------------------------------- Peter Kaiser indicated to the Boston VAX LUG yesterday that Q-bus support for the RRD40 *was* announced (yesterday). ================================================================================ Note 266.17 NAME THAT DEVICE 17 of 25 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 6 lines 11-JAN-1989 12:11 -------------------------------------------------------------------------------- > Digital Review claims that the RZ22, RZ23 and RZ55 are "off the > shelf" non-DEC drives (can't remember the vendor). From the January 9, 1989 issue of Digital Review: The RZ2x is manufactored by Quantum and the RZ55 by Micropolis. ================================================================================ Note 266.18 NAME THAT DEVICE 18 of 25 EISNER::SHANNON "Terry C. Shannon, IDC" 8 lines 11-JAN-1989 19:33 -< Uneducated Extrapolation >- -------------------------------------------------------------------------------- > Bottom line: RF series of disks may be using SCSI like bus protocols > to communicate on the DSSI but that's about it ... looks more like > HSC configuration than anything else! Sure as you can say Local Area System (or Multiple-Host Processing), the DSSI bus implements VAXcluster SCA and acts as a poor man's (or woman's) CI-bus. DSSI almost certainly lays the groundwork for the next step in distributed processing--mayhap a Local Area VAXcluster Phase III . . . ================================================================================ Note 266.19 NAME THAT DEVICE 19 of 25 EISNER::MCCORMICK_B "Bob McCormick" 8 lines 11-JAN-1989 20:44 -< New product descriptions >- -------------------------------------------------------------------------------- I've posted the complete product descriptions for: o DECstation 3100 o VAXstation 3100 o Watchamacallit 3100S server compute engine in the WORKSTATIONS conference This is where you can get the config info and part numbers for much of what has been discussed in this thread. ================================================================================ Note 266.20 NAME THAT DEVICE 20 of 25 EISNER::KILGALLEN "Larry Kilgallen" 1 line 11-JAN-1989 20:44 -< What do you mean by "next step"? >- -------------------------------------------------------------------------------- They said you can hook two 3400s to the same set of 6 disks. ================================================================================ Note 266.21 NAME THAT DEVICE 21 of 25 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 9 lines 24-JAN-1989 15:01 -------------------------------------------------------------------------------- And for round two... RF71 KFQSA BF213 B215F ================================================================================ Note 266.22 NAME THAT DEVICE 22 of 25 EISNER::MAYHEW "Bill Mayhew" 23 lines 24-JAN-1989 17:52 -< I'll see those, and raise you... >- -------------------------------------------------------------------------------- RF71: Full-height, 5.25", 400Mb DSSI disk (er, Integrated Storage Element) KFQSA: Q-bus DSSI controller BF213: Expansion cabinet for DSSI disks, for use with a DSSI-equipped MicroVAX 3300/3400/3500, includes 12 Q-bus slots (Note: MV 3500 DSSI-based systems also announced) R213F: (Note, first letter is R, not B) DSSI disk expansion cab (holds 3 RF-series disks ONLY, no Q-bus slots) Ans here are some I'll toss out for you... RV64 HSC40 SA550 SA650 TA90 ================================================================================ Note 266.23 NAME THAT DEVICE 23 of 25 EISNER::AZZOLI 16 lines 24-JAN-1989 19:54 -< What I remember from the 1/24/89 Announcement >- -------------------------------------------------------------------------------- RV64 - Optical disk jukebox. DEC purchased the robotics from an outside vendor. HSC40- Low end VAXcluster replacement for HSC-50. Performance of HSC-70, but is limited in number of disks and tapes it supports. Field upgradable to a HSC-70 SA550- Four (I think) 5 1/4 drives in a cabinet with space for you to install RA-81's/RA-82's you may have laying around. SA650- Eight 5 1/4 (400 mb each) drives and two RA-90's TA90 - IBM 3480 compatible tape drive. IBM is manufacturing the drive and DEC enhanced it, or something like that. ================================================================================ Note 266.24 NAME THAT DEVICE 24 of 25 EISNER::LOBMEYER "Fred Lobmeyer City of NY DOT" 7 lines 3-APR-1989 20:59 -< VAX 68xx ??? >- -------------------------------------------------------------------------------- On page 3 of the April 3, 1989 issue of InformationWEEK is an ad for VAXes. The photograph shows a picture of a MicroVAX 2000, MicroVAX 3000 and a VAX 6000 system. Upon closer examination of the 6000 the name plate says "VAX 6860" The ad states "Only Digital's VAX computing solutions let you grow from a $5,000 system to a $10,000,000 system." Any guess as to when this was/will be announced? ================================================================================ Note 266.25 NAME THAT DEVICE 25 of 25 EISNER::SAUNDERS "John Saunders, Pacer Software, Inc" 2 lines 3-APR-1989 21:47 -< Not quite yet >- -------------------------------------------------------------------------------- If this is the same ad as the one on page 31 of Digital News, then it looks to me like "6360".  ================================================================================ Note 267.0 Hard-to-find Parts 15 replies EISNER::KENNEDY "Terry Kennedy" 2 lines 11-JAN-1989 20:32 -------------------------------------------------------------------------------- This topic will be used to post information about hard-to-find part numbers. ================================================================================ Note 267.1 Hard-to-find Parts 1 of 15 EISNER::KENNEDY "Terry Kennedy" 10 lines 11-JAN-1989 20:37 -< Hex-head cabinet latches >- -------------------------------------------------------------------------------- The latches on DEC cabinets are comprised of three parts - the actual hex-head casting (part 12-12972-06, $10.00, PDQ required*), the nylon washer (part 12-12972-03, $5.00, PDQ required), and the clip or retain- ing ring (part 90-08500-00, $0.63, 30 days). *PDQ = Product Demand Quote - The part is not orderable through nor- mal DEC channels. You *MUST* call DEC Unlisted Parts at 603-884-5001 and give them your customer information and the part numbers involved. They will then return a quote which you must complete and return with a Purchase Order in order to place an order. ================================================================================ Note 267.2 Hard-to-find Parts 2 of 15 EISNER::KENNEDY "Terry Kennedy" 7 lines 11-JAN-1989 20:41 -< RK07 batteries >- -------------------------------------------------------------------------------- RK07 disk drives (and other DEC devices) use a 9.6V .250 Amp/hour NiCad battery pack. The part number is 12-10641-01, which is only available by PDQ (see *.1). However, you can create this part with *2* 12-10641-00 parts in series. This part is $27.00, 30 days from DEC. However, it is also available at Radio Shack stores for about 1/4 the price. (4.8V NiCad battery) ================================================================================ Note 267.3 Hard-to-find Parts 3 of 15 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 8 lines 12-JAN-1989 13:18 -< The big drawback to buying small parts from DEC >- -------------------------------------------------------------------------------- >> They will then return a quote which you must complete and return with >> a Purchase Order in order to place an order. With a $50.00 MINIMUM per LINE ITEM!!! You may need only a few of those $1.00 parts, but you will wind up with a LOT of spares. Alan ================================================================================ Note 267.4 Hard-to-find Parts 4 of 15 EISNER::CONROY "Alan Conroy" 5 lines 13-JAN-1989 11:32 -< What to do with your spares... >- -------------------------------------------------------------------------------- >With a $50.00 MINIMUM per LINE ITEM!!! You may need only >a few of those $1.00 parts, but you will wind up with a >LOT of spares. Sell 'em to your friends! ;-) ================================================================================ Note 267.5 Hard-to-find Parts 5 of 15 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 3 lines 13-JAN-1989 12:13 -< The DECUS Library slogan was "Steal from your friends" >- -------------------------------------------------------------------------------- >> Sell 'em to your friends! ;-) In DECUServe SWAP_MEET! :-) ================================================================================ Note 267.6 Hard-to-find Parts 6 of 15 EISNER::MERRIMAN "George Merriman - CCA/NY" 35 lines 7-MAR-1989 17:29 -< keycaps >- -------------------------------------------------------------------------------- The moderators of this conference feel that we don't need two "Where-to-find-it" type notes, so I have copied this note here. This note has been retitled to indicate that notes need not be limited to part number information. <<< EISNER::DUA0:[NOTES$LIBRARY]HARDWARE_HELP.NOTE;1 >>> -< HARDWARE_HELP >- ================================================================================ Note 317.1 Where to find it? 1 of 1 EISNER::MAYHEW "Bill Mayhew" 21 lines 4-MAR-1989 18:25 -< Keycaps >- -------------------------------------------------------------------------------- Prompted by a comment in the DECUS_SYMPOSIA conference... < Note 21.12 by EISNER::HARVEY "Jack Harvey" > -< Practical Giveaways >- > o Keyboard key caps. RECYCLE at the Boston Children's Museum often has bins with hundreds or thousands of LK201 keycaps. Some of these appear to me to have minor printing flaws (i.e. they're "irregulars") though not nearly as irregular as some I've seen on actual LK201s. RECYCLE collects industrial scrap and sells it by the bag (a few bucks for a grocery bag) primarily to teachers and other such folks for use in arts and crafts projects, or what have you. There are similar endeavors at other museums (esp. children's museums) across the country. One of (or is it *the*) Digital's LK201 plants is located nearby, in Roxbury. I have no connection with RECYCLE or the Boston Children's Museum, though I used to work there. ================================================================================ Note 267.7 Hard-to-find Parts 7 of 15 EISNER::KENNEDY "Terry Kennedy" 7 lines 22-MAR-1989 02:44 -< RM02/03/05, TM03 unit select plugs >- -------------------------------------------------------------------------------- The unit-number plugs for RM02/03/05 drives are not the same as the vanilla CDC 9762/66 unit select plugs. The CDC part for the special DEC plugs is 82389500 (for a complete set of 0-7). As of this date, the set costs $13.00 from CDC, or $50.00 from DEC (with a 180-day lead time). The same plugs appear to work in TM03 tape drive formatters (at least, the '0' plug from the CDC set mentioned above makes a TM03 formatter #0). ================================================================================ Note 267.8 Hard-to-find Parts 8 of 15 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 11 lines 22-MAR-1989 05:35 -< hign # blue = low # black >- -------------------------------------------------------------------------------- > The unit-number plugs for RM02/03/05 drives are not the same as the vanilla >CDC 9762/66 unit select plugs. The CDC part for the special DEC plugs is >82389500 (for a complete set of 0-7). The vanilla CDC 9766 plugs have blue letters, the TU16, TM02/3, etc plugs have black letters. There are only 4 bits so there are 16 combinations. If you get a full set of 16 with a 9766, you will find you can match your TU16 buttons, but the labeling will be off. It has been a few years, but I think black 0-7 TU16 ones are blue 8-15, and maybe even backwards. ================================================================================ Note 267.9 Hard-to-find Parts 9 of 15 EISNER::WOOD "Jim Wood (609)423-1040" 9 lines 5-APR-1989 19:54 -< APL keyboard for VAXstations >- -------------------------------------------------------------------------------- The APL keyboard for VAXstations is LK201-RC current cost is 214.00 with a 90-day delivery, according to DECdirect tech support. I imagine the 90-day lead time is the time it takes somebody to paint the silly-little APL characters on the front side of the keycaps. ================================================================================ Note 267.10 Hard-to-find Parts 10 of 15 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 3 lines 9-APR-1989 00:45 -< *Real* APL programmers have memorized the keyboard! >- -------------------------------------------------------------------------------- >I imagine the 90-day lead time is the time it takes somebody to >paint the silly-little APL characters on the front side of the keycaps. ================================================================================ Note 267.11 Hard-to-find Parts 11 of 15 EISNER::BURNET "John Burnet" 9 lines 15-AUG-1989 21:09 -< Tilt/swivel base for VR260 19" monochrome monitor >- -------------------------------------------------------------------------------- Okay, so maybe this one isn't so terribly hard to find, but it's not in the DECdirect catalog, the DECdirect "pre-sales assistance" person I spoke to couldn't give me any assistance (pre-sales or otherwise), and I don't have a price book. Does anyone know what the part number is for the VR260 tilt/swivel base? Do you have the price handy as well? Thanks much. ================================================================================ Note 267.12 Hard-to-find Parts 12 of 15 EISNER::HARVEY "Jack Harvey" 10 lines 16-AUG-1989 03:05 -------------------------------------------------------------------------------- > -< Tilt/swivel base for VR260 19" monochrome monitor >- From "VAX Systems and Options, 1989 April-June", p III.36, Tilt/Swivel VR10X-AA for 19-inch monochrome monitor. From Oct 1, 1988 (sorry, note hour) U.S. Price Book: VR10X-AA VR100/VR260 Tilt/Swivel $295 ================================================================================ Note 267.13 Hard-to-find Parts 13 of 15 EISNER::MAYHEW "Bill Mayhew" 3 lines 16-AUG-1989 13:55 -< VR260 tilt/swivel not compatible with VR262, by the way >- -------------------------------------------------------------------------------- Note for the record that DEC says the VR10X-AA is NOT compatible with the VR262, which is the replacement for the VR260. There is no known tilt/swivel that works with the VR262. ================================================================================ Note 267.14 Hard-to-find Parts 14 of 15 EISNER::MAYHEW "Bill Mayhew" 4 lines 14-SEP-1989 22:47 -< Customized keyboard legend strips? >- -------------------------------------------------------------------------------- Does anybody know of a company that makes LK2xx-compatible customized keyboard legend strips, preferably DEC-like (plastic, with clear spots for the LEDs)? Ideally, at a reasonable price in modest quantities (hundreds, not thousands)... ================================================================================ Note 267.15 Hard-to-find Parts 15 of 15 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 22 lines 16-SEP-1989 15:28 -< Keyboard strips are no problem >- -------------------------------------------------------------------------------- > Does anybody know of a company that makes LK2xx-compatible > customized keyboard legend strips, preferably DEC-like > (plastic, with clear spots for the LEDs)? Ideally, at a > reasonable price in modest quantities (hundreds, not > thousands)... This company does not specialize in LK2xx strips, but makes *lots* of similar products. The strips would be a piece of cake. Their quality is quite good (they make the plastic front panels for Tektronix). They can match any color you like. (The terminal on their color matching system has an LK201, so they can easily check the fit, etc.) Small orders like this are no problem. Just remember that the set-up costs will be a high percentage of the price. Nelson Name Plate Co. 3191 Casitas Avenue Los Angeles, CA 90039 (213) 663-3971 Conflict of interest notice: I am their DEC consultant, so the more you spend with them, they more they can spend with me. :-)  ================================================================================ Note 268.0 What are the differences ? VAXstation 3100 vs 3200 7 replies EISNER::LOBMEYER "Fred Lobmeyer City of NY" 11 lines 12-JAN-1989 14:21 -------------------------------------------------------------------------------- Can someone please explain to me the difference between the Vaxstation 3100 vs 3200? As I see it so far the only difference seems to be that the 3200 may be more expandable. What should a 3200 be used for that a 3100 should not ? The environment which my questions are targeted at is for developement in a graphics display application. Thanks in Advance ! Fred ================================================================================ Note 268.1 What are the differences ? VAXstation 3100 vs 3200 1 of 7 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 10 lines 12-JAN-1989 14:34 -< Does the 3100 have a Q-Bus (or any other?) >- -------------------------------------------------------------------------------- >> Can someone please explain to me the difference between the Vaxstation >> 3100 vs 3200? As I see it so far the only difference seems to be >> that the 3200 may be more expandable. ^^^^ More? Since I saw nothing in the announcement that even implied that the 3100 has a Q-Bus, I suspect that it does not. Alan ================================================================================ Note 268.2 What are the differences ? VAXstation 3100 vs 3200 2 of 7 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 21 lines 12-JAN-1989 14:55 -< Cost ! >- -------------------------------------------------------------------------------- A 3200 is very similar to the VAXstation II/GPX in the BA23 box. In other words it uses a Q-Bus ! However the 3200 has very limited expansion capability becasue of the limited number of slots in the BA23 box (eight) and power supply (230 watts). Typical usage is as follows 1 CVAX CPU 2 MS650-AA 8Mbyte memory 3 reserved for additional MS650-AA 4 VCB02-xx video board (1 plane) 5 reserved for VCB02-xx for 4 plane graphics 6 reserved for VCB02-xx for 8 plane graphics 7 TQK50 and DEQNA 8 RQDX3 The 3100 has no bus. But it does support connection of external device via the SCSI (Small Computer System Interconnect). If you don't need a Q-bus, don't pay for it ! (ie. a typical 3200 will probably be $$$ >> than a 3100) ================================================================================ Note 268.3 What are the differences ? VAXstation 3100 vs 3200 3 of 7 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 3 lines 12-JAN-1989 15:01 -< What about graphics functionality ? >- -------------------------------------------------------------------------------- Do you lose any graphics functionality in a 3100 vs 3200 configuration comparison? ================================================================================ Note 268.4 What are the differences ? VAXstation 3100 vs 3200 4 of 7 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 7 lines 12-JAN-1989 19:00 -< Not exactly >- -------------------------------------------------------------------------------- In the case of systems with color displays, no. The standard for the monochrome system is a single plane VSII like graphics engine. There is an upgrade to make it a gray-scale system and add the graphics accelerator. I cant think of a good reason to buy a 3200 now and in fact im about to cancel a couple of them that I have on order and change to 3100s. ================================================================================ Note 268.5 What are the differences ? VAXstation 3100 vs 3200 5 of 7 EISNER::MAYHEW "Bill Mayhew" 5 lines 12-JAN-1989 23:25 -< It's easy to take a 3100 home with you >- -------------------------------------------------------------------------------- ...but a 3200 is a bit more of a problem, unless you enjoy hefting BA23s around. Now, this is not necessarily an advantage in all environments. If *you* can take it home, so can someone else... ================================================================================ Note 268.6 What are the differences ? VAXstation 3100 vs 3200 6 of 7 EISNER::MCCORMICK_B "Bob McCormick" 7 lines 13-JAN-1989 07:43 -< A 3100 may be a better workstation >- -------------------------------------------------------------------------------- There may be a few things the 3100 does faster than the 3200, as I believe they've implement 32 bit datapath's between various functional units on the multi-function CPU module. True, the VAX is (as in the MicroVAX 2000) doing more of the work for peripherals, experimentation with the 2000 found that it was better at some of the graphics manipulation than the equiv Q-Bus (with QVSS) config. ['cuz you don't got to go over the Q-Bus to dink with bits] ================================================================================ Note 268.7 What are the differences ? VAXstation 3100 vs 3200 7 of 7 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 3 lines 13-JAN-1989 09:19 -< Faster disk channel >- -------------------------------------------------------------------------------- I don't recall the stats but I'm pretty certain that the SCSI disk interface on a 3100 will out perform a RQDX3 (but probably not a KDA50).  ================================================================================ Note 269.0 Un-dual-porting an RP06 4 replies EISNER::KENNEDY "Terry Kennedy" 10 lines 16-JAN-1989 03:09 -------------------------------------------------------------------------------- Here's my entry for bizarre question of the year: How do you convince a dual-ported RP06 that it isn't dual-ported? I've yanked the Port B interface cards, which at least lets it work, but the blasted thing is still returning DP in its Massbus ID register. All of you with 197x Peripherals Handbooks, speak up! Thanks, Terry ================================================================================ Note 269.1 Un-dual-porting an RP06 1 of 4 EISNER::FULLER_S "From the terminal of Stu Fuller" 16 lines 16-JAN-1989 13:12 -< Look it up in the prints! >- -------------------------------------------------------------------------------- On the ECC board (M7776), there are a variety of jumpers. Most of them are used for the drive serial number. W9, however, is for the "dual port" bit (bit 11, drive type register). It is located next to E24 (an 8234 mux). With the board in hand, fingers down, components toward you, E24 is on the 2nd row from the top, 5th chip from the left. Jumper W9 is the jumper (or the empty location for one) closest to pin 9 of the chip. There's a couple of 1k resistors to the right of the jumper location. I believe the jumper is currently out, so you'll need to warm up the soldering iron. Information obtained from the RP06 DCL prints, page EC8 (page 10 of 11) and the module layout (page 1 of 11). Stu (DEC Field Service - there is only 1 'C' in "SERVICE") ================================================================================ Note 269.2 Un-dual-porting an RP06 2 of 4 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 1 line 16-JAN-1989 15:37 -< and how many did you rebuild for EDS ? >- -------------------------------------------------------------------------------- ================================================================================ Note 269.3 Un-dual-porting an RP06 3 of 4 EISNER::FULLER_S "From the terminal of Stu Fuller" 1 line 16-JAN-1989 16:57 -------------------------------------------------------------------------------- Me? Only a couple. And, I didn't do _any_ dual port kits. ================================================================================ Note 269.4 Un-dual-porting an RP06 4 of 4 EISNER::KENNEDY "Terry Kennedy" 16 lines 16-JAN-1989 19:47 -< Thanks, comments on manual pilferage >- -------------------------------------------------------------------------------- First, thank you. Second (to Jack), don't let this guy get away... Last, these drives (along with some CPU's) were donated to us. The donor had DEC deinstall them. DEC told the donor 'you can't give them the prints or manuals', and left with all the documentation. You know and I know that it isn't true, but when DEC tells the donor, what are they to believe? The docs are now filling up some rep's basement, and I'm not about to go out and drop $3,000.00 on printsets and manuals (again, I just did it for the last donated system) So, there are 'good guys' and 'bad guys' in FS - I'm just glad (at least) one of the good guys is on this system...  ================================================================================ Note 270.0 RS232 at high speed / Cable Capacitance / KMV1A 11 replies EISNER::BYRNE_C "Charlie Byrne" 9 lines 16-JAN-1989 22:53 -------------------------------------------------------------------------------- Is there anything that would prevent a device (say a KMV1A) to transmit at 56Kbps with an RS232 output option? I realize that over a distance you want to go to something differential (eg V.35). However is there any reason to believe that an RS232 output running, say 5 feet, to a 232-V.35 converter, will have problems? ================================================================================ Note 270.1 RS232 at high speed / Cable Capacitance / KMV1A 1 of 11 EISNER::KENNEDY "Terry Kennedy" 14 lines 17-JAN-1989 01:19 -< Depends on the hardware - a few things to check >- -------------------------------------------------------------------------------- > Is there anything that would prevent a device (say a KMV1A) > to transmit at 56Kbps with an RS232 output option? Provided the device doesn't 'know' what the output option is and limit you, no. But read on for the answer to 'but does it work?' > I realize that over a distance you want to go to something > differential (eg V.35). There is, as far as I can see in the RS-232 [-C] spec, no rating for any speed greater than 9600. The main limit will be the stray capacitance in the cable, so use low-cap cable. Some (older) devices had capacitors in the RS232 section to limit the slew rate. These may have to be clipped. (Remember the 'Module Modified - 100pf' stickers on DH11's?) ================================================================================ Note 270.2 RS232 at high speed / Cable Capacitance / KMV1A 2 of 11 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 9 lines 17-JAN-1989 09:51 -< Other things to check ! >- -------------------------------------------------------------------------------- Although capacitance of the cable is an issue, it is not the only one. Assuming a perfect cable (remember in college when all elctrical device where perfect), the biggest limitation of RS232 is the "slew rate" of the drivers. In laymens terms this is just how fast you can go from a low to a high state and vice versa. The bottom line ... try it, it might work ! ================================================================================ Note 270.3 RS232 at high speed / Cable Capacitance / KMV1A 3 of 11 EISNER::BYRNE_C "Charlie Byrne" 3 lines 17-JAN-1989 11:25 -< Need PUSART slew rate >- -------------------------------------------------------------------------------- According to the KMV User Guide, the line controller is a "NEC micro-PD 7201 PUSART or equivalent". I assume that the data sheet for this would have the slew rate? Can anyone supply? ================================================================================ Note 270.4 RS232 at high speed / Cable Capacitance / KMV1A 4 of 11 EISNER::BYRNE_C "Charlie Byrne" 4 lines 17-JAN-1989 11:48 -< PUSART has no output options >- -------------------------------------------------------------------------------- But now that I think about it, I bet the PUSART is only the line controller, not the line driver; there must be something after the PUSART that determines if your going 232, 423 or whatever. That's where to look for slew rate. Correct? ================================================================================ Note 270.5 RS232 at high speed / Cable Capacitance / KMV1A 5 of 11 EISNER::BYRNE_C "Charlie Byrne" 18 lines 17-JAN-1989 12:32 -< Now I see importance of capacitance >- -------------------------------------------------------------------------------- Nothing like talking to yourself to help figure things out. Yes, the NEC 7201 (similar to Intel 8274) will typically be connected to a Interface chip (a line driver). Now I'm looking at a Texas Instruments Data Book, and they have a SN75188 quad line driver. "meets RS-232C specs". Transition times: TYP MAX Unit Low to High, Capacitance = 15pf 55 100 ns H to L, 15 pf 45 75 ns L to H, 2500 pf 2.5 microsec H to L, 2500 pf 3.0 microsec So at 56,000 bps we have 112000 transition/sec. If our output capacitance is, say, 2500 picofarads, then we are spending appx 308 ms of every second just changing states. And is it correct to say that this does not include the time that the signal is in steady state? I think I'm getting on the right track here... ================================================================================ Note 270.6 RS232 at high speed / Cable Capacitance / KMV1A 6 of 11 EISNER::BYRNE_C "Charlie Byrne" 11 lines 17-JAN-1989 14:54 -< If you've read this far...thanks! >- -------------------------------------------------------------------------------- Continuing my discussion, I have now learned that the line driver in the KMV is listed only as "26LS29". A local source claims this is an AMD 423 single ended tristate line driver. Can anyone verify? Also, I have looked in Black Box and they have regular cables at about 30 to 50 picofarads/foot. Their low capacitance cable is rated at 12 +-2 pf/foot. Does anyone have a ball park guess for what ribbon cable capacitance is? Horrible I'd imagine. ================================================================================ Note 270.7 RS232 at high speed / Cable Capacitance / KMV1A 7 of 11 EISNER::KENNEDY "Terry Kennedy" 26 lines 17-JAN-1989 19:00 -< AMD specs >- -------------------------------------------------------------------------------- > Continuing my discussion, I have now learned that the line driver > in the KMV is listed only as "26LS29". A local source claims > this is an AMD 423 single ended tristate line driver. Can > anyone verify? Indeed. It's on Page 10-20 of the 1985 AMD Bipolar Microprocessor Logic and Interface Data Book. Relevant extracts are: "Individual rise time control for each output" "High capacitive load drive capability" "Meets all requirements of RS-423" "A slew rate control pin allows the use of an external capacitor to control slew rate..." "Rise time (Rl=450 ohm, Cl=500pF, Cc=50pF) 3.0 uS =0 120 nS Fall time (same)" Rl= resistance of output to ground presented by cable/external device, Cl= Capacitance ... Cc= Slew rate capacitor (which connects to pins 9, 12, 13 & 16) > Does anyone have a ball park guess for what ribbon cable capacitance > is? Horrible I'd imagine. Probably. But worst of all, it is variable... ================================================================================ Note 270.8 RS232 at high speed / Cable Capacitance / KMV1A 8 of 11 EISNER::BYRNE_C "Charlie Byrne" 9 lines 18-JAN-1989 20:56 -< Ribbon cable low capacitance? >- -------------------------------------------------------------------------------- >> Does anyone have a ball park guess for what ribbon cable capacitance >> is? Horrible I'd imagine. > Probably. But worst of all, it is variable... Black Box called me back and their usually excellent staff gave me an answer I find hard to believe. Said capacitance for ribbon was in area of 40 pf/foot. Now their regular 232 cables are about that, and their special low cap cable is appx 14pf/foot. Is this answer reasonable? ================================================================================ Note 270.9 RS232 at high speed / Cable Capacitance / KMV1A 9 of 11 EISNER::HARVEY "Jack Harvey" 18 lines 18-JAN-1989 23:04 -< Ribbon Cable Capacitance >- -------------------------------------------------------------------------------- Roughly speaking, cable capacitance per foot is related to it's characteristic impedance, which in turn is related to the geometry, conductor diameters and spacing. The roughly part is tied to the dielectric (insulation) characteristics. Assuming no really strange plastic is used in the ribbon cable, I would guess ribbon cable would approximate 72 ohm twinlead, and Black Box's number is in the ball park. But if it were tightly clamped with other parallel cables or a metal trough, the capacitance would increase because the transmission line geometry is different. I think this is the variability Terry was referring to. Another point is that 56 kb is only about 5.8 times as fast as 9.6 kb. So your five foot cable ought to have about the same effect at 56 kb as a 29 foot cable of the same material at 9.6 kb. Not very worrisome. But the slew rate limits of the line drivers are a very real concern. ================================================================================ Note 270.10 RS232 at high speed / Cable Capacitance / KMV1A 10 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 13 lines 18-JAN-1989 23:10 -------------------------------------------------------------------------------- > capacitance for ribbon was in area of 40 pf/foot. . . Is this answer > reasonable? Could be. I think it depends on how cap/ft is measured. If it is the capacitance between any single conductor and all of the other conductors, including the shield, then a ribbon cable has an advantage: Only two of the other conductors are physically adjacent to the conductor under test, and these two would tend to shield the rest from the test conductor. In typical round cable all the conductors are twisted around each other so there is more non-test conductor area near the conductor under test. (This is a theory only. Anyone out there know anything about it?) ================================================================================ Note 270.11 RS232 at high speed / Cable Capacitance / KMV1A 11 of 11 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 25 lines 19-JAN-1989 15:52 -< Rise-time should not be a problem. >- -------------------------------------------------------------------------------- The RS-232 spec was originally written before IC's (maybe in vacuum tube days?). Discrete transistor circuits of that era didn't like to see very fast rise and fall times -- which were likely caused by noise instead of signals, anyhow. There was a MINIMUM rise time built into the spec; the signal change had to occupy at least that amount of time. Well, came the millenium (the IC's). The chips couldn't care less about receive rise times; the transmit chips would change states very much faster than the transistor stages, so rapidly in fact that an external pin was provided to slow them down to meet the RS232 rise-time spec. As time went on and more circuits changed to use the chips, more and more designs just left the rise-time shaping capacitor off. OK, the condition now is that most devices that send RS-232 signals have very fast rise times; most receivers will accomodate very slow rise times. It is therefore possible to use very long RS-232 lines with no rise-time-related problems (we have several lines in our building that are >400 feet long, shielded pairs, running 19,200 baud RS232 with zero difficulty). If your drivers do not have the rise-time shaping capacitors, I'd say you have a very good chance for 56KB for 5 feet. Try it and see.  ================================================================================ Note 271.0 3200 BPI TAPE DRIVES / EMULEX & DILOG / CIPHER F880 10 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 14 lines 17-JAN-1989 00:10 -------------------------------------------------------------------------------- I just bought my first Emulex board after a three ban on their products. I have a "friend" who sells hardware and got me to try their QT13 tape coupler. I was planning on using this board with a Cipher F880 dual density tape drive. When I set the Cipher to high density the sub system hangs. I found a note in the Emulex manual that states it will not work with all 3200 BPI tape drives. 1. Does anyone know if this will work? (calling Emulex tommorow) 2. Does anyone know of a coupler that will? (BTW - the old Emulex TC02's do work but are out of production) 3. Why is 1600/3200 different from 1600/6250 ? ================================================================================ Note 271.1 3200 BPI TAPE DRIVES / EMULEX & DILOG / CIPHER F880 1 of 10 EISNER::KENNEDY "Terry Kennedy" 30 lines 17-JAN-1989 01:29 -< F88 PE burst (or lack thereof) is the culprit >- -------------------------------------------------------------------------------- Is this an older F880? If so, the problem is in the drive. Here's the scoop: Cipher designed the F880 before there was a spec for 3200 BPI. 1600 BPI tapes have a 'PE burst' at the front, written when the tape is first initialized. That's how drives distinguish between 800/1600. Cipher drives didn't write (or indicate) a burst when in 3200. This was so they might (in the future) auto-detect 1600/3200. Later, ANSI standardized 3200 as a format and required the burst. Now, your controller is looking for the burst and doesn't find it, so it complains (on RSTS, usually with ??Fatal I/O error'). The fix is quite simple, i*if* you have prints for the controller (of course, Emulex doesn't supply prints). All you do is find the burst flip- flop in the coupler and clip the preset line (used to go to the burst line). No connect the preset to the BOT line. Voila - burst every time you traverse the BOT reflector. Note - there *is* one gotcha. The controller will report burst to the operating system. On VMS, this means VMS will play with the tape for a few minutes trying to figure out why there was a valid burst but no readable data. It's fine on RSTS. Alternatively, you can get Cipher to swap the F880 logic for *only* a few grand. The do not make (as of last time I asked) an ECO/FCO kit for this problem. Of course, if you really wanted to get fancy, the F880 reports density on one of the interface lines. You could always make burst conditional on 3200 being asserted. Needless to say, any of the above will likely void your warranty (if any). ================================================================================ Note 271.2 3200 BPI TAPE DRIVES / EMULEX & DILOG / CIPHER F880 2 of 10 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 2 lines 17-JAN-1989 08:32 -< DRIVE IS 45 DAYS OLD >- -------------------------------------------------------------------------------- > Is this an older F880? If so, the problem is in the drive. Here's the > scoop: ================================================================================ Note 271.3 3200 BPI TAPE DRIVES / EMULEX & DILOG / CIPHER F880 3 of 10 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 14 lines 17-JAN-1989 17:53 -< EMULEX QT13, TC02 AND THE CIPHER F880 >- -------------------------------------------------------------------------------- Well here is the story..... Emulex does *NOT* market any tape coupler that will support the Cipher F880 in 3200 BPI mode. I talked to Cipher and they PEIB (Phase Encode Initial Brust) on 1600 but not on 3200. All other Cipher drives have an option switch but the F8%% series. So if your coupler needs a PEIB at 3200 you are SOL. I view this as a combined Cipher / Emulex Sales Rep. screw up. Cipher for not cleaning up their F880 3200 BPI act now that the standard is out. Emulex sales for listing the QT13 as a replacement for the TC02 without listing the exception. Cipher recommends a DQ120, DQ130, or DQ132. I am now off to take to Dialog. ================================================================================ Note 271.4 3200 BPI TAPE DRIVES / EMULEX & DILOG / CIPHER F880 4 of 10 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 19 lines 17-JAN-1989 20:28 -< Latest Dilog is better choice >- -------------------------------------------------------------------------------- > Cipher recommends a DQ120, DQ130, or DQ132. I am now off to take > to Dialog. Ciphers notes must be a bit old... DQ120 is too old to talk about. DQ130 is TM11 emulation - I doubt you want it. DQ132 is MS11 but may not be supported in the latest VMS release. DQ142 is MS11 and is 'ok' , but is a quad card and costs more than the dual width DQ153. DQ153 is TMSCP, and, like the 132 + 142, supports 4 tape drives daisy chained on the Pertec bus cables. The DQ153 does work on a Cipher F880, but I don't know about 3200 BPI mode. This controller is good for FAST GCR drives. ================================================================================ Note 271.5 3200 BPI TAPE DRIVES / EMULEX & DILOG / CIPHER F880 5 of 10 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 23 lines 17-JAN-1989 23:44 -< Fan of Dilog and Cipher >- -------------------------------------------------------------------------------- >> DQ132 is MS11 but may not be supported in the latest VMS release. I am using a Dilog DQ132 on my VAXstation-II and am quite happy with it. It uses the TSDriver (MS:) under VMS v5.0-2. I have heard of end-of-reel problems under v5.0, but don't know if it is software (I think it is) or hardware. >> DQ142 is MS11 and is 'ok' , but is a quad card and costs more than the >> dual width DQ153. I THOUGHT that the DQ142 was dual-width. The DQ132 is a quad-width board. If I were starting from scratch today, I would go with the DQ153. Besides being a dual-width board, you get all the TMSCP advantages, including BOOTING under VMS! The tape drive is a Cipher 910 (75 ips vacuum-column), which is the nicest vacuum-column drive I have ever used. It is so quiet that I had one operating in my living room for several weeks until I could move it into the computer room. Alan ================================================================================ Note 271.6 3200 BPI TAPE DRIVES / EMULEX & DILOG / CIPHER F880 6 of 10 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 13 lines 18-JAN-1989 00:34 -< DQ132 bad with MV3, DQ142 is QUAD >- -------------------------------------------------------------------------------- > I am using a Dilog DQ132 on my VAXstation-II and am quite > happy with it. It uses the TSDriver (MS:) under VMS v5.0-2. > I have heard of end-of-reel problems under v5.0, but don't > know if it is software (I think it is) or hardware. There IS a problem, but I may not have all the details. I think if you UPGRADE to a MV3 board, you will need to change the DQ132 for sure. > I THOUGHT that the DQ142 was dual-width. The DQ132 is > a quad-width board. You got me worried for a second. I just LOOKED. It is QUAD. One of the two cables exits over the A/B side and the other over the C/D. ================================================================================ Note 271.7 3200 BPI TAPE DRIVES / EMULEX & DILOG / CIPHER F880 7 of 10 EISNER::NAGY "Frank J. Nagy, VAX Guru" 13 lines 18-JAN-1989 07:57 -< DQ132 working on MV-III (Gigastore tapes) >- -------------------------------------------------------------------------------- >> There IS a problem, but I may not have all the details. I think if >> you UPGRADE to a MV3 board, you will need to change the DQ132 for sure. Hmm. That's funny since I'm nearly 100% sure that our Gigastores are handled using a DQ132 controller and we have upgraded our boot nodes to MicroVAX-3200s. The Gigastores are connected to the MV-IIIs and have been running so for about a month with no problems. We also have a DQ153 connecting an STC 2925 (Aviv ????) to one of the boot nodes. This is working very well (including the VMS V4.7 bug of needing PHY_IO to run BACKUP so the cache can be enabled) and surprisingly fast. ================================================================================ Note 271.8 3200 BPI TAPE DRIVES / EMULEX & DILOG / CIPHER F880 8 of 10 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 16 lines 18-JAN-1989 16:39 -< DQ132 + MV3 aren't supported >- -------------------------------------------------------------------------------- > handled using a DQ132 controller and we have upgraded our boot nodes > to MicroVAX-3200s. The Gigastores are connected to the MV-IIIs and > have been running so for about a month with no problems. I just checked with Dilog. The dq132 in NOT supported for MV3s, but they did have the microcode space to fix the Q bus timing on the 142, but you NUST have DQ142 microcode REV C. The DQ153 doesn't have the problem. This does not mean that the DQ132 will not appear to run well, only that there is some timing window and potential liability that can't be fixed on the DQ132 when it is used with a MV3. I was wrong before about the relation to VMS V5, this surfaced around then, but is only Q-bus + MV3 related. ================================================================================ Note 271.9 3200 BPI TAPE DRIVES / EMULEX & DILOG / CIPHER F880 9 of 10 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 6 lines 24-JAN-1989 18:26 -< A DILOG SOLUTION >- -------------------------------------------------------------------------------- Well I talked to Dialog. You can strap their DQ153 so it stops looking for PEIB which will allow Cipher F880 to work at 3200 BPI. It is a chip pin to chip pin jumper. Tech support claims there will be new firmware in the future with an option to stop looking for PEIB. To my knowledge the modified Dialog DQ153 is the only controller in production that will support the F880 in both densities. ================================================================================ Note 271.10 3200 BPI TAPE DRIVES / EMULEX & DILOG / CIPHER F880 10 of 10 EISNER::KENNEDY "Terry Kennedy" 8 lines 24-JAN-1989 20:43 -< Depends on the meaning of 'support' >- -------------------------------------------------------------------------------- > To my knowledge the > modified Dialog DQ153 is the only controller in production that will > support the F880 in both densities. Depends what you mean by 'support'. I've diddled with about 5 other controllers to fix the problem, before I got wise and just figured out how to zap the drive into doing it. That's easier than finding the right spot on each different controller...  ================================================================================ Note 272.0 High count in IO$GL_SCB_INT0 - is this normal? 10 replies EISNER::ANGLEY "Kevin Angley" 30 lines 20-JAN-1989 14:12 -------------------------------------------------------------------------------- En route to solving a CLUEXIT problem on our 8300, Colorado pointed to IO$GL_SCB_INT0 and said, "Ahah! You are having hundreds of thousands of invalid level 0 interrupts. It is servicing these spurious interrupts that causes the system to miss the maintenance timer on the CI750 (or CIBCI). Solve that and you'll solve the problem" Well, we and DEC field service have spent several days trying to find the source of these interrupts. We started eliminating devices one by one and I wrote some procedures to capture what the activity was on the system at the time the count went up significantly. It appeared from that that UNIBUS activity on any device would cause it to grow. We then determined that a DWBUA with an absolutely empty box still manifested these interrupts. Also, a SHOW/UNIBUS under SYSGEN will create 1000 or so interrupts instantly. [Also, a device appears at 772140 which is a mystery to me]. Swapped out the DWBUA - same behavior. Meanwhile, I discovered that a SHOW/UNIBUS on a Nautilus machine (which has a DWBUA) will behave the same way, and we are not having CLUEXIT problems there. Have we been chasing a wild goose? Are these level 0 interrupts "normal noise" or what? Should I not still be concerned about the performance implications of handling some 50,000 of these an hour? If you have a Scorpio (82xx/83xx) or Nautilus VAX (85xx,87xx,88xx), do you have a high count of these interrupts ($ EXAM 80004500 for a V5 system)? ================================================================================ Note 272.1 High count in IO$GL_SCB_INT0 - is this normal? 1 of 10 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 8 lines 20-JAN-1989 17:43 -< Location 80004500 on my 8200 >- -------------------------------------------------------------------------------- > If you have a Scorpio (82xx/83xx) or Nautilus VAX (85xx,87xx,88xx), > do you have a high count of these interrupts ($ EXAM 80004500 for > a V5 system)? I have an 8200 with a DWBUA. Upon first examining location 80004500 the value was 63 H (system uptime 2 + days). After executing SHOW/UNIBUS the value changed to 103E H. ================================================================================ Note 272.2 High count in IO$GL_SCB_INT0 - is this normal? 2 of 10 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 17 lines 21-JAN-1989 00:47 -------------------------------------------------------------------------------- If I'm not mistaken, IO$GL_SCB_INT0 is a count of "passive releases". A passive release occurs when you have 2 (or more) cpus responding to an interrupt. The 1st cpu to get there handles the interrupt, and the others simply don't get an interrupt vector from the device performing the interrupt (since the 1st one already got it). This explains why SHOW/UNIBUS increases the INT0 count. While SYSGEN is scanning the Unibus for addresses, most of the addresses will return SSYN timeout, which causes an interrupt from the DWBUA. This could/would result in an INT0. The reason you had an INT0 count before the SHOW /UNI could be due to SYSGEN attempting to size the Unibus at boot time. I believe that some INT0 are "normal"; but I don't know if 50k/hour could be considered normal . . . Stu ================================================================================ Note 272.3 High count in IO$GL_SCB_INT0 - is this normal? 3 of 10 EISNER::GLEASON "CyberPunk" 24 lines 22-JAN-1989 02:31 -< INT0s >- -------------------------------------------------------------------------------- I found an explanation of the INT0 counter in a paper presented at a symposium a while back - the handout says it was by Paul A. Tourigny of DEC, dated April 27, 1987. Although entitled VAX 86XX Unibus Performance issues, it has a lot of facts about Unibus in general, and several pages on the INT0 counter. As he described it, INT0 is a counter of passive releases. He goes on to explain, that the biggest cause of these is from a (mostly) obsolete feature called "Grant Steal". Grant Steal was used to attempt to get NPRs serviced at a higher priority than BRs. Grant Steal, when enabled on a device, allows that device to monitor the NPR line before passing on a BG to the next device in the chain. If NPR is asserted, it blocks the BG, and issues a false SACK. After 10 micro seconds, a passive release is done to disable the BR cycle, and allow the NPR to proceed, and IO$GL_UBA_INT0 is incremented. This feature is not recommended on VAX Unibuses. To disable Grant Steal on cards that use the DC013 Unibus Control Chip, tie pin 3 of this chip HIGH The recommended maximum number of allowable INT0s is stated in this document to be 100 per day. Of course, since your problem is on a BI VAX, and occurs even when no cards are in the adapter, this may have nothing to do with your situation...in which case, I aplogize for wasting the NOTES space. ================================================================================ Note 272.4 High count in IO$GL_SCB_INT0 - is this normal? 4 of 10 EISNER::KENNEDY "Terry Kennedy" 4 lines 22-JAN-1989 04:09 -< Possible FCO activity >- -------------------------------------------------------------------------------- I seem to recall some FCO activity within recent months on the BI- Unibus adapter. I'm not near the fiche tank at the moment but I'll check it tomorrow. I seem to recall it was to address the passive release problem,,, ================================================================================ Note 272.5 High count in IO$GL_SCB_INT0 - is this normal? 5 of 10 EISNER::ANGLEY "Kevin Angley" 26 lines 22-JAN-1989 13:19 -< I'm convinced I have been chasing a wild goose >- -------------------------------------------------------------------------------- > and occurs even when no cards are in the adapter, this may have nothing > to do with your situation...in which case, I aplogize for wasting the > NOTES space. I don't think there is any such thing as a waste of notes space, especially in the instant case. [But in general, I would say that offering irrelevant information should NEVER be criticized; only failure to contribute by NOT putting up information - ah, but this is the stuff that should be in another conference]. Regardless, the information you give is useful and I believe very relevant. Putting 2 and 2 together (actually .2 + .3 + ...) I think that my SHOW/UNIBUS manifestation of the problem is not the same goings on as the INT0's that are occurring in the normal case. These are probably just the normal passive releases (it IS an SMP system where these occur in normal operation). Also, I will check out the grant steal possibility. In any event, I think this whole issue is a big non-issue w.r.t. the CLUEXIT problem that we were originally trying to solve - Colorado just had me jousting at a windmill. BTW, any thoughts on the device that appears at 772140 when the UNIBUS is empty? Is this perhaps some sort of status registers on the DWBUA? It occurs on both my DWBUA'ed systems. ================================================================================ Note 272.6 High count in IO$GL_SCB_INT0 - is this normal? 6 of 10 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 16 lines 22-JAN-1989 14:16 -------------------------------------------------------------------------------- The DWBUA, as part of its self test, converses with the terminator module on the Unibus. Basically, it tells the terminator to perform some types of transactions in order to test the DWBUA's Unibus handling. As someone else once said, I'm not near my fiche tank, but I suspect that 772140 is the address of the CSR on the terminator that is used to start those transactions. That's why, if you don't have the correct terminator (is it M9313? I forget), the DWBUA will fail its self test. WRT the CLUEXITs, simply missing the maint timer on the CI port should not cause a CLUEXIT. You should see the CI port reset and the system continue. If your system "goes to sleep" greater than RECXINTERVAL (sp?) seconds, then that will cause a CLUEXIT. Stu ================================================================================ Note 272.7 High count in IO$GL_SCB_INT0 - is this normal? 7 of 10 EISNER::KENNEDY "Terry Kennedy" 16 lines 22-JAN-1989 21:01 -< The FCO information >- -------------------------------------------------------------------------------- Here's the FCO info I promised: The DWBUA has been FCO'd (in 1987 and again in '88) for the following problems: o SSYN problem o NPR problem o DMZ32 problem o Passive release on multi-master systems o Crash after power-fail restart The quick check is to look in E67 of the T1010 module - should have either 23-329F4-00 or 23-313F4-00 there. 329F4 fixes all the above, 313F4 fixes all but the last one. Hope this helps... ================================================================================ Note 272.8 High count in IO$GL_SCB_INT0 - is this normal? 8 of 10 EISNER::ANGLEY "Kevin Angley" 6 lines 23-JAN-1989 14:34 -< An update >- -------------------------------------------------------------------------------- Thanks for the info ... the DWBUA is supposedly up to rev at rev "H" but I will check the specifics the next time I'm inside. We were kind of having a mix of CLUEXIT's and CI port maintenance timeouts. DEC is now addressing why the thing "went to sleep" but we've almost decided the UNIBUS is not implicated. ================================================================================ Note 272.9 High count in IO$GL_SCB_INT0 - is this normal? 9 of 10 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 8 lines 2-FEB-1989 00:09 -< The answer appeared before I could ask the question >- -------------------------------------------------------------------------------- >> I suspect that 772140 is the address of the CSR on the terminator I wish I had logged onto the system last week. This was driving me buggy since the device I was working with is nearby and I got them confused. Is this device documented anywhere? Alan ================================================================================ Note 272.10 High count in IO$GL_SCB_INT0 - is this normal? 10 of 10 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 12 lines 2-FEB-1989 14:15 -------------------------------------------------------------------------------- > Is this device documented anywhere? I just checked, and 772140 is the address of the terminator/exerciser CSR. It's documented in the DWBUA technical manual, appendix F. The manual is EP-DWBUA-TM-001 (as printed on the top of my fiche). I think the same device is used on the 11750,and may be documented in the DW750 manual. Stu  ================================================================================ Note 273.0 1/4 in. tape systems/backup performance 7 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 10 lines 21-JAN-1989 21:04 -------------------------------------------------------------------------------- Has anyone tried hooking up 1/4" tape cartridges to Q-bus machines? Whose tape drive did you use? Whose controller did you use? What was the cost? How is the performance? ================================================================================ Note 273.1 1/4 in. tape systems/backup performance 1 of 7 EISNER::KENNEDY "Terry Kennedy" 45 lines 21-JAN-1989 22:48 -< One set of experiences >- -------------------------------------------------------------------------------- > Whose tape drive did you use? I've used the Wangtek series of drives. They come in 60, 120 and 150 Mb. capacity (when using DC600 cartridges). They have interfaces of QIC-02 (high-level 1/4" standard), QIC-36 (low-level interface), SCSI and floppy-style interfaces. > Whose controller did you use? For the QIC-02 drive, I've used the MTI controller. Dilog sells it as the 302. Emulex used to sell it. Frankly, the firmware *stinks*. Even Dilog can't get MTI to fix it. Essentially, once the drive has an error (say, attempting to write on a write-protected tape) you have to cycle power to the controller to get it back. For the SCSI drive, I've used (among others) the Dilog 703. How- ever, I had special firmware. It may be a supported drive now. The drive *might* work to replace a TK500Z, but I'd have to try it. For the floppy-interfaced drive, I've used it on an RQDX3. How- ever, one needs to talk to the RQDX directly, and a special driver would be needed for VMS (or RSTS). > What was the cost? The controllers list from $500 to $1200, depending on vendor and interface. As for drives, I don't know. I run a refurb operation on the as a sideline, so I get 'em for free. Commenting on what I sell refurb units for would be self-serving, so... Anyway, you can call Wangtek (or your choice of vendor) for pricing - Wangtek's in Simi Valley, CA. > How is the performance? Very good. The drives are 90 IPS (but remember they're bit-serial, unlike 9-track tapes.) All controllers mentioned stream these drives at the full 90 IPS under RSTS. The 60 Mb drives are 9-track, 120's are 15-track, and 150's are 18-track. As mentioned previously, 600' cartridges are used. From this, you can determine the effective data rate. Remember, though, that a VAX will not stream a TSV05 in BACKUP, so testing is definitely advised. The 120/150 Mb SCSI Wangtek drives have a 64K cache (with parity) on the drive as well as whatever the controller gives you, so that *does* help. ================================================================================ Note 273.2 1/4 in. tape systems/backup performance 2 of 7 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 14 lines 23-JAN-1989 23:07 -< New BACkup changes rules >- -------------------------------------------------------------------------------- > Remember, though, that a VAX will not stream a TSV05 in BACKUP, > so testing is definitely advised. The 120/150 Mb SCSI Wangtek drives > have a 64K cache (with parity) on the drive as well as whatever the > controller gives you, so that *does* help. This will soon be fixed by the NEW backup, but cached drives (that have to LIE about previous QIO's success and that have to be super careful about end of tape situations) may best let BACkup handle everything, as there will be MANY MANY buffers queued, but when the EOT warning marker is seen all outstanding QIOs will be immediately aborted to preclude running off the end of the tape. I.E. the new BACkup won't need controller cache for performance, but will need honest and timely drive status. ================================================================================ Note 273.3 1/4 in. tape systems/backup performance 3 of 7 EISNER::KENNEDY "Terry Kennedy" 23 lines 24-JAN-1989 01:02 -< Drives are smarter than that! >- -------------------------------------------------------------------------------- > This will soon be fixed by the NEW backup, but cached drives (that > have to LIE about previous QIO's success and that have to be super careful > about end of tape situations) may best let BACkup handle everything... The issue of lying about success is a controller issue, not a drive issue. You can use as little or as much of the cache on the drive as you like. Remember that 1/4" drives have a fixed 512-byte record length, and with VMS BACKUP's preference for large recordsizes (32K), the cache will hold only one full BACKUP record anyway. Thus, EOT is reported for the correct VMS block anyway. Also, the drive guarantees that anything in the drive cache will be written to tape. Just like 9-track tape, current 1/4" cartridges have a bit of slack at the end. If you are thinking about the old (4Mb to 10Mb) capacity drives, things have changed a *lot* in the technology. By yoyoing the tape at BOT, a drive can determine the tape length, density, etc. *without* even having to read any data - it's all encoded in the holes! > I.E. the new BACkup won't need controller cache for performance, but > will need honest and timely drive status. Back up a sec (no pun intended). *New* BACKUP? From DEC? They've seen the error of their ways? When? ================================================================================ Note 273.4 1/4 in. tape systems/backup performance 4 of 7 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 30 lines 24-JAN-1989 02:36 -< new BACkup soon >- -------------------------------------------------------------------------------- > Back up a sec (no pun intended). *New* BACKUP? From DEC? They've seen > the error of their ways? When? There was a session in the BIG arena at Anaheim, and the same guy (who did the changes at home!) gave a session at the local LUG in Camb. MA a month and a half ago. His changes are mostly to the input side, and will gobble as much memory as possible for buffers. It then can spend a LONG time planning strategy, and preloading buffers. It will do a one direction sweep over the disk as few times as possible, filling zillions of buffers with file fragments as they pass under the heads. Full buffers get queued for output. The idea is to not start a streamer until one can readily feed it, and to not rattle the heads all over the place. He told a story of an internal DEC operator that paniced after 15 minutes of nothing happening on the tape drive when he started backup. He ran for his supervisor who eventually decided to kill the job. Just then tape took off, and ran flat out for several reels to the end of the disk. It was a BIG cpu with LOTS of memory. The changes will be phased in, as there has to be the ability of the previous version to read new (especially distribution) tapes, and so not all changes will be turned on at first. He also swears that some ODS-1 support bugs are fixed, and bootability of ODS-1 disks will be maintained. With dual-ported drives, guess what I am going to use instead of BRU to backup an RSX system... ================================================================================ Note 273.5 1/4 in. tape systems/backup performance 5 of 7 EISNER::KENNEDY "Terry Kennedy" 31 lines 24-JAN-1989 03:49 -< Oh, *that* backup >- -------------------------------------------------------------------------------- > His changes are mostly to the input side, and will gobble as much memory > as possible for buffers. It then can spend a LONG time planning strategy, > and preloading buffers. It will do a one direction sweep over the disk > as few times as possible, filling zillions of buffers with file fragments > as they pass under the heads. Full buffers get queued for output. Oh, you mean the 'optimized BACKUP'. I thought this was something truly new. I'm not complaining - BACKUP can use all the help it can get! How- ever, nothing the the current BACKUP architecture precludes it from being a hog - it's just implemented weirdly (not necessarily badly, but *definite- ly* weirdly). On other operating systems that support the BACKUP format, performance is substantially better than that of VMS BACKUP. DEC's RSTS implementation will stream a TSV05 or TU80 (100 IPS) on an 11/23 CPU. My own version for MS/PC-DOS streams a TSV05 drive (really a Cipher F880, same thing) on a 4.77 Mhz original PC, and a user has reported success at 200 IPS on an IBM AT. Perhaps DEC just needs to 'throw this one away' and do it over with the experience they've gained. That would solve the compatibility problems, certainly. Actually, what I wish for in VMS BACKUP is a slight change to BACKUP /PHYSICAL. Rather than copying the whole disk to tape, just dump the allocated blocks. Of course, generate the XOR and CRC blocks as well. To finish it, do a /RECORD pass on the whole disk. No matter what qualifiers I've tried, backups on our 780, even when it is unloaded, generate long pauses on the TU78, as well as only running the TU78 at 1/2 the speed it is capable of. /PHYSICAL can do the whole disk in 1/20th the time that /IMAGE or /FAST or various combinations take. ================================================================================ Note 273.6 1/4 in. tape systems/backup performance 6 of 7 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 0 lines 24-JAN-1989 18:50 -< I ASSUME THAT THE WANGTEK 5150ES IS THE DRIVE? >- -------------------------------------------------------------------------------- ================================================================================ Note 273.7 1/4 in. tape systems/backup performance 7 of 7 EISNER::KENNEDY "Terry Kennedy" 1 line 24-JAN-1989 20:44 -< YES >- -------------------------------------------------------------------------------- > -< I ASSUME THAT THE WANGTEK 5150ES IS THE DRIVE? >-  ================================================================================ Note 274.0 many ethernets to one bus 5 replies EISNER::ROBITAILLE "From the Mac of Mike Robitaille" 7 lines 24-JAN-1989 00:28 -------------------------------------------------------------------------------- I've been trying to find out from DEC if there is a VAX processor platform that will take at least three Ethernet interfaces. According to Dec Technical Pre-Sales folks, there's nothing listed. They referred me to my local office which had sent me to pre-sales to begin with! Does anyone out there have any idea what the smallest Vax platform that will take three Ethernet boards is? ================================================================================ Note 274.1 many ethernets to one bus 1 of 5 EISNER::ROECKEL 15 lines 24-JAN-1989 12:39 -< >- -------------------------------------------------------------------------------- > I've been trying to find out from DEC if there is a VAX processor > platform that will take at least three Ethernet interfaces. WOW! I didn't think it mattered how many Ethernet Cards you put in a machine. We have a system on order from an OEM that is suppose to have up to *four* ethernet cards in an 8810. We will also be placing *three* cards in our existing 8530. I would be VERY interested if someone else with more knowledge on this subject explain the details behind this question !! P.S. We will even be running TCP/IP on one of the cards. ================================================================================ Note 274.2 many ethernets to one bus 2 of 5 EISNER::BYRNE_C "Charlie Byrne" 1 line 24-JAN-1989 19:54 -< For What It's Worth, >- -------------------------------------------------------------------------------- ...I have 2 DELQA's merrily running on a MV3600. ================================================================================ Note 274.3 many ethernets to one bus 3 of 5 EISNER::BYRNE_C "Charlie Byrne" 25 lines 16-FEB-1989 21:03 -< Doesn't exactly answer the question, but anyway... >- -------------------------------------------------------------------------------- Cross posted: ================================================================================ Note 178.14 Peripherals and uVAX/VS-3xxx Compatibility 14 of 14 EISNER::BYRNE_C "Charlie Byrne" 18 lines 25-JUL-1988 13:50 -< Two or MORE DELQA's are *supported* >- -------------------------------------------------------------------------------- > < Note 178.8 by EISNER::BYRNE_C "Charlie Byrne, Freeport NY" > > -< Max number of DELQA's on MV3x00? >- > > Anyone know how many DELQA's one can put on an MV3600. After research, From the Vax-DECNET V5.0 SPD: >(1) When two or more Ethernet controllers are used on a > single CPU, each controller must be connected to a > different Ethernet cable. These cables must not be joined > by a bridge or a repeater, though they may be joined > by another DECnet router. >May 1988 >AE-D624Y-TE This seems to clearly indicate that "two OR MORE" Ethernet controllers are *supported*. As Terry pointed out though, it's another story of if/how well they will work. ================================================================================ Note 274.4 many ethernets to one bus 4 of 5 EISNER::BYRNE_C "Charlie Byrne" 7 lines 17-APR-1989 23:09 -< I can't imagine limit of 1 ether on 8530, but... >- -------------------------------------------------------------------------------- >< Note 274.1 by EISNER::ROECKEL > > -< >- > We will also be placing *three* cards in our existing 8530. The original 8530 announcement said it came with 1 Ethernet interface. I want to add another on a system I am getting. Anyone have 2 NI's on an 85xx? ================================================================================ Note 274.5 many ethernets to one bus 5 of 5 EISNER::ROBITAILLE "MRobitaille@PaxRv-NES.Arpa" 14 lines 10-MAY-1989 23:43 -< DEC's answer (thus far...) >- -------------------------------------------------------------------------------- Well folks, after much hemming and hawing, I've managed to come up with a set of answers. If you are only considering DEC boards, the answer is two EtherBoards per Q-Bus. Likewise, two boards per BI bus. Fortunately, one can purchase a BI machine with more than one BI, hence supporting more than two ethernets that way (and pay through the nose for poorly populated BI buses). However, since one of the three nets I wanted to support was to be exclusively TCP/IP, I have received DEC agreement to use two DEC EtherBoards and one each Excelan board for the TCP/IP side. Curiously, the newer 3300/3400's (I'm getting a quote for a 3300 to use as the H/W platform) have an Ethernet controller right on the CPU aboard. That, in turn had me scurrying to DEC with the obvious question of whether I could have two of the newer QBus EtherBoards (DESQA??) as well. It turns out that you can't. Ah, well....  ================================================================================ Note 275.0 Tape Drive Controllers/TD SYSTEMS/MTI 3 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 6 lines 24-JAN-1989 18:51 -------------------------------------------------------------------------------- Has anyone had any experience with "TD SYSTEMS" controllers? QTA for Q-BUS to QIC QDA for Q-BUD to CDROM ================================================================================ Note 275.1 Tape Drive Controllers/TD SYSTEMS/MTI 1 of 3 EISNER::KENNEDY "Terry Kennedy" 8 lines 24-JAN-1989 20:47 -< I'm interested (but uninformed) >- -------------------------------------------------------------------------------- > QTA for Q-BUS to QIC No, but if it's not based on the MTI one I'm interested! You need to find out from them *what* QIC format - should be QIC-02. Also, what the maximum number of tracks supported is (for 150 Mb, you need 18; for 120 Mb, 15). Also inquire about how they handle EOT - do they put a special length code on the tape the first time it is inserted or do they do it dynamically? ================================================================================ Note 275.2 Tape Drive Controllers/TD SYSTEMS/MTI 2 of 3 EISNER::LANG 7 lines 7-FEB-1989 22:18 -< What's Wrong with MTI? >- -------------------------------------------------------------------------------- Question to .-1, why are you concerned about the TD systems controller being like the MTI one...have you had problems with MTI's tape controller??? Bob Lang ================================================================================ Note 275.3 Tape Drive Controllers/TD SYSTEMS/MTI 3 of 3 EISNER::KENNEDY "Terry Kennedy" 25 lines 8-FEB-1989 03:14 -< MTI - far worse than Emulex >- -------------------------------------------------------------------------------- > Question to .-1, why are you concerned about the TD systems controller > being like the MTI one...have you had problems with MTI's tape > controller??? Many problems. Bear in mind this is the QIC-02 tape controller, not a standard 1/2" coupler. A few of the wonders are: 1) Doesn't support 18 or 26-track drives. Nothing like running a 320 Mb drive as a 125 Mb unit. 2) Doesn't support any tape except 600' ones. Although 150, 300, 450 foot tapes are listed as supported, the controller ignores them (see next item). 3) The controller 'sizes' a tape and writes a header block on it the first time it meets a tape. There is *no need* for this, as the tape type can be determined from the BOT hole spacing, as well as from other methods. A side effect of this is you can't say 'I wonder if this tape has MTI stuff on it' and pop it in the drive - if it wasn't a MTI tape, the sizer will zap the tape. 4) Simple errors, like trying to write a write protected tape or doing a DISMOUNT/UNLOAD kill the controller, and you have to cycle CPU power to get it back. These problems were reported to MTI over a year ago. I was promised a new set of firmware for September 1988. It never came. In fact, they admitted to never even doing the update!  ================================================================================ Note 276.0 Bad disk drive (RA81) 15 replies EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 19 lines 26-JAN-1989 15:02 -------------------------------------------------------------------------------- We are currently having problems with an RA81 disk drive. It looks like logical block replacements, but the number of them, or whatever, sends the system associated with the drive off into never,never land and then the system marks the drive as "write locked". This is OK, but we cannot dismount the disk since it has ALL-IN-1 on it... (By the way, this is a shaddow set, the other disk is OK, but the shaddow set is marked "write locked", and the physical drive in question is dismounted, leaving the other drive mounted.) This is a big headache since we must reboot the cluster to regain the good drive of the pair. The real big pain is that hardware service has come in twice, run ILEXER from the HSC for a couple of hours each time and not been able to make the drive fail. At all. I was wondering if there were any way to get the drive/hsc/system to write and then read every block on the whole disk a couple or three times? If we had bad blocks might this not get rid of them? Seton ================================================================================ Note 276.1 Bad disk drive (RA81) 1 of 15 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 24 lines 26-JAN-1989 16:11 -< Try FORMAT then ILEXER again.... >- -------------------------------------------------------------------------------- If you don't need the disk for a while and you have all the data backed up then..... Have them run FORMAT on the drive. This will go out and at least write to every block on the drive. Also you can have ILEXER write to every block and then do a read compare. You need to let it access the user data area. One of the options as you are going through the ILEXER prompts asks if you want read compare after ever write. In order for the ILEXER to work correctly you must answer "YES" to the initial write question. Pick all the options for display of errors and warnings. Also since ILEXER is a timed diagnostic allow plenty of time for the run or else the initial write will not finish before it starts doing the read/write combo. WARNING: I'm not from field service. I have just played with these diagnostics and found that this usually clears up my problems or at least demonstrates to my local FS that the drive has a problem. I usually present them with the output of the FORMAT, VERIFY and ILEXER runs with all the reporting options turned on and say "Oh my look at all those bad blocks"(actually revectored blocks). This usually prompts them into ordering a new HDA. ================================================================================ Note 276.2 Bad disk drive (RA81) 2 of 15 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 29 lines 27-JAN-1989 15:35 -------------------------------------------------------------------------------- > The real big pain is that hardware service has come in twice, run > ILEXER from the HSC for a couple of hours each time and not been > able to make the drive fail. At all. By default, ILEXER only runs on the diagnostic block LBNs (DBNs). If your bad blocks are in the LBN area, and no in the DBN area, ILEXER will not find them. Personally, when a drive appears to be broke "that bad", I don't bother with the diagnostics, since the error log from the system tells you everything you need to know to troubleshoot the problem. VAXsimPLUS will tell you what's broke, and will provide your DEC FS engineer the information needed. I also have a "home grown" tool for looking at error logs that, among other things, will convert an LBN to a cyl/trk/sec. Based on the information provided by FSTerr (my program) and/or VAXSimPLUS, it's very easy to determine whether the problem lies in the HDA or the electronics. > I was wondering if there > were any way to get the drive/hsc/system to write and then read > every block on the whole disk a couple or three times? If we had > bad blocks might this not get rid of them? Yes, $ INIT/ERASE, followed by SYSGEN> CREATE BIGFILE.JUNK/SIZE=800000, followed by ANA/DISK/READ, and repeat as needed. An inspection of the error log, or a simple SHOW ERROR will reveal any bad blocks. Stu (DEC Field Service) ================================================================================ Note 276.3 Bad disk drive (RA81) 3 of 15 EISNER::BATES "Ken Bates" 23 lines 27-JAN-1989 23:20 -< Try the HSC VERIFY utility >- -------------------------------------------------------------------------------- > The real big pain is that hardware service has come in twice, run > ILEXER from the HSC for a couple of hours each time and not been > able to make the drive fail. At all. I was wondering if there > were any way to get the drive/hsc/system to write and then read > every block on the whole disk a couple or three times? If we had > bad blocks might this not get rid of them? Instead of destroying all your data with a format or INIT, why not just run VERIFY? It's a non-destructive HSC utility which will read every block on the disk. If a block is bad, VERIFY will tell you, and also replace it at the same time. In addition, VERIFY will check the DSA consistency of the device: it will scan the RCT (Replacement Control Table), and compare it with EVERY header on the disk. If a block is revectored incorrectly, or there are inconsistencies in the table, VERIFY will report it. If I recall, it will also check the XBN and DBN area, something VMS has no access to. All in all, VERIFY will do considerably more than either of the other methods, and will also save you from doing a backup and restore. It's fully supported, and should be documented in the HSC manual. If field service is unaware of this utility, you might want to show them the manual... - Ken ================================================================================ Note 276.4 Bad disk drive (RA81) 4 of 15 EISNER::RICE "Gary Rice" 7 lines 28-JAN-1989 10:03 -< Get the Media replaced - PRONTO! >- -------------------------------------------------------------------------------- The key here is the "it becomes write-locked" statement. We suffered through this condition also. FS did their stalling tactic of diags and several visits but ultimately replaced the HDA (twice). You are suffering from the "failed Glue" problem. Gar ================================================================================ Note 276.5 Bad disk drive (RA81) 5 of 15 EISNER::BRUCKER "Cyberspace Jockey" 11 lines 30-JAN-1989 20:07 -< revector table probably full >- -------------------------------------------------------------------------------- We also went through this same situation, and it turned out that there was a large number of blocks that had been re-vectored and the table had filled! If memory serves me correctly (mine, not my CPU's) then there is a way to find out what blocks on the disk have been revectored from the HSC. I don't recall the command, but I do vaguely recall some indication at least as to how many blocks have been revectored. Bottom line...we had the HDA replaced and our problem was solved. - Ken ================================================================================ Note 276.6 Bad disk drive (RA81) 6 of 15 EISNER::BURNET "John Burnet" 10 lines 30-JAN-1989 22:09 -< Full revectored-block table? You've *really* got problems >- -------------------------------------------------------------------------------- > We also went through this same situation, and it turned out that > there was a large number of blocks that had been re-vectored and > the table had filled! That must have been one whacked-out HDA! You've gotta remember that it's the *RA60s* that you use the cleaning brush on, not the RA81s. 8-} I can't remember what percentage of the disk surface is reserved for revectored blocks on an RA8x, but I remember that it's pretty large -- maybe even into the double digits. ================================================================================ Note 276.7 Bad disk drive (RA81) 7 of 15 EISNER::KENNEDY "Terry Kennedy" 12 lines 30-JAN-1989 23:34 -< Yes, a large %age >- -------------------------------------------------------------------------------- > I can't remember what percentage of the disk surface is reserved for revectored > blocks on an RA8x, but I remember that it's pretty large -- maybe even into the > double digits. I recall it's a little bit more than one head's worth, spread around the disk, and also a pool somewhere at the end. This is for RA81's, although a drive is free to implement whatever MSCP-compliant method it chooses. See the MSCP Programming Specification (license-only, big $) for more info. Anyway, failure of one head in an RA81 (so long as there is no contam- ination in the HDA) shoulkd not diminish the drive's rated capacity. Of course, it *does* eat up almost all your replacement blocks. ================================================================================ Note 276.8 Bad disk drive (RA81) 8 of 15 EISNER::BATES "Ken Bates" 15 lines 1-FEB-1989 20:18 -< More than 2 digits worth... >- -------------------------------------------------------------------------------- > I can't remember what percentage of the disk surface is reserved for revectored > blocks on an RA8x, but I remember that it's pretty large -- maybe even into the > double digits. All current RAxx drives reserve one revectored block per track. I don't remember offhand the number for all RA drives, but on an RA82, that's 21,405 individual blocks. If you know the number of tracks per group, groups per cylinder, and total cylinders in host space, it's easy to calculate. I know that VMS will supply the drive geometry information in a GET$something call. As for showing which (or how many) blocks are revectored, I *think* that the HSC utility VERIFY will do that. I may have it confused with another utility, however. - Ken ================================================================================ Note 276.9 Bad disk drive (RA81) 9 of 15 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 2 lines 1-FEB-1989 22:03 -< try: SHO DEV DUxxx:/FU >- -------------------------------------------------------------------------------- > I know > that VMS will supply the drive geometry information in a GET$something call. ================================================================================ Note 276.10 Bad disk drive (RA81) 10 of 15 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 13 lines 2-FEB-1989 14:07 -< Actually, less than 2%! >- -------------------------------------------------------------------------------- The space reserved for spare blocks (RBNs) is less then 2% of the total LBN space available on the disk for the RA81. There are 17528 spare blocks, and 891072 LBNs on the disk; you can figure out the math. The other stuff on the disk (XBN, DBN space) is trivial. So, unless Seton's drive developed some bad electronic failure which would give the appearance of zillions of bad blocks to the controller, it's highly unlikely that the drive ran out of spare blocks. Besides, the error log would record that fact with a special MSLG$W_EVENT code. Stu ================================================================================ Note 276.11 Bad disk drive (RA81) 11 of 15 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 12 lines 3-FEB-1989 08:38 -< Why is the RA81 shadow set left Write Locked ? >- -------------------------------------------------------------------------------- > it... (By the way, this is a shadow set, the other disk is OK, > but the shadow set is marked "write locked", and the physical drive > in question is dismounted, leaving the other drive mounted.) This > is a big headache since we must reboot the cluster to regain the > good drive of the pair. Back to the original question. Why is the volume shadowing software and I guess DSA not just dropping this device as BAD and then letting operation continue on the good drive? Also does the volume Shadowing software get the good data from the valid drive if it revectors a block on the BAD member that it can't restore when trying to read verify the block gone bad or suspected as going bad? ================================================================================ Note 276.12 Bad disk drive (RA81) 12 of 15 EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 15 lines 4-FEB-1989 09:23 -< Problem solved, questions remain >- -------------------------------------------------------------------------------- 1) The problem is solved, at least for this week: a) Field Service replaced the Read/Write logic board in the drive. b) I ran FORMAT and VERIFY on the disk from the HSC, revectored about 76 blocks. c) I ran ILEXER on the drive for 60 hours with no errors, hard or soft. d) Drive placed back into operation. 2) The error log indicated that the HSC did, in fact, replace unreadable blocks on one disk from the other disk. 3) I am still would like to know why the shadowing software didn't simply drop the disk. I would drop it on all but one node, on that one node it would drop it *and* write lock the disk... ================================================================================ Note 276.13 Bad disk drive (RA81) 13 of 15 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 18 lines 4-FEB-1989 13:44 -< More questions >- -------------------------------------------------------------------------------- > 3) I am still would like to know why the shadowing software didn't > simply drop the disk. I would drop it on all but one node, > on that one node it would drop it *and* write lock the disk... > When VMS gets an error while reading the BITMAP.SYS file, it will software write lock the disk to prevent allocation on the volume, since it can't be sure of the state of the bitmap file. When the write lock occured, did the WRITE PROT light up on the drive in question? Or, did SHOW DEVICE say that the drive was write locked due to errors in bitmap? If the WRITE PROT wasn't lit, then it was software write locked. If it was (and the button wasn't depressed), then the HSC write locked it for one reason or another. Stu ================================================================================ Note 276.14 Bad disk drive (RA81) 14 of 15 EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 8 lines 6-FEB-1989 13:06 -< Appears to be bit map? >- -------------------------------------------------------------------------------- Show dev said the device was write locked with some sort of AST pending (I did not copy it all down). The devices was available from other nodes in the cluster. When this sort of write lock occurs can it be cleared without dismounting the drive? Seton ================================================================================ Note 276.15 Bad disk drive (RA81) 15 of 15 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 16 lines 6-FEB-1989 13:52 -------------------------------------------------------------------------------- > < Note 276.14 by EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" > > -< Appears to be bit map? >- > > Show dev said the device was write locked with some sort of AST > pending (I did not copy it all down). The devices was available > from other nodes in the cluster. > AST Pending? New one on me. > When this sort of write lock occurs can it be cleared without > dismounting the drive? > In all the cases that I've seen this problem (drive is software write-locked), it had to be dismounted to clear the write-lock. Stu  ================================================================================ Note 277.0 Long distance printer connections 4 replies EISNER::SKANRON "Closet DEChead" 5 lines 26-JAN-1989 16:48 -------------------------------------------------------------------------------- Is there any way to connect a 2000 LPM (Parallel) printer more than 350 feet from a host VAX, and still expect it to operate at maximum speed. It seems to me that even using a terminal server would not support the 2000 LPM requirement, even if the ethernet was only being used by the terminal server. HELP? ================================================================================ Note 277.1 Long distance printer connections 1 of 4 EISNER::MCCORMICK_B "Bob McCormick" 3 lines 26-JAN-1989 17:26 -< Parrell fiber based long line driver >- -------------------------------------------------------------------------------- I think I saw a recent blurb on a fiber optic based long line driver for parrell printers ... if nobody responds here try DECdirect technical consulting at 800.343.4040 ================================================================================ Note 277.2 Long distance printer connections 2 of 4 EISNER::MAYHEW "Bill Mayhew" 5 lines 26-JAN-1989 17:59 -< LLF01 >- -------------------------------------------------------------------------------- The LLF01 is a long-line, fiber optic-based driver for parallel line printers, up to and including the LP29 (2000 lpm) with "no speed degradation". The printer can be up to 500 meters (~1500 feet) from the host. Introduced October '88; USLP $2175, plus fiber cable. ================================================================================ Note 277.3 Long distance printer connections 3 of 4 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 22 lines 26-JAN-1989 21:56 -< more options >- -------------------------------------------------------------------------------- 2000 line / min is only 4400 characters / sec if the lines are a full 132 chars long. At that speed (44000 bits/sec async) you are just a bit beyond the easily doable 38.4kb. If your print lines on the average are NOT full length, and you have a 38.4 port, consider the EIA async to PARALLEL adapters that are well less than $500. Printing shorter lines, 19.2 would be ok. If you need to go really LONG distance, where real world modems are the limit, there have been printer-specific compression boxes that do // to sync at one end and sync to // at the other, but also do a lot of compression, too. These were around before the currently very succesful and popular Symplex DataMiser type compressor product, that might now be a better choice if coupled with generic EIA sync <-> // adapters. Also, the Emulex LAT box has a // printer port but I don't know what speed it is spec'd at, and there is also a chance that Dec might even take notice of some of the features others have and copy them. ================================================================================ Note 277.4 Long distance printer connections 4 of 4 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 5 lines 27-JAN-1989 14:22 -------------------------------------------------------------------------------- Black Box has a new "parallel" printer extender. I don't remember the specs but it was pretty reasonablely priced. I don't think it did a serial to parallel conversion because it used 4 twisted pairs.  ================================================================================ Note 278.0 Zenith Z-29 Terminal 4 replies EISNER::HARVEY "Jack Harvey" 8 lines 26-JAN-1989 19:02 -------------------------------------------------------------------------------- I need a little help with a Zenith Z-29 dumb CRT terminal I just rescued from the crusher. Actually, it seems to have a pretty good VT102 emulation and is in good shape. Does anyone know how to store (save) the SETUP? Shift S doesn't seem to do it. Also, any way to redefine the function keys? If you have documentation on the beast, I'll be happy to copy it and return. ================================================================================ Note 278.1 Zenith Z-29 Terminal 1 of 4 EISNER::CONROY "Alan Conroy" 5 lines 27-JAN-1989 11:45 -< I won't say it'll work, but... >- -------------------------------------------------------------------------------- > Does anyone know how to store (save) the SETUP? Shift > S doesn't seem to do it. Also, any way to redefine the Have you tried control-S? ^^^^^^^ ================================================================================ Note 278.2 Zenith Z-29 Terminal 2 of 4 EISNER::HARVEY "Jack Harvey" 9 lines 27-JAN-1989 19:45 -------------------------------------------------------------------------------- > Have you tried control-S? > ^^^^^^^ Yep. Many others too. No single character causes a "wait" or "done" message we expect from a VT terminal. I haven't had the patience to do an exhaustive test with a power down/up to see if my setup was saved. Thanks for the thought, however. ================================================================================ Note 278.3 Zenith Z-29 Terminal 3 of 4 EISNER::CONROY "Alan Conroy" 8 lines 30-JAN-1989 12:21 -< It MAY have worked >- -------------------------------------------------------------------------------- > Yep. Many others too. No single character causes a "wait" > or "done" message we expect from a VT terminal. I haven't > had the patience to do an exhaustive test with a power > down/up to see if my setup was saved. You probably should. I've seen several terminals that do not tell you anything, but DO make the changes. It's been so long since I've messed with a Zenith terminal that I don't remember if they were one of these. Good luck. ================================================================================ Note 278.4 Zenith Z-29 Terminal 4 of 4 EISNER::ROECKEL 7 lines 9-FEB-1989 12:48 -< Power Down/Up !! >- -------------------------------------------------------------------------------- > Yep. Many others too. No single character causes a "wait" > or "done" message we expect from a VT terminal. I haven't > had the patience to do an exhaustive test with a power > down/up to see if my setup was saved. We have Z-39's here, and we DO NOT get a message on the screen that says 'Done' like the VT320's do. It probably did work!!  ================================================================================ Note 279.0 8650/DELUA problem/FCO 1 reply EISNER::TACKETT "Galen Tackett" 4 lines 28-JAN-1989 18:27 -------------------------------------------------------------------------------- We have an 8650 with a DELUA. It's a documented problem with this configuration that you can get LAT disconnects if you run Terminal Server Manager. Someone at Colorado CSC told me that the latest FCO of the DELUA would solve this problem. Can anyone confirm this? ================================================================================ Note 279.1 8650/DELUA problem/FCO 1 of 1 EISNER::GRANZOW 15 lines 30-JAN-1989 12:22 -< DELUA FCO >- -------------------------------------------------------------------------------- < Note 279.0 by EISNER::TACKETT "Galen Tackett" > -< 8650/DELUA problem >- < We have an 8650 with a DELUA. It's a documented problem with this > < configuration that you can get LAT disconnects if you run Terminal > < Server Manager. Someone at Colorado CSC told me that the latest > < FCO of the DELUA would solve this problem. Can anyone confirm this?> FCO # DELUA-I002 REPLACES MICROCODE PROMS E31 AND E56 TO P/N 23-451E4-00 AND 23-452E4-00 RESPECTIVLLY. TWO OF THE PROBLEM SYMPTOMS ARE: LAVC CLUSTERS KEEP BREAKING CONNECTIONS DUE TO HOST TIME-OUTS, AND TERMINAL SERVERS DISCONNECT DUE TO HOST TIME-OUTS.  ================================================================================ Note 280.0 DSV11 Information 7 replies EISNER::BYRNE_C "Charlie Byrne" 5 lines 30-JAN-1989 10:33 -------------------------------------------------------------------------------- Does anyone have any info on this board? As far as I know, it will run DDCMP and HDLC. Not sure how many lines, what s/w is available, or much else. I could swear I had some info on it several months ago, but I can not find it now. Part of problem is I think it is CSS product, possibly from Europe. ================================================================================ Note 280.1 DSV11 Information 1 of 7 EISNER::MAYHEW "Bill Mayhew" 7 lines 30-JAN-1989 11:55 -< See Jan-Mar '89 VAX Systems&Options Catalog p. 3.26 >- -------------------------------------------------------------------------------- "The DSV11 is a two-line synchronous interface which provides up to 256 Kbps wide area network capability for stand-alone VAX Q-bus systems. The module supports the DDCMP, HDLC, and SDLC protocols at speeds up to 256 Kb/s for one-line-only operation, and speeds up to 64 Kb/s for two line operation. The Bisync protocol is also supported at speeds up to 9.6 Kb/s per line, regardless of whether one or two lines are in operaton." ================================================================================ Note 280.2 DSV11 Information 2 of 7 EISNER::NADEL "S. Freud" 3 lines 30-JAN-1989 17:27 -< Will this help? >- -------------------------------------------------------------------------------- There's also the DUP11 (unibus model) and DPV11 (Q bus). Both handle DDCMP and HDLC, single line, at 9.6. The DMV11 (Q bus) is faster at 56K. ================================================================================ Note 280.3 DSV11 Information 3 of 7 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 5 lines 30-JAN-1989 18:31 -< DSV - see IND 117 >- -------------------------------------------------------------------------------- The DSV's announcement is buried here in "INDUSTRY_NEWS" 117. I believe it came out of Reading from the same folks that brought us the MicroServer (about 2 years late), and that support the various flavors of P.S.I. ================================================================================ Note 280.4 DSV11 Information 4 of 7 EISNER::MCCORMICK_B "Bob McCormick" 33 lines 31-JAN-1989 06:59 -< DSV11 experiences GOOD! >- -------------------------------------------------------------------------------- I've got whole bunch of the DSV11 beasties. Some of the documentation is produced in Reading England. It has two ports. It needs the WAN_DRIVER PAK loaded on VMS V5.0 to work: this is _included_ with the device (and a TK50 with the driver software) ... The driver is the SJDRIVER(.EXE) and also includes a MICROCODE file SJDRIVER.ULD which magically gets loaded into the DSV11 at AUTOGEN CONFIG/ALL time. If you don't have VMS V5.0-2 or -2a upgrade prior to trying to use the DSV. Prior versions only autoconfigure properly with a real pain in the &@!# workaround (involving the SYSGEN parameter WINDOW_SYSTEM). We're using it for DECnet DDCMP ... so in NCP its called DSV-0-0 and DSV-0-1, as you may expect. It comes with two short cables that attach to the module and provide either X.24 (db25) or X.35 (that weird connector) outputs. They also provide a M-F adaptor that you should use if you have an EIA-232-D device (versus X.24) due to conflicts in pinouts of the two specs (BEWARE!). They also have a cab kit for the BA23/BA123 systems. This board costs much more than the DMV11 (which you may consider for MicroVAX II systems). The DSV11 is not protocol limited as the DMV11 is, so if you have future views of other protocols I'd strongly suggest the DSV11. The DSV11 is one of the better engineered boards I've seen out of DEC. If I remember correctly its an 8 or 10 layer board, complete with a Motorola 68000 class CPU and a whole host of other neat functionality. We've only used a single port on them: haven't had the opportunity to run the second port. If anyone needs additional information I'll be glad to post it (short of typing in the entire DOC set that comes with it!) ================================================================================ Note 280.5 DSV11 Information 5 of 7 EISNER::BYRNE_C "Charlie Byrne" 19 lines 1-FEB-1989 11:04 -< How 'bout "VAX DSV11 Syncronous Device Driver" >- -------------------------------------------------------------------------------- My requirement is a follows: 56/64Kbps HDLC data link DMA output, single line, for an MV 3600, BROADCAST! I have been trying the KMV, but recently discovered, much to my (and my bosses) dismay, that the firmware that one can get for this (KMV1A Driver and X.25 Link Level Software) is supported only to 19.2 KBps although the KMV itslef can run to 64K. The stuff indeed starts flaking out as you start externally clocking at fater than 9600. Apparently the microcode is too dumb too realize, for my application, all it has to do is generate CRC and Flags (which is actually done in hardware anyway)! Now as far as needing VMS V5.0 aand thyis WAN stuff for the DSV, I have before me an SPD for "VAX DSV11 Syncronous Device Driver" Sept 1988 28.49.01 It looks like this fits the bill. Thanks to all so far, additional comments? ================================================================================ Note 280.6 DSV11 Information 6 of 7 EISNER::MCCORMICK_B "Bob McCormick" 20 lines 2-FEB-1989 08:41 -< More DSV11 info >- -------------------------------------------------------------------------------- The DSV11 Synchronous Device Driver Manual's DEC P/N is AA-JG79B-TE Errors in the Documentation: SJ$STARTUP.COM doesn't exist, this is really done by magic - refer to note on DSIN regarding the change. Other notes: Extended Initialisation (note Kings English) block requires two pages (1.0K bytes) of memory - physically contiguous and available when the first DSV11 device is connected during a DSV11 driver program ... use SYSGEN parameter SPTREQ ... two Q22 bus map registers are also required. DSV11 driver wants DSR and DCD from modem (or other interface). DSV11 driver requires that CTS is dropped by modem if DSV11 driver drops RTS. There are also a number of other things to watch for with modem control signals if you use the DSV11 for something other than just with DDCMP/DECnet WAN support. The DSV11 was brought to you by the Wide Area Communications people from DEC Networks and Communications Europe, P O Box 121, Reading RG2 0TU UK ================================================================================ Note 280.7 DSV11 Information 7 of 7 EISNER::ARNOLD "Steve Arnold, Joiner Associates" 16 lines 18-FEB-1989 21:33 -< DSV11 earns "four stars" >- -------------------------------------------------------------------------------- > The DSV11 was brought to you by the Wide Area Communications people > from DEC Networks and Communications Europe, P O Box 121, Reading > RG2 0TU UK We are pressing DSV11s to the limit with P.S.I., SNA, and BSC. I think the DSV is an excellent product, and the development group in Reading has been very responsive to our special needs. I recommend the DSV without reservation. See the Wide Area Networking Device Drivers Specifications for useful tables documenting most of Digital's synchronous devices with supported protocols, electrical interfaces (e.g., V.35, RS232), maximum speeds, and numbers of lines. Is there a specific question you have or application you're considering?  ================================================================================ Note 281.0 Emulex QD33 upgrade for VMS V5 No replies EISNER::NORTON "Bill Norton" 7 lines 30-JAN-1989 11:59 -------------------------------------------------------------------------------- We have an Emulex QD33 controlling a CDC 9772 that has worked fine for about 2 years. As soon as we upgraded to VMS V5.0-2 it started logging errors. Emulex says the QD33 needs firmware rev 'E' or later to work with VMS V5.  ================================================================================ Note 282.0 Talaris (QMS) T2400 firmware problem (tray switching) No replies EISNER::TILLMAN "Brian Tillman, Smiths Industries." 5 lines 30-JAN-1989 16:04 -------------------------------------------------------------------------------- There is a problem with Talaris (QMS) QUIC firmware v1.39 that causes T2400 laser printers to loose track of what they're doing when they need to switch paper trays. This problem is supposedly fixed in v1.41 of the firmware, which we received free when we reported this problem to Talaris.  ================================================================================ Note 283.0 Rack Mountable VT200/300 1 reply EISNER::GLADSTONE "Gary Gladstone (201)943-7100" 2 lines 30-JAN-1989 18:53 -------------------------------------------------------------------------------- Does anyone know of or have experience with a rack mountable VT200/300 terminal? Any help/suggestions whould be greatly appreciated. ================================================================================ Note 283.1 Rack Mountable VT200/300 1 of 1 EISNER::MAYHEW "Bill Mayhew" 4 lines 30-JAN-1989 23:39 -< Digital Has It Now >- -------------------------------------------------------------------------------- Just answered this for you over in some other system (don't even remember which, at this point), but DEC makes "industrial" VT3xx terminals with a rack-mount option. See the "industrial products" section of the Jan-Mar '89 VAX Systems and Options Catalog.  ================================================================================ Note 284.0 Do you need both DSSI/KFQSA and SCSI/KRQ50? 4 replies EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 16 lines 1-FEB-1989 16:13 -------------------------------------------------------------------------------- It seems that Digital has recently "announced" support for two new "buses". I am of course refering to DSSI and SCSI. They have also "announced" two new (and DIFFERENT) controllers for them. The KFQSA for DSSI and the KRQ50 for SCSI. (Neither are available as seperate items, but both come in upgrade kits; the KFQSA with the RF71 and the KRQ50 with RRD40). It also SEEM OBVIOUS that you can not mix and match devices (ie. RZxx devices can not be mixed with RFxx devices). But what about the CDROM reader (RRD40) ?!!! If I decide that I want to junk the RQDX3 on my VS II/GPX and go with RF71's or I buy a 33/3400 do I have to buy a KRQ50 so I can read my CDROM ? ================================================================================ Note 284.1 Do you need both DSSI/KFQSA and SCSI/KRQ50? 1 of 4 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 0 lines 1-FEB-1989 19:17 -< IF YOU WANT RF & RRD DISKS YOU NEED *TWO* CONTROLLERS >- -------------------------------------------------------------------------------- ================================================================================ Note 284.2 Do you need both DSSI/KFQSA and SCSI/KRQ50? 2 of 4 EISNER::MCCORMICK_B "Bob McCormick" 5 lines 2-FEB-1989 08:43 -< DSSI disks only external >- -------------------------------------------------------------------------------- And you can't put your (neet) RF71 disks in your BA23/BA123 system... They come in the BA200 series boxes... I have been unable to find what size pannel ("A" or "B") the KFQSA needs to cable the DSSI out of the box. Anyone know? ================================================================================ Note 284.3 Do you need both DSSI/KFQSA and SCSI/KRQ50? 3 of 4 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 5 lines 2-FEB-1989 09:40 -< BACK TO TWO BOX SYSTEMS >- -------------------------------------------------------------------------------- > And you can't put your (neet) RF71 disks in your BA23/BA123 system... > They come in the BA200 series boxes... Yes - DEC makes a BA200 style box you place next to your BA{1}23 box ================================================================================ Note 284.4 Do you need both DSSI/KFQSA and SCSI/KRQ50? 4 of 4 EISNER::KENNEDY "Terry Kennedy" 3 lines 2-FEB-1989 17:17 -< The tall skinny one - it's a 50-pin like RQDXE >- -------------------------------------------------------------------------------- > I have been unable to find what size pannel ("A" or "B") the KFQSA > needs to cable the DSSI out of the box. Anyone know?  ================================================================================ Note 285.0 LN03s spacing problem 17 replies EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 21 lines 2-FEB-1989 11:46 -------------------------------------------------------------------------------- More LN03 problems : We have an LN03s (AKA LN03 Plus). When you print a job with /COPIES =n, n >1, the top line of the first page of copies 2-n is shifted 16 spaces right : NOW IS THE TIME FOR ALL GOOD VAXEN it should be TO BOOT UP ON THE CLUSTER like this NOW IS THE TIME FOR ALL GOOD VAXEN but it looks TO BOOT UP ON THE CLUSTER like this The flag page of all jobs shows the same effect. This is in Portrait mode, after DECSTR reset; landscape-mode shows the same effect, but only a 2-space shift. This has been seen at another site in the company, so it isn't the printer. Anybody want to take a whack? ================================================================================ Note 285.1 LN03s spacing problem 1 of 17 EISNER::HAHN "DECUServe Vice Chair" 13 lines 2-FEB-1989 12:16 -< Try SEARCH >- -------------------------------------------------------------------------------- Have you checked the other conferences ? I have defined 2 search capabilities in my login $ st :== "SEARCH DECUSERVE_TITLES" $ sat :== "SEARCH decuserve_extracts:all_titles.txt" then: st ln03, or sat ln03 will elicit lots of titles with conference name and note numbers. you may want to use lno3 as some of us do make errors. ================================================================================ Note 285.2 LN03s spacing problem 2 of 17 EISNER::WICHMAN "Jon Wichman" 27 lines 2-FEB-1989 12:43 -< Extra spaces following SetUp string >- -------------------------------------------------------------------------------- >When you print a job with /COPIES =n, n >1, the top line of the >first page of copies 2-n is shifted 16 spaces right : >The flag page of all jobs shows the same effect. This is in >Portrait mode, after DECSTR reset; landscape-mode shows the same >effect, but only a 2-space shift. >This has been seen at another site in the company, so it isn't >the printer. I noticed this too a couple of years ago when we got our first LN03's. I've also noticed it on a LA100. My theory is that the spacing is caused by the SETUP string. Apparently, the print symbiont (or printer) counts the set-up string as characters that it has printed and thus, having printed 16 characters already, it starts printing the real data 16 spaces over. The "solution"? Put 16 ^H (backspace) characters at the end of the setup string which will re-position the (logical) printhead position back to column 1. Note: This was a long time ago, but I _think_ that's what I did... Also, if you use a flag page, the extra spacing will appear on that rather than your printed file. ================================================================================ Note 285.3 LN03s spacing problem 3 of 17 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 2 lines 2-FEB-1989 15:51 -< Use control-string/string-terminate >- -------------------------------------------------------------------------------- You also should see improvement if you surround your setup string with ================================================================================ Note 285.4 LN03s spacing problem 4 of 17 EISNER::SCOPELLITI "Whatsa behind is uva no importanc" 1 line 2-FEB-1989 23:54 -< Easier way? >- -------------------------------------------------------------------------------- Also, try instead of the ^H characters. ================================================================================ Note 285.5 LN03s spacing problem 5 of 17 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 41 lines 3-FEB-1989 17:30 -< insert CRs to fix. >- -------------------------------------------------------------------------------- > My theory is that the spacing is caused by the SETUP string. > Apparently, the print symbiont (or printer) counts the set-up string as > characters that it has printed and thus, having printed 16 characters > already, it starts printing the real data 16 spaces over. Close, but not quite. We saw the same thing. The cause: On the LN03 (and LA100), a form feed doesn't reset the horizontal position to zero. If you've told the print symbiont to generate a left margin for you, it does this for EVERY line you print, even lines containing a lone formfeed. So the margin spaces get inserted, and you get a form feed, but the horizontal position stays where it was, just as it would on a hardcopy printer. The symbiont then inserts another set of margin spaces for the first line of the next page. Our cure was to put a at the end of each form setup module used with the printer, and to also define a page setup module that you specify for all forms used with the LN03, said module containing a lone . If other printers share the same form definitions, their device control libraries of course have to have a like-named module, which can be empty. btw, we found it was a great help to put all of our DEFINE/FORM commands in a command procedure. Another command procedure contains all the commands necessary to create the device control library for each printer, e.g. $ libr/text sys$library:ln03ctl sys$input:/module=landscape 15m ... $ libr/text sys$library:ln03ctl sys$input:/module=portrait ... This way you can include DCL comments to explain the obscure command strings to the next poor person who has to maintain this stuff. It's all too easy to fix a problem by typing a DEFINE/FORM command interactively, and then go crazy trying to remember what you did when the folks who run the VAX down the hall have the same problem! Same thing for queue setups. Sure, the system remembers them, but will you? ================================================================================ Note 285.6 LN03s spacing problem 6 of 17 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 5 lines 6-FEB-1989 13:01 -< Symbiont doesn't recognize >- -------------------------------------------------------------------------------- Also keep in mind that the print symbiont counts as a character but not [. I found this out by experimentation. If you use the [ method, the character count won't be incremented by the symbiont so all the previous suggestions (including mine, perhaps) won't be necessary. ================================================================================ Note 285.7 LN03s spacing problem 7 of 17 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 12 lines 13-JUN-1989 06:21 -< How it worked out >- -------------------------------------------------------------------------------- Finally got back to report on progress : Using the BACKSPACEs or the worked OK. Replacing with [ still left me with a four-space shift. The did not seem to do anything. All things considered, the seems to be the best solution; I don't have to count spaces, and it's a simple fix to publish for the user community. Thank you, one and all. ================================================================================ Note 285.8 LN03s spacing problem 8 of 17 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 15 lines 23-JUN-1989 06:50 -< Now it's lines per page... >- -------------------------------------------------------------------------------- Spacing in another direction ... I can get 80 columns on a Portrait page or 132 columns on Landscape easily enough with the DECSLRM (set left-right margins) command. However, I can't get more than 59 lines per page, in Portrait or Landscape. I've tried using DECSLPP (set lines per physical page) and DECSTBM (set top-bottom margins); no combination seems to get more than 59 lines. I can move them up and down on the page, but that's not what I'm after. I'm starting to think we should just live with our LN01 and forget dreams of using new technology :-} john r ================================================================================ Note 285.9 LN03s spacing problem 9 of 17 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 10 lines 23-JUN-1989 10:22 -< Specify spacing rather than lines >- -------------------------------------------------------------------------------- Try altering the vertical spacing by using "Spacing pitch increment" CSI Pn1 ; Pn2 SP G Pn1 = vertical spacing increment Pn2 = horizontal spacing increment SP = a space character Note the spacing is in either decipoint or pixel units (selected by SSU sequence). Ref: LN03 Programmer Reference Manual page 173. ================================================================================ Note 285.10 LN03s spacing problem 10 of 17 EISNER::HAHN "DECUServe Vice Chair" 3 lines 23-JUN-1989 19:33 -< Top and Bottom Margins ? >- -------------------------------------------------------------------------------- Your problem may be with you top and bottom margin settings DECSTBM. The programmer ref Manual (EK-oln03-rm-002) pages 87... should give you the proper code. If you are in doubt send MAIL. ================================================================================ Note 285.11 LN03s spacing problem 11 of 17 EISNER::COY "Dale E. Coy" 7 lines 24-JUN-1989 00:14 -< Take heart - it can be done >- -------------------------------------------------------------------------------- > ...132 columns on Landscape If memory serves (can't access the code at the moment), you can get about 63 useful lines in Landscape mode, with about 168 columns. Example code on a VAX SIG tape under my name. Programs 2CL (2-column lister) and MCL. ================================================================================ Note 285.12 LN03s spacing problem 12 of 17 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 8 lines 24-JUN-1989 04:50 -< we have it now! >- -------------------------------------------------------------------------------- We get 132 col/66 lines in landscape mode, WITH adequate top margin for three-hole-punched paper, and adequate bottom margin too. All it takes is fiddling with the vertical spacing increment. On Monday I will upload a human-readable interpretation of our LN03 device control library, unless someone in the meantime decides that this would constitute "code". (Moderator?) ================================================================================ Note 285.13 LN03s spacing problem 13 of 17 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 19 lines 27-JUN-1989 11:26 -< Code or not, here it comes... >- -------------------------------------------------------------------------------- This is what I've been using for portrait mode setups : DECSTR (reset) PFS (page format select of 0 SP J for portrait, normal paper) SGR (SGR 10m for built-in-font 1) PUM (position unit mode : character) DECSHORP (horizontal pitch - use font default) DECSVERP (vertical pitch - use font default) DECSPM (pitch mode : font default) DECSTBM (set top/bottom margins to 4, 70) DECSLRM (set left/right margins to 1,80) SPI (spacing increment : font default) DESCLPP (set lines per physical page : doesn't matter what I set, 60, 66 or 70, I get 59 lines printed maximum). For landscape, I use PFS for landscape, SGR 15m, DECSLRM of 1,132. I still only get 59 lines printed on the page. john r ================================================================================ Note 285.14 LN03s spacing problem 14 of 17 EISNER::MAYHEW "Bill Mayhew" 89 lines 27-JUN-1989 13:38 -< Analysis of command interactions >- -------------------------------------------------------------------------------- OK... let's analyze this from the top. All references are to the LN03 Programmer's Reference Manual, EK-OLN03-RM-002 (NOTE -002; the old -001 book, which stopped shipping as of ~2 years ago I think, is dreadful...) > DECSTR (reset) This resets the printer to "initial operating values", which include a top margin of 0.00 inches and a bottom margin of 10.56 inches. Note that these are measured from the left edge of the printable area (0.25" from the top-left corner). (Page 123.) It also sets the lines-per-page attribute, i.e. the form length, to 11.00 inches, and the origin placement (i.e. the point of reference for all measurements) to the corner of the printable area (0.25" from the top-left corner, again). (Page 124.) Vertical spacing is set to 6.25 lines/inch. (Page 125.) > PFS (page format select of 0 SP J for portrait, normal paper) I think this is one error. This sets the "page end" position -- the point at which an implicit form-feed is introduced to move to the next page -- at the 9.17-inch mark, which allows only 57 lines of text. I think you want ?20 SP J, which gives you a full 66 lines... and which is also the setting that the Reset sequence sets up, implicitly... so you don't really need to do this at *all*... unless you are trying to do something different than my understanding. (Page 90-91.) > SGR (SGR 10m for built-in-font 1) > PUM (position unit mode : character) > DECSHORP (horizontal pitch - use font default) > DECSVERP (vertical pitch - use font default) The font default for the DBULTN1 font is 6.25 lines per inch, which is fine. > DECSPM (pitch mode : font default) (Note: should be DECPSM, right?) > DECSTBM (set top/bottom margins to 4, 70) Here's another problem point... but I'm not sure it matters, considering you're issuing the DECSLPP below. This sequence sets the top and bottom margins and the "page home line", which is the place where a form-feed puts you on the next page. (Page 98.) The settings you have chosen cause the top margin to be set to line 4 -- which at 6.25 lines/inch is 0.64" -- plus the 1/4" unprintable area (remember, the reset sequence sets all these measurements up to be relative to the printable area, not to the overall paper size) -- which puts you at 0.89" from the top of the paper. The bottom margin is off the end of the world; there aren't 70 lines available, so it is effectively set at the bottom of the form... the 10.5" point in the printable area, or 10.75" from the top of the page. 10.75" less 0.89" leaves you with a net text area of 9.86", which at 6.25 lines/inch is 61.625 lines per page. But factor into this arithmetic the fact that a "line" occupies space on the page... so is the *top* of line 4 0.89" from the page-edge, or the bottom, or the middle? Beats me, which is why I always do these things in pixel resolution instead of in character-cell resolution. > DECSLRM (set left/right margins to 1,80) > SPI (spacing increment : font default) OK. > DESCLPP (set lines per physical page : doesn't matter what I set, 60, 66 or 70, I get 59 lines printed maximum). DECSLPP defines the form length and un-does a fair amount of what you did with DECSTBM. If it's issued properly, it sets the top margin to 1 and the bottom margin to the form length (i.e. the parameter of DECSLPP). The manual does not make it totally clear how this interacts with the Page Format Select you did back at the beginning. The bottom line here is that while I can't quite come up with 59 lines as a result -- though ignoring the DECSLPP gets me pretty close! -- it seems you are doing a bunch of "extra work" and issuing sequences that countermand earlier ones. I hope this semi-guided tour helps at least a little. I would recommend switching to pixel mode for all the settings, using the initial-value page format (?20 SP J), and then using DECSTBM to set the top/bottom margins as you like, measured in pixels (300/inch). This has always worked for me, except when I screw up ;-} ================================================================================ Note 285.15 LN03s spacing problem 15 of 17 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 18 lines 29-JUN-1989 12:01 -< Looking in the wrong place... >- -------------------------------------------------------------------------------- Thank you, Bill, for the explanation. I've spent hours on the manual and it still doesn't make a lot of sense to me. I tried the modifications you suggested; cleared out the contradictory stuff, and just left a few basic setups. I still wasn't getting anywhere. I tried the sample setups the Programmer's Manual has (page 93), and got the same old 60 lines. I finally went tinhorn and called CSC. The person I talked to asked what the form looked like; SHOW QUE/FORM/FULL showed the DEFAULT form had a six-line margin at the bottom, which was forcing a form-feed after line 60. I defined a form with /MARGIN=BOTTOM=0 and set the queue to it : Voila! 66 lines, portrait and landscape, no questions asked. Sometimes the interconnectedness of all things on a VAX gets to be too much. I don't know if I ever would have thought of checking out the form definition; it certainly never bothered the LN01s what the margin was supposed to be! Thank you, one and all. ================================================================================ Note 285.16 LN03s spacing problem 16 of 17 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 11 lines 30-JUN-1989 02:09 -< side comment about excessive top margin >- -------------------------------------------------------------------------------- One other comment, just so someone else doesn't go nuts trying to solve it. On the back of the LN03 is a switch that selects 8 1/2 x 11 paper or A4 size paper. The switch is read _ONLY_ when the LN03 is first turned on. So, if someone tries to print a test page and bumps the switch, you won't notice anything wrong until the LN03 is turned off/on, which might not be for several months. The problem shows up as excessive top margin and you can spend a long time chasing all sorts of other possible causes. ================================================================================ Note 285.17 LN03s spacing problem 17 of 17 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 12 lines 30-JUN-1989 12:18 -< Same switch, different data >- -------------------------------------------------------------------------------- > On the back of the LN03 is a switch that selects 8 1/2 x 11 paper < > or A4 size paper. The switch is read _ONLY_ when the LN03 is first < > The problem shows up as excessive top margin and you can spend a < > long time chasing all sorts of other possible causes. < Another manifestation of that switch being wrong is that some characters cannot be downloaded properly from a RAM cartridge; the square blobs will be printed where the letters should be, and error 24 will be displayed on the front panel. And yes, we spent several weeks on that one!  ================================================================================ Note 286.0 STATIC BAGS (PLEASE NOT UNLISTED PARTS) 5 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 6 lines 2-FEB-1989 17:28 -------------------------------------------------------------------------------- This is a dumb question.... We are mothballing a bunch on UNIBUS hardware. Does anyone know where I can buy hex height and quad height static bags? I need about 50 ex height and 25 quad height bags. ================================================================================ Note 286.1 STATIC BAGS (PLEASE NOT UNLISTED PARTS) 1 of 5 EISNER::KENNEDY "Terry Kennedy" 11 lines 3-FEB-1989 08:11 -< Vendor, part, phone numbers >- -------------------------------------------------------------------------------- > We are mothballing a bunch on UNIBUS hardware. Does anyone know > where I can buy hex height and quad height static bags? I need about > 50 ex height and 25 quad height bags. DEC's vendor is Chiswick. They have bag # 07-1218, a 12x18 top-grade bag, for $81.00 per hundred. Since you only need 75, just by 100 and fold for quad boards. Their phone # is 800-225-8708. They will ship within 24 hours. By the way, these are the metallized bags. The pink organic bags are not good for extended storage... ================================================================================ Note 286.2 STATIC BAGS (PLEASE NOT UNLISTED PARTS) 2 of 5 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 2 lines 3-FEB-1989 09:02 -< BY ANY CHANCE DO YOU HAVE THE QUAD SIZE INFO ALSO? >- -------------------------------------------------------------------------------- > DEC's vendor is Chiswick. They have bag # 07-1218, a 12x18 top-grade ================================================================================ Note 286.3 STATIC BAGS (PLEASE NOT UNLISTED PARTS) 3 of 5 EISNER::KENNEDY "Terry Kennedy" 30 lines 3-FEB-1989 12:41 -< The *complete* list >- -------------------------------------------------------------------------------- > -< BY ANY CHANCE DO YOU HAVE THE QUAD SIZE INFO ALSO? >- Ok, you asked for it... Their part numbers are 07-xxyy, where xx and yy are dimensions in inches. Stock sizes and costs/100 are: 3x5 12.60 4x4 13.00 4x6 15.80 4x24 44.00 5x8 21.60 5x10 25.40 6x10 29.00 8x8 29.50 8x10 34.30 8x12 40.50 10x12 48.10 10x14 54.00 10x24 92.00 11x15 64.00 12x16 72.00 12x18 81.00 14x18 94.00 15x18 98.00 18x18 116.00 Don't forget to allow for slack in folding the bag. Also, the official DEC yellow sticker is part # 130-3 for a 2x2 warning sticker, 29.95 the 1000. The 4x4 size is part # 130-04 for 4x4, 49.95 the 1000. ================================================================================ Note 286.4 STATIC BAGS (PLEASE NOT UNLISTED PARTS) 4 of 5 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 0 lines 3-FEB-1989 12:57 -< THANKS TERRY! >- -------------------------------------------------------------------------------- ================================================================================ Note 286.5 STATIC BAGS (PLEASE NOT UNLISTED PARTS) 5 of 5 EISNER::KENNEDY "Terry Kennedy" 0 lines 3-FEB-1989 17:08 -< No problem... >- --------------------------------------------------------------------------------  ================================================================================ Note 287.0 Need 1200 lpm multipart printer 5 replies EISNER::ROBERTS_D "don roberts #241" 18 lines 3-FEB-1989 11:23 -------------------------------------------------------------------------------- <<< EISNER::DUA0:[NOTES$LIBRARY]VMS.NOTE;1 >>> -< VMS and bundled utilities >- ================================================================================ Note 404.0 need 1200 lpm printer 3 replies EISNER::ROBERTS_D "don roberts #241" 10 lines 2-FEB-1989 17:11 -------------------------------------------------------------------------------- We are in need of a printer with the following specs: 1200 LPM ability to print 17 cpi can handle multipart forms not solid character (can handle graphics, maybe) I don't know which conference this goes in, so I put it here. Could anyone tell me what vendors have printers that meet these specs? Of course, it must work with VMS. Don ================================================================================ Note 287.1 Need 1200 lpm multipart printer 1 of 5 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 27 lines 9-FEB-1989 02:07 -< Printronix P9012 claims 17cpi >- -------------------------------------------------------------------------------- > We are in need of a printer with the following specs: > 1200 LPM > ability to print 17 cpi > can handle multipart forms > not solid character (can handle graphics, maybe) I am looking at an ad for a Printronix P9012 and it lists 17 cpi condensed print as a capability. It would seem to also meet your other needs. The ribbon looks like their regular one on the older 300 and 600 lpm jobs. These ribbons run out of ink long before they wear out. C Itoh makes similar printers, but I don't know about speed or 17 cpi, but the C Itoh folks use a finer ribbon fabric that makes sharper characters, and I believe they are interchangable between the 2 brands. We once bought a reinker that worked well for these ribbons and only cost about $75. We don't use it any more simply because we have a pile of new ribbons to go, and hardly use our P600 since getting a laser. If you plan heavy printing (someone asking for 1200 lpm must be) and have an operator who can watch it (you have to keep your eye on it lest it over ink, or run out of ink) get a reinker and run your ribbons until the fabric starts to wear out (I did one 6 times and it still looked new). ================================================================================ Note 287.2 Need 1200 lpm multipart printer 2 of 5 EISNER::ROBERTS_D "don roberts #241" 9 lines 9-FEB-1989 11:34 -< We print a little >- -------------------------------------------------------------------------------- >If you plan heavy printing (someone asking for 1200 lpm must be) We have an LG01 (that is a story in itself, what pain, Digital says it fails so often cause we overwork it), a C.Itoh something or other that is SYS$PRINT, four Printronix P300's, three LA210's, two LA120's, an LN03, a couple other printronix models, and a couple zillion PC printers. Yah, we do a little printing here. Like most companys it is a waste of paper for the most part, but they don't listen to me on that... ================================================================================ Note 287.3 Need 1200 lpm multipart printer 3 of 5 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 15 lines 9-FEB-1989 13:59 -< Printronix printers just keep chugging along >- -------------------------------------------------------------------------------- >> -< Printronix P9012 claims 17cpi >- Just a general comment on Printronix printers. They work. And work, and work, and work... A number of my clients use the older P300 and P300 printers with a work load that would kill any other printer in its class. After years of constant use, they still produce good print quality and rarely require service. A former employer is still thrilled with the P300 he bought in 1977. Now, I realize that this doesn't mean that the new P9012 is built to the same high standards, but I would certainly give it a close look. Alan ================================================================================ Note 287.4 Need 1200 lpm multipart printer 4 of 5 EISNER::ROBERTS_D "don roberts #241" 14 lines 9-FEB-1989 14:10 -< On the other hand >- -------------------------------------------------------------------------------- >Just a general comment on Printronix printers. They work. And >work, and work, and work... At my previous position there were five P300 printers that had basically a constant workload eight hours per day. They were all at least seven years old. These printers would each fail at least once per week (=five service calls per week on printers). Normal failures included burned out coils, worn hammers, bad A logic boards and bad hammer driver boards. Occasional problems were bad shuttle assemblies and stepper motors, and sometimes a belt wearing out. They were workhorses, but the frequency of failure was a problem, especially when two others would fail when we were waiting for the FE to work on the first one. Not bad for their age and workload, though. ================================================================================ Note 287.5 Need 1200 lpm multipart printer 5 of 5 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 20 lines 9-FEB-1989 15:42 -< P600 not as solid as P300 >- -------------------------------------------------------------------------------- >Just a general comment on Printronix printers. They work. And >work, and work, and work... A number of my clients use the >older P300 and P300 printers with a work load that would kill >any other printer in its class. After years of constant use, The P600 initially was just a hopped up P300 and failed regularly by banging itself to death. They came out with various NECESSARY 'improvements' such as carbide hammer tips, and 'phase fired' hammer banks and a new motor starter scheme. Ours never really became solid for long heavy use, but works well if serviced. The service $s charged do reflect the trouble they must generally be... It is possible that newer P600s are very rugged, I don't have one. It is also possible that the new 1200 lpm one has been well engineered, but I would love to talk to someone who has run one for several months. The C-Itoh lower speed units produced much better quality print, and don't sound as though they have a death rattle in their throat.  ================================================================================ Note 288.0 KENNEDY TAPE DRIVES 1 reply EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 4 lines 3-FEB-1989 14:38 -------------------------------------------------------------------------------- Does anyone have information on Kennedy tape drives? I need the home office phone number. I want to obtain the name of a local distributor and info on the 6550 units. ================================================================================ Note 288.1 KENNEDY TAPE DRIVES 1 of 1 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 21 lines 3-FEB-1989 15:59 -< Access to Kennedy tape drives >- -------------------------------------------------------------------------------- Kennedy 1600 Shamrock Ave Monrovia CA 91016 818-357-8831 Their east coast repair depot is in Woburn MA. We buy them from: Total Tec Larry Holtzenthaler (our salesman) in NJ: 201-906-6500 They have a woburn MA office, but we use NJ, and they are a Dilog (and everything else) dealer and I think may be Dilog's East Coast Distributor. Almost ANYONE selling 3rd party gear into the DEC market seems to offer Kennedy, but Total Tec buys direct, and is not 'picking them up' for you from elsewhere. (usual stuff: I'm not affiliated with ANY of the above)  ================================================================================ Note 289.0 Will it fly, orville? 2 replies EISNER::ROBERTS_D "don roberts #241" 28 lines 3-FEB-1989 18:23 -------------------------------------------------------------------------------- We are planning to make some changes to our configuration here. Proposed: Ethernet backbone _________________________________________________________________________ | | | | uVAXII 8350--RA60 8350 |-ds200 [there | |-------- CI ----------| |-ds200 are |-RD53 /\ |-ds100 more] (paging & _____/ \_____ |-ds100 swapping) | | HSC50 HSC50 \____________/ |-ra81 (system disk) [dual |-ra82 (5) ported] |-ra81 (two) The two 8350's in a CI cluster, the uVAXII in an NI cluster with one of the 8350's, vms 5.0-2 unless 5.1 is out by then. Current: LAVC 8350 boot node with uVAXII, 3 RA81, 1RA82, 1 RA60, rd53 on uVAXII as local paging and swapping disk with VMS 4.7 Questions: 1) Mixing the NI and CI, how well will (can) this work, will the uVAXII be able to get to the disks through the 8350 LAVC and 2) How complicated will terminal and printer setups be with this mixed cluster? 3) What do we need to know to set this up? We will bring up the 8350 with two HSC's then add the LAVC equipment later, this being all part of a change of locations. ================================================================================ Note 289.1 Will it fly, orville? 1 of 2 EISNER::OSUDAR "John Osudar" 27 lines 4-FEB-1989 10:56 -< MI cluster "flies" fine, wilbur >- -------------------------------------------------------------------------------- Looks like a simple mixed-interconnect cluster to me... A couple of comments, though: (1) You say "the uVAXII in an NI cluster with *one of* the 8350's". Actually, both 8350's *and* the uVAXII will be in one (MI) cluster together. If you were implying that only one of the 8350's will serve the HSC-based disks to the uVAXII, that's different -- but I would recommend that you be VERY careful if you do that (since unavailability of that 8350 will make your uVAXII inoperable until it comes back up.) (2) What "terminal and printer setups" are you concerned about? With your terminals and printers on terminal servers, there shouldn't be much of a problem. For printers, you'll have to create LTAxxx: ports on at least one node -- but you may want to choose to create the ports on both 8350's, call them by the same name, spool them to the same queue; then you start the queue (on one node, obviously). You can then set up an automated switchover procedure in case that node goes down, that'll simply restart all the queues over on the other node. (3) "What you need to know to set this up" should mostly be contained in the V5 manual on setting up a cluster. It's 99.44% true that "a cluster is a cluster is a cluster" under V5 -- and your configuration appears to be one of the more trouble-free ones (ours, a dual-boot-node NI-only cluster, seems to be one of the more difficult ones!) If you have more specific questions, I'm sure there enough cluster expertise here on DECUServe to help you... ================================================================================ Note 289.2 Will it fly, orville? 2 of 2 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 16 lines 6-FEB-1989 13:11 -< Some things we do in our MI cluster >- -------------------------------------------------------------------------------- One thing we've done for our MI cluster is to refrain from offering the cluster service name on the non-CI systems. We do this because our satellites are workstations and we don't want our server load balancing picking a workstation. Likewise, we have the execution queues for the satellites disabled during the day so that batch jobs won't run on them (SYS$BATCH points at *all* nodes in the MI cluster). The satellites also do not define the DECnet cluster alias name, so that foreign SET HOSTs don't pick a satellite. Watch out also for tapes offered by HSCs. If a batch job running on a satellite tries to allocate one, the job just hangs. We have a command procedure to call in the event a job wants a tape drive that acts as a queued allocate. It first checks to see if the job is executing on a satellite. If so, it obtains the batch job's procedure name from the job controller and resubmits the job to a CI-only generic batch queue.  ================================================================================ Note 290.0 VS3100 Non-DEC Peripherals 6 replies EISNER::HARVEY "Jack Harvey" 17 lines 3-FEB-1989 18:46 -------------------------------------------------------------------------------- This topic is compatibility (or problems) of/with non-DEC peripherals on the VAXstation 3100. The 1989 January-March issue of the VAX Systems and Options Catalog contains the following from page III.4 in the VAXstation 3100 discussion: "o Small Computer System Interface (SCSI) allows external expansion of industry-available SCSI peripherals." There have been some indications that many "industry" peripherals do not strictly meet the SCSI specification, and that DEC had to struggle to find chips that would. This raises the possibility that just any old device claimed to be SCSI might not work with the VS3100. Does anyone know of any devices specifically claimed by the manufacturer to be VS3100 compatible, or tested and found to be compatible? Or is this a groundless fear? ================================================================================ Note 290.1 VS3100 Non-DEC Peripherals 1 of 6 EISNER::BURNET "John Burnet" 21 lines 20-JUN-1989 21:58 -< Wanted: non-DEC SCSI disks (& the moon with a fence around it) >- -------------------------------------------------------------------------------- > < Note 290.0 by EISNER::HARVEY "Jack Harvey" > > Does anyone know of any devices specifically claimed by the > manufacturer to be VS3100 compatible, or tested and found to > be compatible? Or is this a groundless fear? If anyone has any information about third-party SCSI-interface disk drives that will work with the VS3100, I (and, I'm sure, many many others besides Jack) would love to hear it. Since DEC's price for a 300-MB RZ55 with expansion box is almost $7k, it's obvious that the market for non-DEC SCSI drives for the VS3100 and DS3100 is just waiting to be exploited. Have any brave souls out there in DECUServeland tried using drives other than the RZ22, 23, 55 with the VS3100? What were the results? I hope this isn't another situation (a la VAXBI) where DEC is the only game in town. I'm asking partly out of desperation... we just got two VS3100s and now we find that we have to upgrade one from a satellite page/swap-disk-only configuration and add another disk to make it a standalone system. The RZ23 is the only DEC disk bigger than 50 MB that we can afford, and we would really like quite a bit more than the 100 MB that the '23 provides. But seven thousand dollars for 300 MB? :-( ================================================================================ Note 290.2 VS3100 Non-DEC Peripherals 2 of 6 EISNER::BURNET "John Burnet" 6 lines 21-JUN-1989 01:08 -< That's a software problem, bub >- -------------------------------------------------------------------------------- Should've poked around a bit more before asking about SCSI disks... I see from Mark S's note 13.0 in WORKSTATIONS that the problem is not necessarily finding SCSI disks that will work with the VS3100, but finding drivers for those disks. Rats. Guess we'll compromise this time around and shell out for an RZ23. ================================================================================ Note 290.3 VS3100 Non-DEC Peripherals 3 of 6 EISNER::NAGY "Frank J. Nagy, VAX Guru" 5 lines 21-JUN-1989 07:15 -< CDC Wren-Vs apparently work as is on 3100 >- -------------------------------------------------------------------------------- I don't have direct experience (no VS3100's available to me :-( as yet), but I heard from several sources that CDC/Imprimis Wren-V disks work as is on the VS3100 SCSI bus (using the Digital drivers). This *may* also mean that the other Wren series disks and (possibly) the CDC Swift (3.5" drives) might also work. ================================================================================ Note 290.4 VS3100 Non-DEC Peripherals 4 of 6 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 13 lines 21-JUN-1989 12:06 -------------------------------------------------------------------------------- >Should've poked around a bit more before asking about SCSI disks... I see >from Mark S's note 13.0 in WORKSTATIONS that the problem is not necessarily >finding SCSI disks that will work with the VS3100, but finding drivers for DEC does intend for us to put random other disks on, the only question is how soon. The guy who wrote their driver did it as a true port/class driver with open-ended expansion possibilities. The Atlanta tape might be worth getting. There was also a BOF session with several DECies eager for input from users. The source for his new driver wasn't out as of Atlanta, but was expected out on the next release (5.2?). Is there a chance you could use a larger drive of similar but in some dimension(s) larger geometry and for now only use part of it? ================================================================================ Note 290.5 VS3100 Non-DEC Peripherals 5 of 6 EISNER::BURNET "John Burnet" 14 lines 21-JUN-1989 20:46 -------------------------------------------------------------------------------- > < Note 290.4 by EISNER::BRUCE "Barton F. Bruce - CCA/MA" > > The Atlanta tape might be worth getting. Thanks -- I'll do that. > Is there a chance you could use a larger drive of similar but in some > dimension(s) larger geometry and for now only use part of it? You mean fake out the driver and tell it that it's talking to an RZ55? That might do it. In any case, I think we'll at least explore the possibility of using the Imprimis drives as Frank suggested. It's encouraging that DEC does intend for 3rd-party disks to work on the bus eventually. ================================================================================ Note 290.6 VS3100 Non-DEC Peripherals 6 of 6 EISNER::KENNEDY "Terry Kennedy" 10 lines 21-JUN-1989 23:56 -< More on SCSI design >- -------------------------------------------------------------------------------- > Is there a chance you could use a larger drive of similar but in some > dimension(s) larger geometry and for now only use part of it? SCSI is self-sizing (one of the commands is "how big art thou?") Unless DEC deliberatly violated the SCSI guidelines *or* your disk is so big 32 Gig? that you exceed some VMS design assumption, you should be ok size-wise. Also, one of the goals of SCSI is compatibility. Any drive using Common Command Set (and implementing it properly) should work.  ================================================================================ Note 291.0 VT330 PRINTER PORT 15 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 6 lines 5-FEB-1989 21:07 -------------------------------------------------------------------------------- Does anyone know how to get the VT330 printer port to work? No matter what I do it will not work. I am using the same DEC Connect cables that I use to attach the comm ports to the system. P.S. status line always shows PRINTER: NONE ================================================================================ Note 291.1 VT330 PRINTER PORT 1 of 15 EISNER::HAHN "DECUServe Vice Chair" 1 line 5-FEB-1989 22:44 -< Check setup >- -------------------------------------------------------------------------------- I presume you have turned the printer on in the VT330 setup. ================================================================================ Note 291.2 VT330 PRINTER PORT 2 of 15 EISNER::HARVEY "Jack Harvey" 6 lines 6-FEB-1989 03:33 -< Printer Setup? >- -------------------------------------------------------------------------------- Sorry, no experience with the VT330. However, I remember having the same problem years ago with a VT220. The solution was correct setup of the printer... Seems the VT220 wouldn't believe it had a printer connected unless the printer had the correct communication characteristics. My memory is dim, but I think it was a parity problem. ================================================================================ Note 291.3 VT330 PRINTER PORT 3 of 15 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 15 lines 6-FEB-1989 08:24 -< IT IS WIRING OR THE TERMINAL >- -------------------------------------------------------------------------------- This printer works OK off the printer port of...... VT102 HDS 2200 MicroTERM 4325 Also I place a break out box on the line and nothing is comming down. BTW - the firmware is V1.1 P.S. Do you realize DEC is the only major hardware vendor in the DEC market place without a free tech support number..... ================================================================================ Note 291.4 VT330 PRINTER PORT 4 of 15 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 5 lines 6-FEB-1989 09:29 -< MORE INFO >- -------------------------------------------------------------------------------- Pierre Hahn suggested I try plugging a VDT into the printer port which I did. I can type on the remote VDT and what I am typing appears on the VT330 screen and is being sent to the computer. But there is no way to get to print to the remote VDT. This is strange.... ================================================================================ Note 291.5 VT330 PRINTER PORT 5 of 15 EISNER::KENNEDY "Terry Kennedy" 3 lines 6-FEB-1989 10:20 -< Need hardware handshake line, I think >- -------------------------------------------------------------------------------- Look at the modem control signals. If this is on a MMJ connector, you don't have all of the RS-232 signals, but look at DSR/DTR and RTS/CTS (which one of the paie is used escapes me at the moment...) ================================================================================ Note 291.6 VT330 PRINTER PORT 6 of 15 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 7 lines 6-FEB-1989 11:02 -< CONFIRMED WITH A BREAK OUT BOX >- -------------------------------------------------------------------------------- > Look at the modem control signals. If this is on a MMJ connector, you > don't have all of the RS-232 signals, but look at DSR/DTR and RTS/CTS The VT330 printer port is raising CTS, DSR, DCD The VDT I have attached to it is raising RTS,DTR ================================================================================ Note 291.7 VT330 PRINTER PORT 7 of 15 EISNER::CONROY "Alan Conroy" 9 lines 6-FEB-1989 11:54 -< DTR is the culprit >- -------------------------------------------------------------------------------- >Pierre Hahn suggested I try plugging a VDT into the printer port which I >did. I can type on the remote VDT and what I am typing appears on the >VT330 screen and is being sent to the computer. But there is no way to >get to print to the remote VDT. This is strange.... Did the PRINTER:NONE status go away with the VDT connected? To the best of my knowledge (and I have messed with this a lot), PRINTER:NONE ONLY comes from the VT330 not seeing DTR on the printer port. Have you tried this on another VT330? ================================================================================ Note 291.8 VT330 PRINTER PORT 8 of 15 EISNER::KENNEDY "Terry Kennedy" 3 lines 7-FEB-1989 00:06 -< FCO Kit >- -------------------------------------------------------------------------------- Try FCO kit EQ-01513-01, available from DECdirect for a mere (!) $190.00. This upgrades the terminal to Version 2.1. Of course, if you're already at V2.1, try something else... ================================================================================ Note 291.9 VT330 PRINTER PORT 9 of 15 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 9 lines 7-FEB-1989 00:08 -< Second the DTR comments >- -------------------------------------------------------------------------------- I agree... if the printer does not raise DTR, the VT will show PRINTER: NONE. Make sure you're using the correct MMJ to 25-pin adapter - there are several which look alike, but interchange send/receive pins. I'm at home - don't have info here. I've got an LA50 plugged into my VT340 (V2.1) that works just fine. Also.. if you're using an LA50, they are known to fail by either dropping DTR permanently or not doing ^S/^Q. ================================================================================ Note 291.10 VT330 PRINTER PORT 10 of 15 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 5 lines 7-FEB-1989 00:32 -< IT HAS GOT TO BE THE ADAPTER OR THE FIRMWARE >- -------------------------------------------------------------------------------- > Make sure you're using the correct MMJ to 25-pin adapter - there > are several which look alike, but interchange send/receive pins. I am using an 8571-A adapter - do you know if it is the right one? ================================================================================ Note 291.11 VT330 PRINTER PORT 11 of 15 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 3 lines 7-FEB-1989 09:10 -< Looks like the right adapter... >- -------------------------------------------------------------------------------- I'm using that exact one (mine says Rev B). So, looks like a firmware problem, or DTR from the printer isn't high. If you're still under warrenty or maintenance, DEC should do the firmware upgrade free. ================================================================================ Note 291.12 VT330 PRINTER PORT 12 of 15 EISNER::PROVOST 7 lines 7-FEB-1989 14:05 -< FREE TECH SUPPORT >- -------------------------------------------------------------------------------- > < Note 291.3 by EISNER::KILLEEN "Jeff Killeen DECUServe Chair" > >P.S. Do you realize DEC is the only major hardware vendor in the DEC >market place without a free tech support number..... But, DEC has DECUS! ================================================================================ Note 291.13 VT330 PRINTER PORT 13 of 15 EISNER::MCCORMICK_B "Bob McCormick" 7 lines 8-FEB-1989 18:50 -< Try connecting Ready-Busy >- -------------------------------------------------------------------------------- Jeff: If you're using the DECconnect H8571-A adaptor, try connecting pins 1 and 6 on the MMJ (DECconnect wire) which is ready - busy... or is it busy - ready... I've found that to be a problem when using the adaptors when the EIA-232-D device expects the DTR stuff... ================================================================================ Note 291.14 VT330 PRINTER PORT 14 of 15 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 18 lines 11-FEB-1989 23:15 -< DEFEATED >- -------------------------------------------------------------------------------- Well I give up.... We have tried.... Four MMJ connectors Two BC16E cables Two VT330 Upgrading both of them to V2.1 firmware I put a tri-state breakout box on the H8571-A and it looks like the VT330 isn't even monitoring DTR. As much as it pains me I am shipping one the terminals off to DEC F.S. carry in. ================================================================================ Note 291.15 VT330 PRINTER PORT 15 of 15 EISNER::KENNEDY "Terry Kennedy" 4 lines 12-FEB-1989 06:52 -< Send me one to beat up >- -------------------------------------------------------------------------------- > Well I give up.... Jeff - send one of 'em down to me (you have the address) and I'll make it work...  ================================================================================ Note 292.0 DMV11 switch setting information needed 4 replies EISNER::BURNET "John Burnet" 22 lines 7-FEB-1989 12:27 -------------------------------------------------------------------------------- Can anyone help us out with switch-setting information for the DMV11? We just added a MicroVAX II to our network, connected via a leased line into a DMV on the MicroVAX end and a DMF on an 11/750 on the other end. The circuit won't come up -- it stays in "on-starting" state. We're not sure that we have the switches set properly on the DMV; the manual is ambiguous and for some of the switches it says something to the effect of "this switch should be off if the software will set it, or on if the software won't set it." (Sorry I can't be more specific -- I don't have the manual here.) We've tried the settings that seemed to be correct, but no luck so far. As far as we can tell, there are no other problems that would prevent the circuit from coming up -- the DECnet PAK is installed, loopback tests show that the leased line and modems are okay, etc. My plea for help is: Can anyone supply any more detailed information on the DMV switch settings -- from a DEC service guide, maybe? I thought I'd see if any wizard here could shed some light on this before we continue the hassles with DEC field circus. Thanks a bunch. P. S. I have seen 164.* in this conference but thought I'd create a new topic for this, since it's not MV3600-related. ================================================================================ Note 292.1 DMV11 switch setting information needed 1 of 4 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 3 lines 7-FEB-1989 13:26 -< more details, please >- -------------------------------------------------------------------------------- What electrical interface (232, 422, etc.) are you using? And what are the rev levels of the card and the distribution panel? ================================================================================ Note 292.2 DMV11 switch setting information needed 2 of 4 EISNER::BURNET "John Burnet" 25 lines 7-FEB-1989 16:07 -------------------------------------------------------------------------------- Since I posted .0, the local field-circus office where the machine is located sprang into action, and between their efforts and our calls to Colorado TSC, we may be on our way to a resolution of the problem. The difficulties may be due to having old or incomplete documentation (our manual and a service-manual page faxed to us by a TSC person show conflicting assignments of the switches on the distribution panel), and the problem seems to have been aggravated by the field-service person's lack of the proper service guides and/or training. We're trying to hook up a sync RS-232 connection via the DMV, and so far we haven't been able to figure out what the switch settings on the board itself and on the dist. panel should be. Sorry, I don't have the rev levels available. Do the switch assignments differ depending on the board and/or distribution panel revisions? In any case, the pressure has been put on DEC to make it work; since there are no technical people at that location at the moment (the machine's in Minneapolis and we're in Chicago), there's not much else we can do but leave it to DEC. I guess the MicroVAX II is now handled within DEC field service by the Unprofitable Products Group or something like that. At least, the service person who's in charge of this newly-installed VAX has always been known to us as our DECmate repairman. ================================================================================ Note 292.3 DMV11 switch setting information needed 3 of 4 EISNER::PRIGOT "Jonathan M. Prigot" 86 lines 7-FEB-1989 16:27 -< DMV11 Switches >- -------------------------------------------------------------------------------- Here is the DMV11-AA information. (Couldn't find you in WHO..., so the information is here, hopefully in time.) On the M8053 processor board are four sets of DIP switches: E113 E101 E54 E53 E53 and S1 and S2 on E54 are the CSR switches ! E53 ! E54 ! CSR 8 7 6 5 4 3 2 1 2 1 ON 760020 ON 760040 ON ON 760060 ON 760100 ON 760200 ON ON 760300 ON 760400 ON ON 760500 ON ON 760600 ON ON ON 760700 ON 761000 ON 762000 ON ON 763000 ON 764000 E54 switches 3-8 are the vector 8 7 6 5 4 3 Vector ON ON 300 ON ON ON 310 ON ON ON 320 : : ON ON ON 700 E113 is the "DDCMP address register. Tributary/Password" Off switches produce a ONE in the associated register. (Normally set to all OFF) E101 switches are: (0 = "ON") S1 - "Mode enable" S2 - Unit # for booting S3 - Auto answer S4 - Power on boot enable S5 - Remote load detect S6-8 Mode when switch 1 is set 6 7 8 Mode ON ON ON HDX pt-pt DMC mode ON ON FDX pt-pt DMC mode ON ON HDX pt-pt ON FDX pt-pt ON ON HDX control station ON FDX control station ON HDX tributary FDX tributary S9 - High speed (set above 19.2Kbps) S10- ON=V.35, OFF=EIA Distribution panel switches should be set to: 1 -OFF 2 -OFF 3 -OFF 4 -ON 5 -OFF 6 -OFF 7 -ON 8 -ON 9 -OFF 10-ON 11-ON 12-ON 13-ON 14-ON 15-OFF 16-OFF 17-ON 18-ON 19-ON 20-OFF ================================================================================ Note 292.4 DMV11 switch setting information needed 4 of 4 EISNER::BURNET "John Burnet" 8 lines 8-FEB-1989 12:29 -< Thank you! Case closed. >- -------------------------------------------------------------------------------- > < Note 292.3 by EISNER::PRIGOT "Jonathan M. Prigot" > > -< DMV11 Switches >- Thanks much for the help. Aided by that information, and after some back-and-forthing with Field Circus, we did finally get the situation straightened out, and the DECnet circuit is now up. As expected, the cause of all the confusion was just old or inappropriate switch and cabling information.  ================================================================================ Note 293.0 8650 will not auto restart 4 replies EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 16 lines 7-FEB-1989 13:02 -------------------------------------------------------------------------------- We have an interesting problem with an 8650: It will not auto reboot. After a shutdown or crash where auto restart is requested the machine is dead, not even the console works. This is a headache more than a problem since simply cycling power reboots the machine (after several minutes of loading microstore...). A snap is generated as the machine attempts to reboot. So far we have changed all boards in the CPU with the exception of Memory, the DW780, and CI780. Our FS engineer, along with central support, have examined a number of the SNAP files and found they all point to errors on the Data path. Extended runs of the 8650 CPU diagnostics have shown nothing amiss. At this point our FS people (CDC to be exact) are suggesting that the DW780 be pulled and checked. They have also mentioned that a new revision of the 8650 V6 console (rev. 9.12) may help. ================================================================================ Note 293.1 8650 will not auto restart 1 of 4 EISNER::HAMPARIAN "Don Hamparian - Battelle" 8 lines 7-FEB-1989 13:22 -< Our 8650 chokes with INT STK NOT VALID >- -------------------------------------------------------------------------------- We see a variation of this problem on one of our 8650s. If you bring it down and then it autoboots or you reboot it manually it will (usually) give you a INT STK NOT VALID, do a snapshot, try again, choke, etc until we power it off. Our field service guys say that the console sw needs to be replaced. I don't know the rev off the top of my head, but if you want I will check to see what console rev they want to take us to. ================================================================================ Note 293.2 8650 will not auto restart 2 of 4 EISNER::GLEASON "CyberPunk" 14 lines 8-FEB-1989 01:17 -< I thought it was just my 8650... >- -------------------------------------------------------------------------------- > -< Our 8650 chokes with INT STK NOT VALID >- > (usually) give you a INT STK NOT VALID, do a snapshot, try again, > choke, etc until we power it off. One of my 8650s started doing this exact same thing, immediately after I upgraded it to VMS V5. Field service has upgraded the console software, changed the front ent T11 board, and swapped out a UNIBUS controller, without fixing the problem. We suspect some UNIBUS problem, since if we set STARTUP_P1 to MIN and do a minimal startup, a restart reboot works fine - the theory is that something (hardware, software or firmware) is unhappy with the state of the UNIBUS(es). ================================================================================ Note 293.3 8650 will not auto restart 3 of 4 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 22 lines 8-FEB-1989 16:42 -------------------------------------------------------------------------------- >< Note 293.0 by EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" > > -< 8650 will not auto restart >- > > We have an interesting problem with an 8650: It will not auto reboot. > After a shutdown or crash where auto restart is requested the machine > is dead, Define "dead". What is the last thing that the console printed? Is the FP key in "Local/Secure" position? Is the console prompt "ROM>"? > They have also mentioned that > a new revision of the 8650 V6 console (rev. 9.12) may help. > My console pack on both 8600 and 8650 (they use the same pack) says "console pack V10.0", and the console software is V9.12. One of the fixes in 9.12 is related to failures in KAF (Keep Alive Failure) handling and snapshot taking. There were some problems with earlier versions of the console where they would report a stack underflow during a snapshot. Stu ================================================================================ Note 293.4 8650 will not auto restart 4 of 4 EISNER::HAMPARIAN "Don Hamparian - Battelle" 8 lines 9-FEB-1989 07:37 -< CONSOLE V10 SUPPOSED TO FIX OUR PROBLEM >- -------------------------------------------------------------------------------- According to our field service guys (pretty good ones), our problem is fixed by console pack V10.0. We haven't had an chance to test though. Our machine fails usually right when VMS comes up and before the connection manager starts spitting out console messages. There are a variety of failure messages, but usually the INT STK error is the one we see.  ================================================================================ Note 294.0 8650 V6 Console Rev 9.12 4 replies EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 4 lines 7-FEB-1989 13:04 -------------------------------------------------------------------------------- CDC Field Service has told me that a new revision of the 8650 V6 console media is available, V6, rev 9.12. This revision is reported to contain some micro code fixes for 8650s that show a timing problem that causes them to be unable to auto restart. ================================================================================ Note 294.1 8650 V6 Console Rev 9.12 1 of 4 EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 11 lines 8-FEB-1989 14:11 -< How does one order the revised console? >- -------------------------------------------------------------------------------- I have been told (by CDC) that the part number for the console and microcode is BD-FI17J-ME. When I attempt to enter this into the Electronic store I get an "invalid format" error. I also get one for the part number from the original V6 console disk we have (BD-FI18F-DE). Is there some sort of trick to this, or do I have to go around the end-zone to order this stuff? (I would ask our DECrep 100 but they are currently being replaced with a DECrep 200, we get to meet them next Monday...) Seton ================================================================================ Note 294.2 8650 V6 Console Rev 9.12 2 of 4 EISNER::GRANZOW 19 lines 8-FEB-1989 15:54 -< 86XX CONSOLE >- -------------------------------------------------------------------------------- < Note 294.1 by EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" > -< How does one order the revised console? >- I have been told (by CDC) that the part number for the console and microcode is BD-FI17J-ME. When I attempt to enter this into the Electronic store I get an "invalid format" error. I also get one for the part number from the original V6 console disk we have (BD-FI18F-DE). Is there some sort of trick to this, or do I have to go around the end-zone to order this stuff? (I would ask our DECrep 100 but they are currently being replaced with a DECrep 200, we get to meet them next Monday...) Seton THE CORRECT NUMBER SHOULD BE BC-FI17J-ME , NOT BD: BD IS A RA60 PACK WHERE BC IS A RL02 PACK. THE NUMBER ENDS IN "ME" BECAUSE THATS THE UNLICENSEED VERSON (WITHOUT DIAG'S) . YOU MIGHT HAVE TO GO THROUGH DEC DIRECT. ================================================================================ Note 294.3 8650 V6 Console Rev 9.12 3 of 4 EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 9 lines 15-FEB-1989 09:45 -< Part Numbers as of 15-Feb-1989 >- -------------------------------------------------------------------------------- I checked the Digital Electronic Store, left a message, and have been given the following new (as of 15-Feb-1989) part numbers: BC-FI17K-ME -- VAX 8600/8650 Console (No diagnostics) BC-FI18K-DE -- VAX 8600/8650 Console W/Diagnostics (Must have diagnostic licence) The store posted *no* version information. I will be checking with our DEC Rep soon... ================================================================================ Note 294.4 8650 V6 Console Rev 9.12 4 of 4 EISNER::KENNEDY "Terry Kennedy" 5 lines 15-FEB-1989 09:54 -< The 'K' is the release code >- -------------------------------------------------------------------------------- > BC-FI17K-ME -- VAX 8600/8650 Console (No diagnostics) ^ This should be the release-specific identifier. Thus, the prior release would be BC-FI17J-ME. From this info and the packing list of your current media, should be able to determine if you are current.  ================================================================================ Note 295.0 EMC MAXPORT/Multiple Spindle Disk Drives 10 replies EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 17 lines 7-FEB-1989 13:12 -------------------------------------------------------------------------------- I have just finished talking with the sales people from EMC2. They showed me the MAXPORT drives -- a system where 1 to 4 CDC Sabre series drives can be run off of a single DSA port. The basic design is an intelligent controller that talks to a single DSA port on one side and up to four spindles on the other. The whole set up is supposed to look like a single physical drive to the DEC side. The intelligent controller is based on one of the 68000 series chips and includes diagnostics, etc. As of now (February, 1989) EMC says that they have had the MAXPORT series out for about six months. They also say they have had no problems with DEC compatibility. Has anyone had any experience with the MAXPORT series drives, or any other EMC disk drives? What about the concept of putting more than one spindle together to make a large geometry drive? ================================================================================ Note 295.1 EMC MAXPORT/Multiple Spindle Disk Drives 1 of 10 EISNER::MERRIMAN "George Merriman - CCA/NY" 6 lines 7-FEB-1989 16:38 -< What about performance? >- -------------------------------------------------------------------------------- > What about the concept of putting more than one spindle together to > make a large geometry drive? Does this eliminate the performance advantages of multiple spindles vs. the same capacity on one giant disk? ================================================================================ Note 295.2 EMC MAXPORT/Multiple Spindle Disk Drives 2 of 10 EISNER::KENNEDY "Terry Kennedy" 6 lines 7-FEB-1989 19:05 -< Decreases throughput, reliability >- -------------------------------------------------------------------------------- > Does this eliminate the performance advantages of multiple spindles > vs. the same capacity on one giant disk? Probably - The UDA is supposed to handle simultaneous transfers on all 4 ports. Also, what happens if one of the disks fails - does the whole 'drive' go down? ================================================================================ Note 295.3 EMC MAXPORT/Multiple Spindle Disk Drives 3 of 10 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 10 lines 8-FEB-1989 07:48 -< One HDA = bad = whole drive = bad ! >- -------------------------------------------------------------------------------- > Also, what happens if one of the disks fails - does the whole > 'drive' go down? YES ! I have been looking at this setup from a few manufacturers and they all answer the question the same: The whole logical drive will fail. Also, Ken correct me if I'm wrong, the HSC has no way currently to take advantage of optimizing seeks and/or transfers for all the drives that are behind the one port. The HSC doesn't know they are there. ================================================================================ Note 295.4 EMC MAXPORT/Multiple Spindle Disk Drives 4 of 10 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 13 lines 8-FEB-1989 08:34 -< WHERE THE MARKET IS HEADING BUT...... >- -------------------------------------------------------------------------------- I know of two manufactures who have these type of systems in the works. The one I know the most details of is planning a cabinet with up to 16 5 1/4 inch drives. They claim in the lab that on a normal timesharing system they are seeing what amounts to zero seek time waits over 80 percent of the time. Why haven't they come to market? Until they have drive shadowing working they feel this is not viable model. They want to be able to run 8 masters with 8 shadows. BTW - IBM uses this approach on one of their storage systems. I have heard from several disk wizards (including the ones from DEC) that multiple spindles is where everyone will be going for performance reasons. ================================================================================ Note 295.5 EMC MAXPORT/Multiple Spindle Disk Drives 5 of 10 EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 11 lines 8-FEB-1989 09:37 -< More on my session with Emulex >- -------------------------------------------------------------------------------- Performance: Emulex claimed that the multiple spindles did not reduce throughput, rather increased it, since the chance of having a head at the proper cylinder when the current read is complete is increased. This is possible since there controller takes care of multiple seeks overlapping data transfere. Emulex also says that they "format" the logical disk in a way that allows them to Reliability: Emulex claims to have this set up running in the IBM midrange area for two years or so. They say they have built in "smarts" to look at logical block replacement on the controller level and do predictive maintenance. ================================================================================ Note 295.6 EMC MAXPORT/Multiple Spindle Disk Drives 6 of 10 EISNER::KENNEDY "Terry Kennedy" 30 lines 8-FEB-1989 20:40 -< Technical summary of tradeoffs >- -------------------------------------------------------------------------------- > Performance: Emulex claimed that the multiple spindles did not > reduce throughput, rather increased it, since the chance of having > a head at the proper cylinder when the current read is complete > is increased. "Throughput" is based on a number of factors. Some are: 1) Transfer rate - the spped at which data comes off the disk 2) Seek time - how long it takes to move the heads to the right place 3) Rotational latency - how long it takes for the disk to spin to the right place 4) Controller/logic overhead The 'multiple-disk-per-DSA-port' approach exhibits *lower* throughput on items 1 and 3, and usually lower on 4. It is better at 2. These comments presume drives with equal performance characteristics. The reasons are: a) Multiple DSA ports can transfer at the same time. Grouping drives on one DSA port loses this advantage. b) With multiple spindles, seek time will be better (if the drives have the same performance as a single biug drive). However, you get this advantage with multiple drives on multiple DSA ports as well... c) The DSA controller (UDA, KSC, etc). 'knows' where each drive is, and orders transfers based on rotational position. With multiple drives per DSA port, this optimization can only be performed by the DSA-to-whatever port adapter. Most drives used in this this application do not provide RPS information, anyway. d) The disk data needs to be buffered in the DSA-to-whatever converter before being sent to the host (due to MSCP protocol issues). Thus, you've just added another fixed delay. ================================================================================ Note 295.7 EMC MAXPORT/Multiple Spindle Disk Drives 7 of 10 EISNER::KENNEDY "Terry Kennedy" 25 lines 8-FEB-1989 20:49 -< Comments about the IBM implementation >- -------------------------------------------------------------------------------- > BTW - IBM uses this approach on one of their storage systems. I have > heard from several disk wizards (including the ones from DEC) that > multiple spindles is where everyone will be going for performance > reasons. The item in question is the IBM 9332 Direct Access Storage Device, Model 400. That is *nothing* like what's being done for DSA. The 9332 has two pos- itioning arms on *one* set of platters (ie., in one drive). The interface is SCSI, and reports 2 units. While the heads move independently, you can queue up transfers for both drives. The one which is nearest the heads rotationally (assuming both actuators are on cylinder) happens first. This is because with one spindle and one control logic, it is very easy to deter- mine which is closer. That is a *lot* harder to do with loosely coupled disks like the DSA implementations being discussed. By the way, although the 9332 has this functionality, it is *not* used when connected to the disk/tape adapter in the 9370 CPU. The adapter is too unsophisticated to do it. It *waits* for one drive to finish before talking to the next one. That adapter is due to be replaced with a second- generation unit sometime in the future. If DEC is looking at doing it, I suspect we'd *never* see it in the RA series, but only in RF and RZ series drives. This *isn't* to imply that the RA90 could not be used as the basis for a new RF or RZ-type drive. After all, the RA70 generated the base design for the RF series... ================================================================================ Note 295.8 EMC MAXPORT/Multiple Spindle Disk Drives 8 of 10 EISNER::HOPKINS "Al Hopkins" 15 lines 19-FEB-1989 08:36 -< MAXPORT works fine in Connecticut >- -------------------------------------------------------------------------------- | Has anyone had any experience with the MAXPORT series drives, or any | other EMC disk drives? We installed 2 MRA1600's just after Xmas. Not only have there been no problems whatsoever, but also the drives are noticeably faster than our Digital 81's. Installation from truck to fireup took less than 2 hours. We will be adding more soon. Also, we have been using EMC's 14" laser drive for over six months. Other than it being very s-l-o-w (which is okay as it is an archival device), we've not experienced any difficulty with it. All in all, we have been very pleased with EMC, their responsiveness to questions and service, and most of all their product. ================================================================================ Note 295.9 EMC MAXPORT/Multiple Spindle Disk Drives 9 of 10 EISNER::BATES "Ken Bates" 69 lines 25-FEB-1989 11:04 -------------------------------------------------------------------------------- Sorry to take so long to respond, but I haven't been able to log in for the past few weeks. In response to .2: You're absolutely correct, there is no way for the HSC (or the KDA/KDB/UDA) to take advantage of these types of drives. There are two basic ways of presenting a multiple spindle drive to a DSA controller: The first is to spread the logical cylinders shown to the user (and controller) across bands on the individual disks. In this case, a 400 cylinder logical drive would map cylinders 1-100 onto cylinders 1-100 of drive 1, cylinders 101-200 mapped to cylinders 1-100 on drive 2, etc. The second is to spread the logical cylinders evenly across all drives. In this case, drive 1 would contain cylinders 1,5,9,..., drive 2 would contain cylinders 2,6,10,..., etc. The theory behind all of this is twofold: Individual drives seek shorter distances, and parallel operations can occur, since a single spindle is no longer a bottleneck. Since only *one* operation can be outstanding on a single SDI port, no parallelism can be gained by this approach. The four drives are seen as one by the HSC, so all operations are serial. If two requests are received by the HSC which would result in one on drive 1 and one on drive 2, the HSC will completely service one request before even considering the second. In fact, if a single request were to be received by the HSC which would span two drives (a transfer which crosses a cylinder boundary), the performance would be much slower on this arrangement than a single drive due to the internal mapping work and delays required for correct operation. If this is not done "correctly" then the delays could be quite severe. All in all, I believe that the overall performance would be somewhat worse in this area. Seeking should be better since there is less distance to move the heads, but the optimization algorithms in the HSC/UDA/KDA/KDB are unaware of this, and will optimize as if a single drive was present. Overall, this shouldn't hurt performance, but it will effectively negate any possible performance gains. Complicating this is the fact that VMS has a very high locality of reference, resulting in many requests with little or no seek distances (50% of VMS requests are for the same cylinder as the last request!). Because of this, very little advantage is gained by having a smaller number of cylinders per drive. After all, if you only seek 1 cylinder, it's there's no way to reduce that distance! Overall, I suspect that the seek component of response time will be about equal to a single drive. In response to .6: Terry has very effectively summarized the issues. A minor nit with "a)": Only one DSA port can transfer data at a time (a DSA port is synonomous with an SDI cable). With individual drives on individual ports, the probability of two transfers occuring at the same time is (generally) quite small, so 4 drives can co-exist happily with very little contention on a single port. The transfer time is only a small part of the overall response time when you take seeking and rotational latency into account. With all drives on a single port, only one can be active, so this advantage is lost. All in all, I believe that the approach of 4 drives on a single SDI port will result in a fairly significant loss of performance compared to having 4 (indentical) drives spread across 4 individual ports on a heavily loaded VMS system. Naturally, it would be easy to write a benchmark to show the opposite, but that's what benchmarks are for, isn't it? - Ken ================================================================================ Note 295.10 EMC MAXPORT/Multiple Spindle Disk Drives 10 of 10 EISNER::KENNEDY "Terry Kennedy" 21 lines 25-FEB-1989 16:23 -< Still more on disk performance >- -------------------------------------------------------------------------------- > The theory behind all of this is twofold: Individual drives seek shorter > distances, and parallel operations can occur, since a single spindle is > no longer a bottleneck. As you mention, this benefit is reduced by other considerations in such subsystems. Also, short seeks (given VMS's locality of reference) do not really benefit from this, while the difference in long seeks will not be as great. For the others listening in: Modern drives do *not* move the heads at a constant rate - they accelerate and deceler- ate at the start and end of the seek. Thus a seek that's twice as long distance-wise may only take 15% longer time to complete. Also, the info given by the manufacturer leads you to think that they have better seek performance than DEC. This may be true for com- posite disks of the *same size*, but the goal was to create a *bigger* disk. Thus, with all of your data on one 'drive', you will be seeking all over the place (so much for locality of reference). Such 'drives' need to be compared based on the composite spec, not that of the in- dividual components. That information may be hard to come by, as it depends not only on the component drives, but also the vendors 'black box' of control logic.  ================================================================================ Note 296.0 CDC 8650 SNAP analysis tool No replies EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 21 lines 7-FEB-1989 13:22 -------------------------------------------------------------------------------- During the past week I have had a chance to see CDC's DEC hardware support operation working on an interesting problem on an 8650 CPU. One of the more impressive things they had was an interactive SNAP analysis program that they used to look at a number of error SNAP files. They did not elaberate very much on the details of the program, but explained it as an automated way of looking at the various paths and modules that have values in the SNAP file. They claim that the program has knowledge of the 8650 CPU and the expected values and such in the SNAP file. From this they attempt to point out which modules are pointed to by the SNAP. It has a slick interface, comes up an a VT200 with a picture of the 8650, from which you can zoom into various parts of the machine. You get the power supply voltages on each of the power supplies, for example. As you get into an area of the machine it will attempt to call out the cards that may be involved, from most to least likely. They use it by taking a dump of the SNAP file and loading into there machine as a text file. They then massage the file using there program. If they find a new error scenario they then load the new knowledge into the program.  ================================================================================ Note 297.0 11/84'S AND TU CLASS TAPES 1 reply EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 4 lines 7-FEB-1989 14:05 -------------------------------------------------------------------------------- There were some ECO's that dealt with 11/84's and TU class tape drives. Does anyone know the numbers? ================================================================================ Note 297.1 11/84'S AND TU CLASS TAPES 1 of 1 EISNER::KENNEDY "Terry Kennedy" 6 lines 7-FEB-1989 19:12 -< Of course... >- -------------------------------------------------------------------------------- > Does anyone know the numbers? It's EQ-01397-01, needed when an 11/84 UBA (M8191) is at CS 'B'. This is a simple rewire - if you're handy with a soldering iron I'll pull the info and post it here.  ================================================================================ Note 298.0 200 (or other non-standard) Baud rates? 6 replies EISNER::LEDERMAN "Bart Z. Lederman" 7 lines 7-FEB-1989 19:14 -------------------------------------------------------------------------------- Is anyone running an async port at 200 Baud (or another non-standard speed) on a VAX? We have an application where we may have to hook up to such a line. DEC hardware with an external clock, third-party hardware, a Terminal server, or a third-party terminal server whould all be acceptable solutions. (If it can't be easily done we can use external "black boxes" to convert to a standard baud rate, but we'd rather do it in the system if it's practical.) ================================================================================ Note 298.1 200 (or other non-standard) Baud rates? 1 of 6 EISNER::KENNEDY "Terry Kennedy" 21 lines 7-FEB-1989 19:35 -< Possible, yes; fun, no >- -------------------------------------------------------------------------------- > Is anyone running an async port at 200 Baud (or another non-standard > speed) on a VAX? No, but here's how: 1) For boards which accept external clock, apply the desired baud rate x16 on the external clock pin, as follows: AY5-8116 baud rate generator -> EIA 1488 driver -> Ext Clk pin Both the 8116 and the 1488 are avilable at Radio Shack stores. The 8116 is a programmable baud rate generator. You'll probably need an oddball crystal, see the data sheet for the 8116 for more info. 2) On the boards which don't have an external input, you can do the same trick, but couple the 8116 output directly to the AY5-1013 or AY3- 1015 UART. 3) If extreme accuracy isn't needed, a 555 timer can be used for infinitely variable baud rates. Put the output into a Schmitt trigger to square it up before passing it to the 1488 driver. ================================================================================ Note 298.2 200 (or other non-standard) Baud rates? 2 of 6 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 13 lines 7-FEB-1989 20:59 -< Here's one answer. May be more than you want to spend. >- -------------------------------------------------------------------------------- What bus are you dealing with? Unibus, Qbus, or VAXBI? We (Simpact Associates, San Diego CA, 619-565-1865) have an async interface product (hardware/software combo) that will do this for the latter two. It has a moderately versatile capability to understand start- and end-of- message sequences, which may help offload your VAX, though of course even single-character $QIOs shouldn't tax a VAX too much at 200 bps. Four ports per board. Nothing for Unibus at present. (We have the Unibus board but the async software hasn't been ported to it yet.) DISCLAIMER: I work for Simpact. For more details, contact our sales department. ================================================================================ Note 298.3 200 (or other non-standard) Baud rates? 3 of 6 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 17 lines 7-FEB-1989 21:12 -< Try a black box (literally) >- -------------------------------------------------------------------------------- Another possibility, much cheaper than .2: Buy one of Black Box's "Communication Adapter Plus" units ($395). "...can perform data rate conversion for any of 21 different speeds, ranging from 45.5 bps to 19.2 Kbps. Just set the DIP switches for each port individually, and the CAP automatically communicates with that port at the appropriate baud rate." Bring your data into a VAX port set to 300 bps (or any other supported speed that's greater than 200). If you had to transmit data as well you'd have all kinds of handshaking/overrun problems, but this setup should work just fine. Black Box: 412-746-5530 for orders, 412-746-5565 for technical questions (like to ask whether the beast really supports 200 bps). Disclaimer: I have no connection with Black Box, other than that my company is a satisfied customer. ================================================================================ Note 298.4 200 (or other non-standard) Baud rates? 4 of 6 EISNER::LEDERMAN "Bart Z. Lederman" 18 lines 8-FEB-1989 07:53 -< Data will be incomming only >- -------------------------------------------------------------------------------- I should have said that the data is going to be one-way only (into the VAX). The Black Box is our alternate solution: but we need several lines, and at $395 per line in addition to having an Async interface on the VAX, we were looking to see if the VAX could do it by itself. The question of an external clock is not so much how to get the clock (we can build a good generator), it's whether or not you CAN feed an external clock into a DEC Async device. The original DH11 had provision for an external clock, but I do not know if any current hardware has a provision for turning off the internal clock and using an external one. Modifying DEC boards is not a viable alternative for us as it would make the system difficult to maintain. Our immediate need is Unibus, but in future will be Q-bus. Thanks for the information so far. ================================================================================ Note 298.5 200 (or other non-standard) Baud rates? 5 of 6 EISNER::LEDERMAN "Bart Z. Lederman" 8 lines 8-FEB-1989 13:57 -< Xyplex MAXserver might be an answer. >- -------------------------------------------------------------------------------- There is a DEXPO in New York City at the moment. I talked to some people from Xyplex, who make the "MAXserver", a terminal server like device which does LAT and TCP/IP. Their technical person tells me that their unit has a setting for 200 Baud, though he doubts if anyone has ever actually used it. A question which should probably be in a separate topic (if anyone has the answer): has anyone used a Xyplex MAXserver? ================================================================================ Note 298.6 200 (or other non-standard) Baud rates? 6 of 6 EISNER::KENNEDY "Terry Kennedy" 9 lines 8-FEB-1989 20:27 -< Suggestion, take two... >- -------------------------------------------------------------------------------- > using an external one. Modifying DEC boards is not a viable > alternative for us as it would make the system difficult to maintain. Well, the DEC DZ-11 module uses the 8116 baud rate generator, so your *only* mod would be to change one crystal on the board. That would affect *all* 8 lines, but from what you say, that would be a plus. DZ11's are so inexpensive right now, you could just get an extra board to use for a spare. Since you won't have wires all over the place using this method, DEC may not have a problem with it.  ================================================================================ Note 299.0 VAXStation-2000 powers itself off 7 replies EISNER::NAGY "Frank J. Nagy, VAX Guru" 20 lines 8-FEB-1989 08:50 -------------------------------------------------------------------------------- Another one of "Has anyone else seen this behavior?" requests: We've had this come and go on various VAXStation-2000s. One in particular is causing problems now. The configuration is a VAXStation-2000, monochrome system, 6 MB memory and a Seagate ST251-1 (RD32) local disk. Thinwire is being used. The problem is that the VS-2000 shuts itself off. The power switch is ON but the system has powered down (no fan noise). The solution is to turn the system OFF, wait a few seconds and then turn the system back ON. We call this "going to sleep". This behavior seems to happen only to an idle system (user not actively typing, mousing or even viewing) and only (right now) to a specific system. We have had similar problems with other systems in the past, but after sufficient playing around by FS the problem went away. In one case, FS claimed that the system was powering itself down every time the person goa a phone call and their answering machine started (because both were plugged into same wall fixture). Anyone else seen this problem? Anyone fixed it? ================================================================================ Note 299.1 VAXStation-2000 powers itself off 1 of 7 EISNER::MCCORMICK_B "Bob McCormick" 10 lines 8-FEB-1989 19:01 -< Yep I've seen it! >- -------------------------------------------------------------------------------- Yep! I've seen the exact same thing... but on an old MicroVAX 2000 (non workstation) with the Ethernet and internal RD54 + external TK50Z. 6mb memory and no sync/async board. Only thing I could figure was powersupply shutdown... but couldn't reproduce it easily to show DEC field service rep! The system still is running ... but once in a while it dies (off). Most often shortly after a power up after a shutdown. ================================================================================ Note 299.2 VAXStation-2000 powers itself off 2 of 7 EISNER::KENNEDY "Terry Kennedy" 9 lines 8-FEB-1989 20:53 -< Cheap ASTEC power supply >- -------------------------------------------------------------------------------- It's probably an Astec power supply. They are so twitchy it's not funny. SOmetimes DEC has a second vendor. Try to get one of those instead (if avail- able). It's easy to tell - when FS has the machine open, look at the power supply. It should have something like the following ailk-screened on the power supply circuit board: A S T E C /\/\/\/\/\/\ ================================================================================ Note 299.3 VAXStation-2000 powers itself off 3 of 7 EISNER::STRIEGEL "Alan Striegel" 9 lines 19-FEB-1989 17:54 -< Saw one similar instance >- -------------------------------------------------------------------------------- I saw a somewhat similar occurrence here, but only one time. MicroVAX 2000, 6MB, RD54, External TK50, Ethernet board. Something caused the entire building's power to go off and then right back on. The MicroVAX 2000 just sat there with its power switch on, but no sound or signs of life. I shut the thing off and turned it back on a few seconds later and still no good. Waited a minute and turned it on again, all was normal. I assumed the power supply was protecting itself. Alan ================================================================================ Note 299.4 VAXStation-2000 powers itself off 4 of 7 EISNER::HARVEY "Jack Harvey" 6 lines 20-FEB-1989 03:57 -< Seen Frequently >- -------------------------------------------------------------------------------- MV and VS 2000's often pass across my desk on their way to customers who think these are expensive PC's and would be indignant if they had to read a manual to learn how to load the software. The first time it happened, I got a bit upset, but couldn't make it repeat. Customers never complain. I guess they expect PC's to behave that way. ================================================================================ Note 299.5 VAXStation-2000 powers itself off 5 of 7 EISNER::WATERMAN "Doug Waterman (609)423-1040" 15 lines 23-FEB-1989 09:56 -------------------------------------------------------------------------------- -- My VS2000 (RD54, 8MB Clearpoint) was doing the same thing daily. It turned out that it was the thermal protection on the power supply tripping. Field Service replaced the PS and now everything is fine. Oh, I also mookved the system unit so that it gets better airflow. The power supplies in these things are known to be prone to this type of behavior. Apparently, diskless models don't seem to have the overheat problem as much. As soon as you stick a disk in it however, the supplies get hot. I started out with a diskless station and later added a Maxtor drive. That's when the problem began for me. ================================================================================ Note 299.6 VAXStation-2000 powers itself off 6 of 7 EISNER::KOZAM 9 lines 25-FEB-1989 00:41 -< Intermittent Power Failure >- -------------------------------------------------------------------------------- Add a VAXstation2000 with RD53 to the list. In this case, the power went out for a short time (< 1 second) during a thunderstorm. About 10 minutes afterward, I noticed that the VAXstation was still sitting there, dead. (No, I wan't using it at the time!) Having read DECUServe, I said Ah!, I'll just flick the power switch off then on again. Nothing happened, even after I repeated the off-on cycle a few times. After letting it sit off for about a minute, I switched it back to on, and everything has been OK every since. ================================================================================ Note 299.7 VAXStation-2000 powers itself off 7 of 7 EISNER::MAYHEW "Bill Mayhew" 15 lines 25-JUN-1989 23:01 -< Does this sound related? >- -------------------------------------------------------------------------------- I've had a couple instances this weekend of... VS2000, 4Mb, RD32/RX33, Ethernet... just sitting there, "idle". There's one open window (VMS 4.7/VWS 3.1) on the screen, and a user is logged in to the system via LAT, but nothing's going on. All of a sudden the raster goes away on the VR262 (i.e. shrinks to a point); less than a second later it comes back. The LED on the VR262 never went out, so its power presumably remained intact (though this is not totally clear as the LED seems to take 1-2 sec to turn off when you cycle the VR262 power switch). When the raster reappears, everything *looks* normal -- but you can't DO anything. No keyboard response. The LAT user got a "disconnect". *Even the HALT switch doesn't produce any (visible) effect.* Power down/up seems to bring it back.  ================================================================================ Note 300.0 VAX SIDs 1 reply EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 86 lines 8-FEB-1989 09:03 -------------------------------------------------------------------------------- Joe sent me a mail message asking about SID's. I sent him VMS topics 300 and 323. As you can see below he is looking for more info. He is writting an application that will look at SIDs. Any responses will be forward back to him. Thanks in advance D E C U S - I N T E R O F F I C E M E M O R A N D U M Date: 7-Feb-1989 11:36pm EST From: Joe H. Gallagher GALLAGHER Dept: DTR/4GL SIG Tel No: 816-276-4235 (work) TO: Jeff Killeen ( KILLEEN ) Subject: RE: FROM YOUR FRIENDS AT DECUServe Thanks every so much for friends at DECUServe. The information was quite interesting and very close to what I was after, but not quite. What I had in mind was something like: The SID is a 32-bit integer. The first byte (bits 31:24) are the processor type. Processor Other Type bits Meaning 0 Reserved to Digital 1 VAX-11/780 and VAX-11/785 Bits 23:22 1--> 780; 0-->785. Bits 21:18 hardware CPU rev level Bits 17:15 system rev level Bits 14:12 manufacturing site Bits 11:00 serial number 1 VAX-11/782 Bits 23:16 CPU hardware rev level Bits 15:12 manufacturing site Bits 11:00 serial number 2 VAX-11/750 Bits 23:16 not used Bits 15:08 CPU microcode software rev level Bits 07:00 CPU hardware rev level 3 VAX-11/730 and VAX-11/725 Bits 23:16 not used Bits 15:08 CPU microcode software rev level Bits 07:00 CPU hardware rev level 4 VAX 8600/8650 Bits 23:16 CPU hardware rev level Bits 15:12 manufacturing site Bits 11:00 serial number 5 VAX 8300/8200 Bits 23:19 processor revision code Bits 18:09 microcode patch rev level Bits 08:08 secondary patch bit Bits 07:00 control store ROM/RAM rev level 6 VAX 8800/8700 (Nautilus) Bits 23:23 left/right processor Bits 22:16 processor revision Bits 15:00 serial number ? VAX 8550/8500 Bits 23:23 left/right processor Bits 22:16 processor revision Bits 15:00 serial number I want to know which model numbers fix into which processor types and whether a part of the SID is a serial number. What I have here is all the old stuff. What I need is the corresponding information for all the new processors. I guess I need to look in the latest VAX Hardware Handbooks, but my latest ones are 1986 vintage (from which I got most of this information). I need the 1989 or 1990 one! Any help there? ================================================================================ Note 300.1 VAX SIDs 1 of 1 EISNER::KENNEDY "Terry Kennedy" 3 lines 8-FEB-1989 20:57 -< Consider uV II, look at LMF >- -------------------------------------------------------------------------------- Remember that some CPU's (like the MicroVax II) don't have a unique serial number in the SID. If I were Joe, I'd talk to DEC about LMF and not try to do it myself.  ================================================================================ Note 301.0 RQDX3 REV D1 VS. REV D2 WITH -DA OR -RA DISKS 10 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 37 lines 12-FEB-1989 00:04 -------------------------------------------------------------------------------- This is more of an FYI item than a question........ We have several systems setup as follows.....  ------------------------- | BA11 CPU BOX | | PDP-11/53 or 73 | | RQDX3---------! | ----------------!-------- ----------------- ! | RX50 DESK TOP |<--------------! ! <- CONNECTED TO J1 ON THE RQDX3 _________________ ! ! ! ! <- CONNECTED TO J1 ON THE RD52-DA CONNECTED TO J1 ON THE RX50-DA->! ! ! ! CONNECTED TO J3 ON THE RD52-DA->! ! ! v ----------------- | RD52 DESK TOP | ----------------- This is a legal configuration and works fine with any RQDX2 controller. It also works fine with any RQDX3 REV C or REV D1 controller. It will not work with a RQDX3 REV D2 controller. Both RQDX3 controllers have the current 243/244 firmware. The REV D1 will access both drives with no problem. The REV D2 will only access DU1 if DU0 (RD52) is turned off - and will only access DU0 if the RX50 if turned off. If both drives are on the subsystem hangs. Is there anyway to "SPR" hardware that is not under contract? BTW - the REV D2 has two undocumented jumpers right next to the firware proms that the old revs do not have. The REV D1 boards have some wire jumpers attached to chip pins that the REV D2 boards don't have. ================================================================================ Note 301.1 RQDX3 REV D1 VS. REV D2 WITH -DA OR -RA DISKS 1 of 10 EISNER::KENNEDY "Terry Kennedy" 12 lines 12-FEB-1989 06:59 -< According to DEC numbering, this should not happen >- -------------------------------------------------------------------------------- > BTW - the REV D2 has two undocumented jumpers right next to the firware > proms that the old revs do not have. The REV D1 boards have some wire > jumpers attached to chip pins that the REV D2 boards don't have. According to DEC numerology rules: CS D1 vs. CS D2 should not have any changes to the schematic (ie., *no* electrical differences). D1/D2 would indicate *either* the etch level of the base module is different (Mxxxx 50xxxxxA vs. Mxxxx 50xxxxxB, for ex- ample), *or* that the firmware is different. I'll research it in the fiche... ================================================================================ Note 301.2 RQDX3 REV D1 VS. REV D2 WITH -DA OR -RA DISKS 2 of 10 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 9 lines 12-FEB-1989 08:33 -< OH..... >- -------------------------------------------------------------------------------- > According to DEC numerology rules: > > CS D1 vs. CS D2 should not have any changes to the schematic (ie., *no* >electrical differences). D1/D2 would indicate *either* the etch level of First I assume the REV numbers are the numbers press'ed into the module handles. The REV D1 boards are reworked etch "C" boards. The REV D2 boards are etch "D" boards. ================================================================================ Note 301.3 RQDX3 REV D1 VS. REV D2 WITH -DA OR -RA DISKS 3 of 10 EISNER::KENNEDY "Terry Kennedy" 25 lines 13-FEB-1989 00:01 -< Long-winded explanation, theories >- -------------------------------------------------------------------------------- > First I assume the REV numbers are the numbers press'ed into the module > handles. Right. This is more appropriately (at least to DEC) the 'CS level' or'CS Rev level'. It is punched into the module bracket (handle) on the non-com- ponent (etch) siode of the board. Usually there is a 3-digit number before the CS info, such as '736 E1'. Boards reworked by CRC's (Customer Returns Center, although they also handle DEC FS work) will have the old rev X-d out and a new one punched in. DEC FS depot rework will *either* X out the old rev with black pen and rubber-stamp a new level (also in black) *or* use Brady wrap markers on the handle. By the way, boards which have been through CRC lately have a sticker with *lots* of small print near the edge of the board just below the handle. It indicates the problem reason/ repair code, or an X in 'NPF' if there was No Problem Found. > The REV D1 boards are reworked etch "C" boards. The REV D2 > boards are etch "D" boards. Then, they are *supposed* to be identical. This doesn't always happen, due to complications. I don't know how you would report such a thing, though. If you had a new system configured/sold that way, you could not accept it until it worked, for example. I'm afraid I don't know of a less drastic method. I'll keep an eye out on the fiche updates... P.S. - Did you know that Mxxxx means Magenta handles? G=Green, etc... ================================================================================ Note 301.4 RQDX3 REV D1 VS. REV D2 WITH -DA OR -RA DISKS 4 of 10 EISNER::MCCORMICK_B "Bob McCormick" 5 lines 13-FEB-1989 07:42 -< What's the "T" for? >- -------------------------------------------------------------------------------- | P.S. - Did you know that Mxxxx means Magenta handles? G=Green, etc... Yes! But what does Txxxx mean? (8000 & 6000 for those who don't know...) ================================================================================ Note 301.5 RQDX3 REV D1 VS. REV D2 WITH -DA OR -RA DISKS 5 of 10 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 13 lines 13-FEB-1989 19:13 -< Do I win? >- -------------------------------------------------------------------------------- > But what does Txxxx mean? (8000 & 6000 for those who don't know...) > Take out one of those modules. Near the fingers (connectors) is a T shaped registration hole.  Yeah, I know that's probably not really why, but it popped into my head. Next question: What does the L in Lxxxx (as in HSC, CI, 11750) modules stand for? ================================================================================ Note 301.6 RQDX3 REV D1 VS. REV D2 WITH -DA OR -RA DISKS 6 of 10 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 18 lines 20-JUL-1989 13:37 -< What is the current RQDX3 revision? >- -------------------------------------------------------------------------------- >> The REV D1 boards are reworked etch "C" boards. The REV D2 >> boards are etch "D" boards. I have an RQDX3 stamped on the handle as "C1" and with the following ROMs: DEC86 DEC86 LM8709 LM8705 244E5 243E5 The diagnostics report that it has "Revisions = 2 and 1". The question is, what is necessary to upgrade it to the current revision (and what *is* the current rev)? I gather that there are some jumpers and new ROMs. Can anyone give me the FCO information so I can update the board myself, or must I return the board to DEC. I can copy ROMs with no problems. The board works OK, except for not being able to format RX33 floppies. ================================================================================ Note 301.7 RQDX3 REV D1 VS. REV D2 WITH -DA OR -RA DISKS 7 of 10 EISNER::KENNEDY "Terry Kennedy" 14 lines 20-JUL-1989 16:25 -< RQDX3-F002 >- -------------------------------------------------------------------------------- > The question is, what is necessary to upgrade it to the current > revision (and what *is* the current rev)? I gather that there > are some jumpers and new ROMs. Can anyone give me the FCO > information so I can update the board myself, or must I return > the board to DEC. You need RQDX3-F002 (April '89), whicch contains the followwing EPROMS: 23-339E5-00 23-340E5-00 It is orderable as kit EQ-01532-02. This fixes *many* problems with the RQDX3, most of which were fixed in various 'internal' revisions which were *not* releazsed to customers. ================================================================================ Note 301.8 RQDX3 REV D1 VS. REV D2 WITH -DA OR -RA DISKS 8 of 10 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 21 lines 21-JUL-1989 11:51 -< Can't find FCO kit on E-Store >- -------------------------------------------------------------------------------- >> You need RQDX3-F002 (April '89), which contains the following EPROMS: >> >> 23-339E5-00 >> 23-340E5-00 >> >> It is orderable as kit EQ-01532-02. This fixes *many* problems with the >> RQDX3, most of which were fixed in various 'internal' revisions which were >> *not* released to customers. Does this kit also include the instructions for the various cut-and-jumper mods I assume are necessary? Since I have a rev C-1 board, do I need any intermediate kits? I just checked the Electronic Store and EQ-01532-* was not listed, nor could I find any RQDX kits at all. Is this one of those kits you have to know someone to get? On the other hand, EQ-01523-01 "VMB Patches Kit" sounds interesting. I wonder what this FCO is for? ================================================================================ Note 301.9 RQDX3 REV D1 VS. REV D2 WITH -DA OR -RA DISKS 9 of 10 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 10 lines 21-JUL-1989 13:17 -< Put away your X-acto knife... >- -------------------------------------------------------------------------------- < Note 301.8 by EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" > >Does this kit also include the instructions for the various >cut-and-jumper mods I assume are necessary? Just having upgraded an RQDX3 to V4 firmware, there was no cut-and-jumper'ing required, just prom-swapping. Not 100% sure if my RQDX was at C1, though... Ed ================================================================================ Note 301.10 RQDX3 REV D1 VS. REV D2 WITH -DA OR -RA DISKS 10 of 10 EISNER::KENNEDY "Terry Kennedy" 13 lines 21-JUL-1989 20:25 -< Unlisted parts >- -------------------------------------------------------------------------------- > Does this kit also include the instructions for the various > cut-and-jumper mods I assume are necessary? Shouldn't be any hacks - the problems were all firmware (at least for this set of fixes). The kit comes with an FCO document listing any prerequisite kits as well. > I just checked the Electronic Store and EQ-01532-* was not > listed, nor could I find any RQDX kits at all. Is this > one of those kits you have to know someone to get? Can you say "unlisted parts"? Or will Sharon wash your mouth out with soap if you do?  ================================================================================ Note 302.0 TAPE MARKS ON TSV05 14 replies EISNER::LIEBERMAN 5 lines 14-FEB-1989 19:07 -------------------------------------------------------------------------------- We are trying to create a tape on a TS11 for reading on an IBM system. The recipient tells us that we are starting the tape with a tape mark. We read the tape using DUMP and ETAPE and can't find the tape mark that seems to be the very first thing on the tape. We also tried to write on a TSV05 with the same result. Any ideas? ================================================================================ Note 302.1 TAPE MARKS ON TSV05 1 of 14 EISNER::KENNEDY "Terry Kennedy" 28 lines 15-FEB-1989 04:45 -< QUestions, solutions for DOS/VSE and VM/SP >- -------------------------------------------------------------------------------- > We are trying to create a tape on a TS11 for reading on an IBM system. > The recipient tells us that we are starting the tape with a tape mark. What utility are you writing the tape with? Under what operating system? I presume you are writing it out in EBCDIC for them. By the way, this should not be a problem for the IBM side. Here are some simple solutions: For DOS/VSE: 1) Mount your tape and a scratch tape on 2 tape drives (lets call the drives 280 and 281, your tape on 280 and the scratch on 281) 2) DITTO FSF,280,1 3) DITTO TT,280,281,99 4) DTTTO WTM,281,2 5) DITTO RUN,280 6) DITTO REW,281 7) This leaves a copy of your tape with the leading tape mark removed on unit 281. For VM/CMS: 1) Change the FILEDEF from: FILEDEF INMOVE TAP1 NL 1 .... to FILEDEF INMOVE TAP1 NL 2 .... ================================================================================ Note 302.2 TAPE MARKS ON TSV05 2 of 14 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 21 lines 16-FEB-1989 14:38 -< Ya wanna look at them little bits on the tape? >- -------------------------------------------------------------------------------- >> We read the tape using DUMP and ETAPE and can't find the tape mark that >> seems to be the very first thing on the tape. We also tried to write >> on a TSV05 with the same result. Any ideas? There are several products that will let you actually see the bits on the tape. I have used them for years in solving these problems and highly recommend them. They consist of finely powdered iron in Freon. You either dip the tape in the liquid or spray it on, let it dry and view the bits! The stuff is available at better hi-fi stores, professional sound equipment suppliers and some electroncs supply companies. The two I have here are: Magna-See by Soundcraft, a division of Columbia Broadcasting System, Inc., Danbury, CT 06810. This comes as a liquid in a half-pint can. Mag View by Nortronics Company, Inc., 8101 Tenth Avenue North, Minneapolis, MN 55427 (612) 545-0401 This comes in a 4 oz spray can. ================================================================================ Note 302.3 TAPE MARKS ON TSV05 3 of 14 EISNER::WICHMAN "Jon Wichman" 6 lines 16-FEB-1989 15:25 -< You only do that once I take it? >- -------------------------------------------------------------------------------- > They consist of finely powdered iron in Freon. You either dip the > tape in the liquid or spray it on, let it dry and view the bits! That don't hurt the tape drive any next time you run it through - do it? :-) ================================================================================ Note 302.4 TAPE MARKS ON TSV05 4 of 14 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 21 lines 16-FEB-1989 16:00 -< tape 'SEE' powder comes off >- -------------------------------------------------------------------------------- > That don't hurt the tape drive any next time you run it through > - do it? :-) It cleans off easily, if necessary. Redunk and wipe quickly rather than let dry, or try a TEX-WIPE. Someone also used to make a little glass or plastic box with the powder in it and you held the tape against the back, shook the box, and then looked through the face at your bits. When the box was moved away from the tape, the 'bits' fall into a pile inside the box. Check out the back of your credit cards, too. High density and low density stripes. Your NUMBER they MUST read, your NAME, well thats nice, but optional. I believe different groups 'own' the different track standards, and ABA and IATA both come to mind, but someone who KNOWS should comment. Also check the inter-record gap on your tapes. We once has a TU10 that always wrote a 2" one and we wondered why its tapes held less. F/S said thats OK, its in spec and ANYONE can READ it. We did get it fixed. ================================================================================ Note 302.5 TAPE MARKS ON TSV05 5 of 14 EISNER::PROVOST 6 lines 17-FEB-1989 08:56 -< Tape Viewer >- -------------------------------------------------------------------------------- 3M makes a "tape viewer" which allows you to "read" the bits without affecting the tape. The metal is suspended in some liquid. When the device is placed on the tape, you can see the bits. I have found these quite useful over the years. They are reasonably costly and die over time. I don't currently have one that works or I could give you more information. ================================================================================ Note 302.6 TAPE MARKS ON TSV05 6 of 14 EISNER::BURNET "John Burnet" 13 lines 18-FEB-1989 18:11 -< Is this really feasible? (Enlighten me...) >- -------------------------------------------------------------------------------- > There are several products that will let you actually see > the bits on the tape. I have used them for years in solving > these problems and highly recommend them. I always wondered about this, and since I have never used these products, this may be a dumb question: Even after the bits are visible, how does one accurately decipher a half-square-inch area containing over 56,000 bits (6250 x 9)? With a microscope, pencil and paper, and the patience of ten saints? Granted that to look at the first (say) 1K bytes on the tape, you don't need to decode very much of the length (1K approx. == 1/6")... still and all, those things are *tiny*! ================================================================================ Note 302.7 TAPE MARKS ON TSV05 7 of 14 EISNER::HARVEY "Jack Harvey" 32 lines 19-FEB-1989 02:01 -< What Can be Seen >- -------------------------------------------------------------------------------- > -< Is this really feasible? (Enlighten me...) >- Can't fool you! You're right, you certainly cannot see the individual bits. What you can see, however, are areas of the tape that have flux reversals (which attract the powder) versus areas that have no data (which are relatively powder free). For 1600 BPI and up, the only areas without flux reversals are the gaps between blocks, the gap between the BOT foil and the first block, and the "tape marks" or EOF marks. But just knowing where these are can frequently spot a problem. For example, if the inter-block gaps are too short, or there is *not* a long gap between BOT and the first block, etc. Measuring the actual length of the data blocks and comparing it with the calculated length for the density and format in use can disclose a software (or hardware) problem. (Of course, now we should explain all the formats...) Sometimes it's useful to visually compare a known good tape of the same format with a problem tape. It's especially important to check known good tapes when you first try one of these products, so you are not misled by the streaks and blobs that occur with the dip and spray schemes. And be aware that streaming drives, such as the TU80 and TU81 can sometimes write a much longer inter-block gap than the standard 0.6 inch gap written (most of the time) by start-stop drives like the TU77 or TU78. (This is legal and compatible.) ================================================================================ Note 302.8 TAPE MARKS ON TSV05 8 of 14 EISNER::BURNET "John Burnet" 6 lines 19-FEB-1989 14:22 -< Very interesting... >- -------------------------------------------------------------------------------- The reason I asked was that I had heard of people using Magna-see to actually decode the bits on the mag stripes on credit cards, as well as (years and years ago) information stored on the original HP-65 magnetic program-storage cards, because HP wouldn't publish the format. (Remember the HP-65? Wonder who's got Steve Wozniak's now...) I guess the density on those things is a lot lower than 6250 bpi. ================================================================================ Note 302.9 TAPE MARKS ON TSV05 9 of 14 EISNER::HARVEY "Jack Harvey" 12 lines 20-FEB-1989 03:41 -< Low Densities Readable >- -------------------------------------------------------------------------------- As I recall, at 800 BPI, one could *almost* make out individual bits. Of course, that was not a Manchester format and a track with all zeros had no flux reversals within a block and would show clear. You could thus distinguish between binary areas and ASCII areas, especially because in 800 BPI days, lower case was rare. At 200 BPI on the seven track tapes, it probably could have been read, but I never did it. I would expect that credit cards and the HP-65 calculator program cards would have to be very low density because of poor contact between the stripes and the read heads. ================================================================================ Note 302.10 TAPE MARKS ON TSV05 10 of 14 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 4 lines 21-FEB-1989 14:03 -< shorter gaps at 6250 >- -------------------------------------------------------------------------------- > standard 0.6 inch gap written (most of the time) by 0.3 is legal in GCR mode. ================================================================================ Note 302.11 TAPE MARKS ON TSV05 11 of 14 EISNER::HARVEY "Jack Harvey" 24 lines 22-FEB-1989 21:25 -< GCR 6250 Details? >- -------------------------------------------------------------------------------- > 0.3 is legal in GCR mode. I must confess ignorance of the gross physical characteristics of GCR tapes. Could you expound a bit on the visible differences that would show with the tape "developer"? 0.3 startles me because that doubles the capstan acceleration required for a given tape speed on a start-stop drive like the TU78. But it makes sense when a big objective of 6250 is, of course, more data per reel. Are long gaps (i.e., backup and erase a long gap when read-after-write detects an error) still 3 inches? Is there a mandatory long gap at BOT? What does a Tape Mark look like? Is any block structure, such as a preamble visible? Given a data block of 3 X 6250 = 18,750 bytes, how long should the physical block on the tape be? Obviously, if there was no pre/postamble, and no error correction bytes, it would be exactly 3 inches. What kind of overhead does the GCR add? ================================================================================ Note 302.12 TAPE MARKS ON TSV05 12 of 14 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 49 lines 22-FEB-1989 23:16 -< a few GCR details >- -------------------------------------------------------------------------------- > I must confess ignorance of the gross physical characteristics > of GCR tapes. Could you expound a bit on the visible > differences that would show with the tape "developer"? > Is there a mandatory long gap at BOT? What does a Tape Mark > look like? Is any block structure, such as a preamble > visible? At BOT there is an ID burst OVER the section with the BOT TAB, but for GCR it is in a different channel (on 1 instead of P, as used in PE mode). (With the oxide UP, BOT on your left, data extending to your right, the channels from NEAR edge to FAR are: 5,7,3,P,2,1,0,6,4, but are also called 1-9 as tape tracks) The ARA (Automatic Raed Amplification Burst follows at no specific distance from the ID burst, BUT will start between 1.5" and 4.3" after the leading edge of the bot marker, and ends between 9.5" and 11.5" after the same point. At the end of this ALL 1s in ALL tracks burst there is an ID character approximately 2" long and is ones in channels 7,3,2,1,6,4, and erasure in 5,P,0. There is then a NORMAL interblock gap to the first data block. EOF marks are 250-400 flux reversals recorded in channels 5,7,P,2,0,6, and erasure in 3,1,4. This is at a rate of 9042 FRPI. PE has a similar EOF, but at 3200 FRPI and writing 64-256 reversals. But PE's channels MAY be the same, but actually channels 5,P,0 can, in any combination, be written or erased, while the others are the same written or not as in GCR. EOF > Given a data block of 3 X 6250 = 18,750 bytes, how long > should the physical block on the tape be? Obviously, > if there was no pre/postamble, and no error correction > bytes, it would be exactly 3 inches. What kind of overhead > does the GCR add? GCR writes raw bits at 9042 per inch, BUT has all sorts of data groups, and sub groups and resync bursts, and preambles, and end mark and residual, and CRC, postamble (preamble for read reverse), etc. There are LOTS of extra bits written to enable recovery, but the net usable data is 6250 CPI when the overhead is factored in. Sorry, it is a bit too much to type ALL the grubby details in. If someone is desperate, and doesn't have this anywhere, I could be convinced to FAX a few pages of it, but poke around in those manuals YOU OWN, but that F/S keeps wandering off with. I am reading in the general area of pps 4-24 - 4-34 of Pertec's Model T1940-96 Operating and Service Manual No. 107930 (called TU78 by DEC). ================================================================================ Note 302.13 TAPE MARKS ON TSV05 13 of 14 EISNER::HARVEY "Jack Harvey" 9 lines 23-FEB-1989 05:17 -< I'll Look at a GCR Tape >- -------------------------------------------------------------------------------- Got it. It isn't practical to predict the exact physical length of a GCR block given the data byte count, when standing in front of the drive with a piece of tape dripping developer solvent. Thanks, Barton, for all the other info, however. First chance I get, I'm going to develop a GCR tape with some known stuff on it, including tape marks. The erased tracks should be very obvious. ================================================================================ Note 302.14 TAPE MARKS ON TSV05 14 of 14 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 44 lines 23-FEB-1989 13:22 -< Here are the ANSI mag tape standards >- -------------------------------------------------------------------------------- >> Sorry, it is a bit too much to type ALL the grubby details in. If someone >> is desperate, and doesn't have this anywhere, I could be convinced to >> FAX a few pages of it, but poke around in those manuals YOU OWN, but that >> F/S keeps wandering off with. I am reading in the general area of pps >> 4-24 - 4-34 of Pertec's Model T1940-96 Operating and Service Manual >> No. 107930 (called TU78 by DEC). You might also order the ANSI standard: X3.54-1976 "Recorded Magnetic Tape for Information Interchange (6250 CPI, Group Ceded Recording)" (The "-1976" in the above number may be a later year if the standard has been updated since then. I don't think it has.) Order from: American National Standards Institute, Inc. 1430 Broadway New York, NY 10018 I don't know what the current price is, but my 1980 catalog lists it at $5.50. I doubt it has more tripled in price. I suggest that you might also want to order: X3.27-1978 "Magnetic Tape Labels and File Structure for Information Interchange" X3.39-1973 "Recorded Magnetic Tape for Information Interchange (1600 CPI, PE) X3.22-1973 "Recorded Magnetic Tape for Information Interchange (800 CPI, NRZI) X3.40-1976 "Unrecorded Magnetic Tape for Ibnformation Interchange" (This one defines the size of the hub, location of the BOT/EOT markers, etc.) I'm sorry, but I don't have the one for 3200 CPI tapes. ANSI can certainly tell you. When you order the above standards, be sure to ask for the current revision, if the one I listed is not the latest.  ================================================================================ Note 303.0 11/53 prom help needed 7 replies EISNER::BRUCE "Barton F. Bruce - CCA/MA" 18 lines 14-FEB-1989 21:36 -------------------------------------------------------------------------------- Does anyone know anything about the proms on an 11/53? They appear to be LARGE proms, too large to fit in the I/O page, and so probably get their contents moved to low memory for execution. Special systems based on 11/53s (e.g. DS550) apparently get shipped with extensive diagnostics built in that get run before any booting over Ethernet is tried. There are 2, labeled low byte, and high byte, and I assume must be like those on an 11/84. The 11/84 manual I am digging through clearly does not expect the users to blast their own proms, and simply treats them as part of the board you get. Is the format of these proms documented in some manual, or has someone figured it out. Any clues will be welcome. ================================================================================ Note 303.1 11/53 prom help needed 1 of 7 EISNER::KENNEDY "Terry Kennedy" 8 lines 15-FEB-1989 04:55 -< >- -------------------------------------------------------------------------------- > Any clues will be welcome. I don't have a '53, but I think the PROMS are segmented so that they can be overlaid in the I/O page. The '23 uses a pair of 68766 (8Kb) PROMS. How big (or what part number) are the PROMS in the '53? A pos- sible help would be the manual for the boot card the 11/73 dual board uses (MXV11-B something, I think). ================================================================================ Note 303.2 11/53 prom help needed 2 of 7 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 11 lines 16-FEB-1989 14:43 -< Are the 11/53 PROMs like those for the 11/73? >- -------------------------------------------------------------------------------- >> Does anyone know anything about the proms on an 11/53? >> Is the format of these proms documented in some manual, or has >> someone figured it out. I had to do a lot of digging around in the PROMs for the 11/73 (quad board) and still have the documentation for that. If you think they are similar, give me a call and I'll try to help you. Alan Frisbie (213) 256-2575 ================================================================================ Note 303.3 11/53 prom help needed 3 of 7 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 3 lines 16-FEB-1989 16:05 -< Proms are 27256 >- -------------------------------------------------------------------------------- > How big (or what part number) are the PROMS in the '53? I am looking at a pair of 27256s. ================================================================================ Note 303.4 11/53 prom help needed 4 of 7 EISNER::KENNEDY "Terry Kennedy" 4 lines 16-FEB-1989 23:50 -< That's a LOT of code! >- -------------------------------------------------------------------------------- > I am looking at a pair of 27256s. Now that's BIG. You've got 32 Kw of code there. I suggest you talk to Alan (see .-2) about his '73 info. ================================================================================ Note 303.5 11/53 prom help needed 5 of 7 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 9 lines 17-FEB-1989 13:14 -< 32 KW in a BOOT ROM??? >- -------------------------------------------------------------------------------- >> Now that's BIG. You've got 32 Kw of code there. It is kind of mind-blowing to realize that the boot ROMs now have more code than the first version of RSX! Remember the original PDP-11 boot ROMs? Sixteen rows of 16 diodes each, which you clip out or leave in to make ones and zeros. It was a big advance over toggling it in via the front panel switches. ================================================================================ Note 303.6 11/53 prom help needed 6 of 7 EISNER::KENNEDY "Terry Kennedy" 4 lines 17-FEB-1989 14:35 -< That's the M792 ROM diode matrix card >- -------------------------------------------------------------------------------- >Remember the original PDP-11 boot ROMs? Sure, that's the M792 module. I have one here with a DU bootstrap in it. Isn't that perverse? ================================================================================ Note 303.7 11/53 prom help needed 7 of 7 EISNER::KOZAM 4 lines 18-FEB-1989 01:04 -< Type-in DU boot code? >- -------------------------------------------------------------------------------- > Sure, that's the M792 module. I have one here with a DU bootstrap in > it. Isn't that perverse? Gee, could you post/mail the boot code you use?  ================================================================================ Note 304.0 VAX Architecture Handbook? 4 replies EISNER::MAYHEW "Bill Mayhew" 3 lines 15-FEB-1989 11:45 -------------------------------------------------------------------------------- I'm looking for a part number for the current VAX Architecture Handbook (if there is such a thing). My only copy is dated '81, but now I have a client who wants one. Is there a newer one? ================================================================================ Note 304.1 VAX Architecture Handbook? 1 of 4 EISNER::PROVOST 15 lines 15-FEB-1989 12:56 -< 1981 is latest, I think. >- -------------------------------------------------------------------------------- >< Note 304.0 by EISNER::MAYHEW "Bill Mayhew" > > -< VAX Architecture Handbook? >- > I'm looking for a part number for the current VAX Architecture Handbook > (if there is such a thing). My only copy is dated '81, but now > I have a client who wants one. Is there a newer one? I have VAX Architecture Handbooks dating from 1977 to 1981. At that point the VAX became obsolete. I do have a PDP-11 Architecture Handbook dated 1983-84. Would you like to upgrade? Seriously, I think that Digital thinks the new multi-volume VAX hardwaree Handbook covers the ground. It doesn't. Don't lose your 1981 book. I think the information is no longer published anywhere. Since you have the 1981 book, I assume you don't need its part number. ================================================================================ Note 304.2 VAX Architecture Handbook? 2 of 4 EISNER::DELARISCH "Arnold S. De Larisch" 9 lines 15-FEB-1989 15:21 -< 1986 VAX Architecture Manual is the Latest! >- -------------------------------------------------------------------------------- I have sitting in front of me a copy of the "VAX Architecture Handbook" Copyright 1986 (EB 26115). Its not as physically large as the 1981 edition, but seems to have all the right kinda stuff in it! I've got a couple of extra copies (3 I think) I'd be willing to part with. On the other hand, I would SURE like a copy of the 1981 version if anyone can spare a copy! (Maybe this belongs in Swap Meet?) -Arnold ================================================================================ Note 304.3 VAX Architecture Handbook? 3 of 4 EISNER::BYRNE_C "Charlie Byrne" 12 lines 15-FEB-1989 15:23 -< Digital Press has a VAX Architecture book >- -------------------------------------------------------------------------------- There is the Vax Architecture Reference Manual edited by Timothy Leonard Digital Press 1987/hardbound/417 pages/ isbn 0-932376-86-x Order Number EY-3459E-DP-B5 $40.00 Phone 1-800-343-8321 ext 178 Don't know if this is basically the same one they used to give away or not; probably just made it pretty. ================================================================================ Note 304.4 VAX Architecture Handbook? 4 of 4 EISNER::KILGALLEN "Larry Kilgallen" 6 lines 15-FEB-1989 16:18 -< Don't bother with the Architecture Handbook >- -------------------------------------------------------------------------------- The latest version left out the details of which registers get trashed on various instructions etc. The new one is worthless, although for some of the coding-type problems, an instruction card will do. Get the Reference Manual (described in .3), which really has "the right stuff".  ================================================================================ Note 305.0 CRT Screen cleaner spray 24 replies EISNER::BRUCE "Barton F. Bruce - CCA/MA" 17 lines 16-FEB-1989 05:50 -------------------------------------------------------------------------------- Dec sells a special spray for cleaning the coated CRT screens. I seem to recall the Windex class sprays are a NO-NO, and further that they don't do this particular job well, anyway. The DEC stuff works well. I also recall the DEC price encourages you to look elsewhere. What is the stuff? What makes it special? What else works - safely for user and screen? And that is also available in LARGE bottles preferably at the SUPER-MARKET? Are TEX-WIPES 91% Alcohol pads (tape drive cleaner stuff) ok? Any other related folk-lore or even truth? ================================================================================ Note 305.1 CRT Screen cleaner spray 1 of 24 EISNER::KILGALLEN "Larry Kilgallen" 6 lines 16-FEB-1989 07:03 -< A vote for Texwipe >- -------------------------------------------------------------------------------- Not the tape drive cleaner stuff, but I have had very good luck with the Texwipe "Clearview Terminal Wipes" (TX803) and "Terminal Screen Restorer" (TX805). Another truly marvelous product from them is a kit to remove the old gunked-on labels from reels of magnetic tape (the sticky labels, that is, not the ANSI labels for which the proper removal device is the TS11). ================================================================================ Note 305.2 CRT Screen cleaner spray 2 of 24 EISNER::LEDERMAN "Bart Z. Lederman" 9 lines 16-FEB-1989 07:56 -< Photographic lens cleaner? >- -------------------------------------------------------------------------------- Have you checked the price for lens cleaner at a local photography supply store? Probably more expensive, but possibly not. Certainly save to use on coated surfaces. In the TV repair business we used ammonia and water on UNCOATED TV screens, which works well even in the dirty atmosphere around New York. It's also good for cleaning up plastic cases (in case your VT terminal has had coffee spilled on it once too often) but don't let it get inside the terminal. ================================================================================ Note 305.3 CRT Screen cleaner spray 3 of 24 EISNER::MAYHEW "Bill Mayhew" 9 lines 16-FEB-1989 10:55 -< Ken Olsen's screen cleaner (read The Ultimate Entrepreneur) >- -------------------------------------------------------------------------------- As far as I have been able to determine, they no longer sell that spray, nor do they ship it with the machines (though I understand that newer DEC CRTs use a different coating that is more resilient to 'other' cleaning products, I have been reluctant to try it :-}). I have a client who uses Texwipes successfully, and also some sort of screen-cleaner that is made by them, on the full spectrum of DEC CRTs (including some that originally shipped with the spray). I'll find out what it is... ================================================================================ Note 305.4 CRT Screen cleaner spray 4 of 24 EISNER::HAHN "DECUServe Vice Chair" 10 lines 16-FEB-1989 13:11 -------------------------------------------------------------------------------- We use what ever we can get for "free". The hospital provides without too much trouble alcohol wipes (used to prep skin for injections), Screen Prep (CSP20) from Chemtronics Inc, Hauppauge NY a 2 pads Pad1 smells like alcohol, pad2 "dries and polish" Antistatic Screen Cleaner from Read Right order #TX1261 The Texwipe Co, Upper Sadle River NJ At home I use rubbing Alcohol with Tissue for drying. It all works and has not yet scratched my VT screen on my Rainbow. ================================================================================ Note 305.5 CRT Screen cleaner spray 5 of 24 EISNER::MAYHEW "Bill Mayhew" 9 lines 16-FEB-1989 15:20 -< I was misinformed... >- -------------------------------------------------------------------------------- > I have a client who uses Texwipes successfully, and also some sort > of screen-cleaner that is made by them... I lied. Turns out they use 'nothing standard'. While the computer dept. uses Texwipes, the end-users use whatever they happen to have around. There are some bottles of other third-party "stuff" but they have not been tested on the older DEC screens. Sorry... ================================================================================ Note 305.6 CRT Screen cleaner spray 6 of 24 EISNER::HARVEY "Jack Harvey" 8 lines 16-FEB-1989 20:07 -< Texwipe Wet & Dry >- -------------------------------------------------------------------------------- Texwipe makes a two-part terminal cleaning pad, called "Wet & Dry". You wash the screen first with the wet pad, and then dry it off with the other pad, which is a small paper towel. This does an excellent job of restoring the non-reflective character of the VT220 screen. Aside from the little perfume spray cans that came new VT220s, I haven't seen any other way that did this, including Texwipe's "Clearview Terminal Wipes". ================================================================================ Note 305.7 CRT Screen cleaner spray 7 of 24 EISNER::BYRNE_C "Charlie Byrne" 9 lines 16-FEB-1989 21:22 -< G Gladstone reply: "Soften your screen" >- -------------------------------------------------------------------------------- Changed new TOPIC to REPLY. - C Byrne CoMod HARDWARE_HELP ================================================================================ Note 306.0 Soften your screen No replies EISNER::GLADSTONE "Gary Gladstone (201)943-7100" 3 lines 16-FEB-1989 20:49 -------------------------------------------------------------------------------- I've been told that if you wipe a TV screen with a rag that was dampened with a liquid fabric softener, it will eliminate a lot of static cling. Haven't tried it yet though. ================================================================================ Note 305.8 CRT Screen cleaner spray 8 of 24 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 18 lines 16-FEB-1989 21:45 -< Softner kills CARPET static, too >- -------------------------------------------------------------------------------- > I've been told that if you wipe a TV screen with a rag that was > dampened with a liquid fabric softener, it will eliminate a lot > of static cling. Haven't tried it yet though. That is probably ok on an uncoated tube like a VT100, and we used to kill static from our 'computer room' CARPET (former location) by using spray bottles loaded with softner cut with enough water to be sprayable. My problem now is the coated ('half-wave' filter?) tubes that have a blueish tint, and that once you touch them your grubby finger marks make horrible reflections and ALL attempts to clean it makes more smears and streaks that seem to glare even more. The DEC spray worked. Finger marks won't stay off long enough for dust to collect, and our last smoker has left the office, and we have some HEPA class filters limiting the circulating crud in the air-conditioner air. So the only problem is FINGERS. ================================================================================ Note 305.9 CRT Screen cleaner spray 9 of 24 EISNER::KENNEDY "Terry Kennedy" 31 lines 16-FEB-1989 23:46 -< More answers >- -------------------------------------------------------------------------------- > What is the stuff? Sassafras extract. No kidding! > What makes it special? Drying characteristics. > What else works - safely for user and screen? > And that is also available in LARGE bottles preferably at the SUPER-MARKET? IBM sells similar stuff for 3270 screens. Larget size is 1 gallon. > Are TEX-WIPES 91% Alcohol pads (tape drive cleaner stuff) ok? For the screen itself - maybe. For the terminal case - no. Use the Texwipe screen cleaner pads instead - far less active chemically. Also, you'll have a better leg to stand on if you complain they stripped the coating. > Any other related folk-lore or even truth? This was one of the first DEC casualties of the new Federal hazardous- labeling laws. DEC had to include a 15-page handout describing, among other things, that rabbits held underwater in this stuff drowned. All the secretaries around here went "Yecchhh! They drown bunnies to test this?" Also required by Federal law are the real manufacturer's name and phone number, so repeat order business was nil for DEC. As far as I know, DEC only sells items requiring Federal labeling if they are *essential* to correct device operation. Thus, you can still get tape drive cleaner, but not screen cleaner. ================================================================================ Note 305.10 CRT Screen cleaner spray 10 of 24 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 6 lines 17-FEB-1989 08:43 -< Texwipe >- -------------------------------------------------------------------------------- Another vote for Texwipes Wet and Dry ! I usually don't clean my tube untill it has a 1/2 inch of crud on it and these really do the job. I haven't tested them on ribs or fried chicken yet ... ;-) ================================================================================ Note 305.11 CRT Screen cleaner spray 11 of 24 EISNER::RENES "Stuart Renes, DFWLug Chair" 17 lines 17-FEB-1989 15:08 -< The eleventh reply: >- -------------------------------------------------------------------------------- Wow... eleven replies on the subject of cleaning your CRT screen! Gee, we all seem to have a lot of free time on our hands ...haha... Well, I've tried a lot of screen cleaing stuff on VT100s, 220s and now the 300-series and found that my trusty can of foamy Windex works as good or better than the rest. There is a newer Windex on the market that uses vinegar instead of Ammonia... doesn't leave a film. The VT200 series screens are the hardest to "Clean" due to the coating - the trick is to use a soft, dry cloth to buff em down as soon as you have sprayed em with cleaner. Takes a little elbow grease but works good. Stuart. ================================================================================ Note 305.12 CRT Screen cleaner spray 12 of 24 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 6 lines 17-FEB-1989 16:50 -< Blame eating at terminals >- -------------------------------------------------------------------------------- > I haven't tested them on ribs or fried chicken yet ... ;-) > Gee, we all seem to have a lot of free time on our hands In our office it is the 'FREE' office provided microwave OILY popcorn that gucks my fingers (and terminal). ================================================================================ Note 305.13 CRT Screen cleaner spray 13 of 24 EISNER::HAHN "DECUServe Vice Chair" 1 line 18-FEB-1989 08:16 -< And what about the Key board >- -------------------------------------------------------------------------------- how do YOU clean the Key board... ================================================================================ Note 305.14 CRT Screen cleaner spray 14 of 24 EISNER::KOZAM 8 lines 18-FEB-1989 11:21 -< Warm Dishwashing Soap for Keycaps >- -------------------------------------------------------------------------------- > how do YOU clean the Key board... I pull off ALL the keycaps and wash them in dishwashing soap and warm water. For the keyboard itself, I hold it upside down and shake it until most of the crumbs fall out. I guess if you did this too often, you'd destroy the keycaps, but I'm hardly that compulsive. ================================================================================ Note 305.15 CRT Screen cleaner spray 15 of 24 EISNER::COY "Dale E. Coy" 16 lines 18-FEB-1989 13:26 -< But how do you figure out where to put them? >- -------------------------------------------------------------------------------- Since I don't have time to write a program to help me put the key caps back in the right places.... :-) A barely-damp soft sponge or towel, and a little soap, seems to do fine for me. Followed by just a damp towel to remove any residual soap. Not enough water in either case to run down between the keycaps. Air dry for a minute or two before turning on the terminal. I've never seen a problem leaving the keycaps on the keyboard, unless the sides of the keys get really dirty. Even with really old keyboard designs (20 years ago), an occasional small drop of water running down between the keys wouldn't cause a problem - if allowed to dry before power is applied. More modern keyboard designs ATTEMPT to cope with the more common problem of spilling your coffee on the keyboard. Most will survive a cup of black coffee, but sweetened drinks may gum up the works. ================================================================================ Note 305.16 CRT Screen cleaner spray 16 of 24 EISNER::BURNET "John Burnet" 5 lines 18-FEB-1989 18:13 -< rubbing alcohol >- -------------------------------------------------------------------------------- > ... spilling your coffee on the keyboard. Most will survive a cup of > black coffee, but sweetened drinks may gum up the works. If so, try an alcohol bath (for the keyboard, not its user :-) and that may get rid of the worst of the mess. ================================================================================ Note 305.17 CRT Screen cleaner spray 17 of 24 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 7 lines 21-FEB-1989 17:09 -< we did some of this before >- -------------------------------------------------------------------------------- >>> ... spilling your coffee on the keyboard. Most will survive a cup of >>> black coffee, but sweetened drinks may gum up the works. >If so, try an alcohol bath (for the keyboard, not its user :-) and that >may get rid of the worst of the mess. And see 6.53-6.62 right here in Hardware for earlier related comments. ================================================================================ Note 305.18 CRT Screen cleaner spray 18 of 24 EISNER::BONE "George Bone, MINSY" 5 lines 27-FEB-1989 10:40 -< Hold the world a sec - I need a definition. >- -------------------------------------------------------------------------------- Okay, I'm dumb. What is the 'coating' being discussed here, and what's the problem with WINDEX or other general window cleaners? As for keyboards, occasionally I use compressed gas to blow miscellaneous crumbs etc out of them. Works pretty well! ================================================================================ Note 305.19 CRT Screen cleaner spray 19 of 24 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 18 lines 27-FEB-1989 11:40 -< watch the ozone layer >- -------------------------------------------------------------------------------- > Okay, I'm dumb. What is the 'coating' being discussed here, and > what's the problem with WINDEX or other general window cleaners? There is some sort of coating that one is warned to be carefull of, and I seem to recall, but am not positive, that WINDEX class stuff is a NO-NO. > As for keyboards, occasionally I use compressed gas to blow > miscellaneous crumbs etc out of them. Works pretty well! Compressed air is great stuff for cleaning where a vacuum can't, but check the can label. Many such cans are Freon-12. So are the boat canned fog horns. A tiny can may not seem like much, but the ozone layer cares. Most large buildings have compressed air somewhere, even if it is just a hose tapping the compressor that otherwise only runs the pneumatic air-conditioning controls. Keyboards are very portable. ================================================================================ Note 305.20 CRT Screen cleaner spray 20 of 24 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 10 lines 27-FEB-1989 15:58 -< Compressed Air is DIRTY. >- -------------------------------------------------------------------------------- >> Most large buildings have compressed air somewhere, even if it is >> just a hose tapping the compressor that otherwise only runs the >> pneumatic air-conditioning controls. Keyboards are very portable. Most compressed air is very dirty, contains lots of oil and water which collect in seldom-used portions of the piping. If using a non-conditioned air supply, DON'T point it at a keyboard or other delicate stuff until you have blown out all of the crud. If you blow it through a piece of white cloth you may be surprised at what all comes out... ================================================================================ Note 305.21 CRT Screen cleaner spray 21 of 24 EISNER::LEDERMAN "Bart Z. Lederman" 14 lines 1-MAR-1989 08:03 -< Screen coating is very thin: strong chemicals no good. >- -------------------------------------------------------------------------------- The coating on the screen is a thin layer of material (used to be magnesium fluoride or something similar) 1/4 wave length of light thick. It's basically the same as the anti-reflection coating on a camera lens. Windex and similar window cleaners are ammonia based, and may take off some of the coating, especially if you rub too hard. I agree about the building compressed air supply. Also note that even if you see what looks like a filter near your air tap, it may actually be a filter and oiler, which injects oil into the air to lubricate air tools. You will not want to blow this into your keyboard. Of course, if you happen to work in an industry where you really have clean air available, a LOW POWER air current should work. ================================================================================ Note 305.22 CRT Screen cleaner spray 22 of 24 EISNER::MERRIMAN "George Merriman - CCA/NY" 1 line 1-MAR-1989 15:48 -< What is the wavelength of light? >- -------------------------------------------------------------------------------- > a thin layer of material . . . 1/4 wave length of light thick. ================================================================================ Note 305.23 CRT Screen cleaner spray 23 of 24 EISNER::HARVEY "Jack Harvey" 7 lines 1-MAR-1989 19:56 -< Light Waves Now >- -------------------------------------------------------------------------------- > -< What is the wavelength of light? >- Red is about 0.7 microns, violet about 0.4 microns. Sometimes multiple coatings are used to "tune out" different wavelengths. The incomplete cancellation at certain wavelengths causes the faint color cast seen on camera lenses and very clean VT220 screens. ================================================================================ Note 305.24 CRT Screen cleaner spray 24 of 24 EISNER::LEDERMAN "Bart Z. Lederman" 16 lines 1-MAR-1989 20:37 -< We're talking THIN >- -------------------------------------------------------------------------------- On CRT screens (and old cameras) it's usually a single layer (because the material is usually harder, and it's less expensive). It can be fairly well optimized by centering on green light (about .55 microns) as the eye is most sensitive to that color. You can tell because, if you look at the screen at an angle reflections are purple (actually magenta), as the screen reflects a little red and blue light. In any event, the layer is "thin". This is slightly off course: but if you have a "white" monitor set it for black letters on a white screen. Glare, reflections (and often dirt) don't show up as much. Black letters on a white background is also the way most written and printed material has been done since we graduated from clay tablets, and there is less contrast when you go from the screen to a printed page.  ================================================================================ Note 306.0 HSC Loading / Channel Cards / Disk Access 3 replies EISNER::ROLLER "Sir Osis of Liver" 5 lines 21-FEB-1989 18:11 -------------------------------------------------------------------------------- Can anyone provide me with any advice (sage or otherwise) about how much to load (when to buy another) HSC5x-BA? Are they capable of performing simultaneous operations, therefore running into capacity problems? Or can they only do one thing at a time, leading to collision problems? ================================================================================ Note 306.1 HSC Loading / Channel Cards / Disk Access 1 of 3 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 41 lines 21-FEB-1989 19:26 -< Performance Article in Feb '89 Quorum >- -------------------------------------------------------------------------------- In this months issue of Quorum VAXcluster Systems (Volume 4 issue 3 February 1989) there is an article that I think might meet your needs. The article is entitled "Understanding I/O Bottlenecks in VAXcluster Systems", by Hank Jakiela, VAXcluster Performance Engineering High Performance Systems. The highlights are as follows: o I/O Characteristics of Workloads o Typical I/O Characteristics o Performance Tradeoff: Throughput and Response Time o Components of the VAXcluster I/O Subsystem o Disk Drive Characteristics o Common Disk Bottlenecks o Suggestions for Eliminating Disk Bottlenecks o HSC Bottlenecks * o HSC Disk Channel o HSC50 and HSC70 o VAX CI Adapters o CI Adapter Characteristics o CI Bus The summary indicates the following items: o Not all I/O loads are alike, and it is important to understand the characteristics of the workload when trying to identify and correct I/O bottlenecks in VAXcluster systems. o Disk drives are probably the most common I/O bottleneck in VAXcluster systems, especially when workload consists of many small I/O requests randomly distributed over the disk surface. o The CI adapter may be a bottleneck in fast VAX processors running I/O-intensive applications. * o The HSC50 and HSC70, disk data channels, and the CI bus are not likely bottlenecks in most situations. Quorum VAXcluster Systems is a _FREE_ publication from Digital. If you would like a copy of the article and/or a copy of the subscription form send me a MAIL message with your FAX or mailing address. ================================================================================ Note 306.2 HSC Loading / Channel Cards / Disk Access 2 of 3 EISNER::BATES "Ken Bates" 49 lines 25-FEB-1989 11:29 -------------------------------------------------------------------------------- I'll be giving a session on configuring for performance at the next DECUS which will cover this in depth, but here's a quick response: The channel card interface on an HSC supports up to 4 ports. All 4 ports can have a seek command outstanding. The port will service the seek completions as they occur, not in the order they were received, so you gain a fair amount of parallelism. The disks on all 4 ports are monitored for their current rotational position. If more than one disk lands on cylinder at the same time, the first one to reach the selected block will begin to transfer. As before, this parallelism allows increased performance. Only *one* data transfer per k.sdi (channel card) can occur at a time. If one begins and another one is ready, it is not done and must wait until the first one has completed before it may begin. Naturally, this will cause a rotational miss, since the disk has to once again rotate back to the block of interest. From the above, it can be seen that performance will suffer if: 1. Two or more disks have work at the same time, AND 2. All disks land on cylinder at the same time, AND 3. The data is under the heads of all disks at exactly the same time. The probabilioty of the above occurring is, as Hank points out, very dependent on the VMS workload. If you issue large requests in a sequential file, double buffer them, and do the same on multiple disks on the same channel card, then you would probably suffer a performance loss. On the other hand, consider a "typical" VMS system, such as interactive timesharing, office automation, or OLTP, which has an average transfer size of 8 sectors, with a median of 2 sectors, and a mode of 1 sector. If the average seek time is about 10-15 ms (the seeks which occur under VMS, not what the drive spec says) and the average rotational latency is 8.3 ms, then the chance of having a collision between two transfers is extremely small. Not only that, but you must also have both requests outstanding at the same time. In general, you are probably safe having up to 4 drives on a single channel card, *if* you have a typical VMS workload. In this case, you will notice a small degradation in response time only at very high request rates. For example, if you have 4 RA82 drives on a single channel card, the average response time per drive will be about 75 ms with a combined rate of 175 requests per second. If each drive were on a separate channel card, the average response time would drop to about 65 ms at the same rate. Below about 100 requests per second (25 per drive), there is no difference between 1 drive per channel card and 4 drives per channel card. - Ken ================================================================================ Note 306.3 HSC Loading / Channel Cards / Disk Access 3 of 3 EISNER::ROLLER "Sir Osis of Liver" 6 lines 28-FEB-1989 08:18 -< Dual porting mitigates further >- -------------------------------------------------------------------------------- Thanks very much. Here's another consideration that I've encountered as well. If one dual ports all their disks evenly, then there are at most two disks actually active on any given k.sdi, therefore making the liklihood of transfer collision even less. I feel very comfortable about loading 'em up now.  ================================================================================ Note 307.0 Color Screen Image Processing/Video Standards & Gadgets 11 replies EISNER::LOBMEYER "Fred Lobmeyer City of NY" 12 lines 21-FEB-1989 21:34 -------------------------------------------------------------------------------- One of our sites is using a Seiko CH5312 with a CADD system. This printer sits between the monitor and the 3500 and steals the RGB output. The problem is that other users at there stations would like to print their screens and rather than buying many of these I would like to use one shared printer. Not knowing the workstation software, although there is a print screen option on one of the pull down menus, can a color printer similar to this be attached to a port and served via a cluster common queue? If so please specify an alternate printer and a contact. Thanks in advance! ================================================================================ Note 307.1 Color Screen Image Processing/Video Standards & Gadgets 1 of 11 EISNER::WICHMAN "Jon Wichman" 12 lines 22-FEB-1989 09:47 -< How about a video mux? >- -------------------------------------------------------------------------------- I don't believe you can run this on a queue since it does use video signals, however Seiko does make a 4-port video multiplexer which works well. This runs about $3k. For Vaxstations, you'll also need a Video Signal Adapter kit for _each_ Vaxstation. Also, you'd probably want to check out the remote copy switches which lets you initiate a print from the workstation (rather than the mux panel). With the proper cables, you can go to about a 140' radius from the printer to each workstation. Contact Seiko or your CADD vendor for more info. I have a Seiko reps number if you'd like/need it. ================================================================================ Note 307.2 Color Screen Image Processing/Video Standards & Gadgets 2 of 11 EISNER::BURNET "John Burnet" 13 lines 22-FEB-1989 18:56 -< What's a video signal adapter? >- -------------------------------------------------------------------------------- > < Note 307.1 by EISNER::WICHMAN "Jon Wichman" > > For Vaxstations, you'll also > need a Video Signal Adapter kit for _each_ Vaxstation. ^^^^^^^^^^^^^^^^^^^^^^^^ What is this, exactly? Is it needed only to connect the VAXstation to the video mux? Or does it do some sort of scan-rate conversion? I'm curious because we've been looking at film recorders for use with PC graphics cards (1280 x 1024) and VAXstations, and the VAXstation's weird 1024 x 864 resolution (what's the scan rate?) is rather inconvenient. I would almost rather have the lower 1024 x 768 resolution in order to be able to use other devices with it that use the 64 kHz horizontal scan. ================================================================================ Note 307.3 Color Screen Image Processing/Video Standards & Gadgets 3 of 11 EISNER::WICHMAN "Jon Wichman" 56 lines 23-FEB-1989 10:21 -< It's something that adapts Video Signals? >- -------------------------------------------------------------------------------- > need a Video Signal Adapter kit for _each_ Vaxstation. ^^^^^^^^^^^^^^^^^^^^^^^^ | What is this, exactly? Is it needed only to connect the VAXstation to | the video mux? Or does it do some sort of scan-rate conversion? Unfortunately, I haven't worked with this stuff for about 9 months now so, my memory is a little hazy on some of this... As I recall, the Video Signal adapter mainly does interface conversion, splitting out and/or putting together the sync signal from the G signal or going from one type of cable to another. Seiko has 4 different adapters that I know of. From the 3/86 Seiko CH5300 brochure CH-620; BNC connector (75 ohms) (For use with non-loop through monitors) CH-621; TTL level 37P D-SUB connector (For use with IBM PC/G) CH-622; Differential video signal twin axial connector (For use with IBM PC/GX) From notes of a phone conversation with Seiko 8/87 CH624-01; Supports either CGA or EGA on a IBM/PC, includes cables (I think they all do), installation is extra if not installed with the copier. | I'm curious because we've been looking at film recorders for use with PC |graphics cards (1280 x 1024) and VAXstations, and the VAXstation's |weird 1024 x 864 resolution (what's the scan rate?) is rather inconvenient. Is scan-rate the same as the refresh rate? If so, it's 60hz non-interlaced According to "A Technical Summary of Digital's Vaxstation Family" (EB-29389-51), the Vaxstation supports the HP7510 film recorder and the MPS-2000 film recorder. You might be interested in a company called PDS Video Technology, Inc. 1152 Santa Barbara St., San Diego, CA 92107, (619) 222-7900 I've never dealt with them, however, they were recommended by Seiko to solve a particular customer problem. They manufacture a line of Video scan converters which convert Various Tektronix terminal and EGA/PGA PC display output to NTSC. Also, they have adapters for: RGB output for Tek 4205 and 4207 terminals. Mono output for Tek 41xx and 42xx terminals for use with monochrome hardcopy units. Thier price list show that they also have: Photo/Slide Making equipment ("PDS integrated/tuned") Wide-Screen video projectors Color Monitors Video cables. I have no relationship with Seiko or PDS. ================================================================================ Note 307.4 Color Screen Image Processing/Video Standards & Gadgets 4 of 11 EISNER::LEDERMAN "Bart Z. Lederman" 6 lines 23-FEB-1989 15:47 -< Will it do RS-170 to TTL? >- -------------------------------------------------------------------------------- Do you happen to know if any of those adaptors would turn analog RS-170 RGB video sync on green (what comes out of DEC equipment like PROs, Rainbows, VT240s, etc.) into RGBH TTL as is used in IBM PC CGA type adaptors? I have a computer which emits the former, and a color monitor which only accepts the latter and don't have time to wire my own. ================================================================================ Note 307.5 Color Screen Image Processing/Video Standards & Gadgets 5 of 11 EISNER::BURNET "John Burnet" 16 lines 23-FEB-1989 18:53 -< RS-170A <--> TTL RGB? >- -------------------------------------------------------------------------------- > < Note 307.3 by EISNER::WICHMAN "Jon Wichman" > Thanks a bunch for the info on Seiko and PDS. We will follow up those leads. > < Note 307.4 by EISNER::LEDERMAN "Bart Z. Lederman" > > Do you happen to know if any of those adaptors would turn analog > RS-170 RGB video sync on green (what comes out of DEC equipment > like PROs, Rainbows, VT240s, etc.) into RGBH TTL as is used in IBM > PC CGA type adaptors? Another thing we'd like to do is exactly the opposite of what Bart mentions... we have a low-res film recorder which takes an RS-170 signal, currently from a Rainbow, and we'd like to use the same device with a PC EGA board. It certainly seems that there's a market for both of these adapter types. ================================================================================ Note 307.6 Color Screen Image Processing/Video Standards & Gadgets 6 of 11 EISNER::MCCORMICK_B "Bob McCormick" 17 lines 24-FEB-1989 18:40 -< Beware of devices that are RS170 "like" >- -------------------------------------------------------------------------------- Thing to watch: Be carefull of RS170 or RS170A "like" equipment. I worked on a workstation project where we generated video "to the spec" ... which isn't easy 'cuz the numbers don't divide easily by usual (inexpensive) logic ... So a lot of times you'll loose something: maybe actual video, maybe quality. Check in your area for companies that supply the television broadcast industry. They often have boxes that will do these kinds of conversion, providing gen lock et all... As for RS170, I've got the spec if anyone really wants more detail. For an SASE I can foto copy it -- mail to me and I'll tell you how much postage! -- or if interest, I can post general synopsis here. ================================================================================ Note 307.7 Color Screen Image Processing/Video Standards & Gadgets 7 of 11 EISNER::LOBMEYER "Fred Lobmeyer City of NY" 19 lines 26-FEB-1989 19:13 -< Color output question restated? >- -------------------------------------------------------------------------------- > Not knowing the workstation software, although there is a print > screen option on one of the pull down menus, can a color printer > similar to this be attached to a port and served via a cluster common > queue? If so please specify an alternate printer and a contact. > I don't believe you can run this on a queue since it does use video > signals, however Seiko does make a 4-port video multiplexer which > works well. > With the proper cables, you can go to about a 140' radius from the > printer to each workstation. This is not an option because the distance is over Seiko's MAX. What I really want to know is there a way to capture the screen to a color printer that doesn't sit between the system and the monitor? (ie. some type of print screen option toggled from the keyboard?) ================================================================================ Note 307.8 Color Screen Image Processing/Video Standards & Gadgets 8 of 11 EISNER::MAYHEW "Bill Mayhew" 1 line 26-FEB-1989 23:24 -< What about the LJ250 or LCG01 color printers? >- -------------------------------------------------------------------------------- ================================================================================ Note 307.9 Color Screen Image Processing/Video Standards & Gadgets 9 of 11 EISNER::HASSINGER "Bob Hassinger" 3 lines 27-FEB-1989 10:39 -< I would be very interested. A useful reference here. >- -------------------------------------------------------------------------------- > As for RS170, I've got the spec if anyone really wants more detail. > For an SASE I can foto copy it -- mail to me and I'll tell you how > much postage! -- or if interest, I can post general synopsis here. ================================================================================ Note 307.10 Color Screen Image Processing/Video Standards & Gadgets 10 of 11 EISNER::MCCORMICK_B "Bob McCormick" 3 lines 6-MAR-1989 19:27 -< Posting coming shortly >- -------------------------------------------------------------------------------- OK ... I'm at home (the spec's at work) ... It may take me another week to get it on here as I'm going on the road as usual... ================================================================================ Note 307.11 Color Screen Image Processing/Video Standards & Gadgets 11 of 11 EISNER::MERRIMAN "George Merriman - CCA/NY" 1 line 28-MAR-1989 15:01 -< for further reading . . . >- -------------------------------------------------------------------------------- See also topic 17 in WORKSTATIONS.  ================================================================================ Note 308.0 VS2000 Mouse Problem 3 replies EISNER::WATERMAN "Doug Waterman (609)423-1040" 29 lines 23-FEB-1989 10:26 -------------------------------------------------------------------------------- I'm having a problem with my VS2000 related to the mouse. I've got VMS V5.0-2 and VWS V4.0 on a VS2000 with an 8MB Clearpoint board and an RD54. I'm a member of an LAVC consisting of three similar stations (except 6MB of DEC memory, RD53), two 11-750's for boot nodes and an HSC disk farm, and a MV-II. The other stations do not have this problem. Here's the scenario: In the morning, I click the mouse and get the "Workstation Options" menu as expected. I then go through normal procedures to log in a couple of terminal emulators. When the emulators are running, I cannot grab the window edge to move them, nor do I get any results when I click on the "Menu" box in the upper left corner. When I change the window to an icon (using an escape sequence), I cannot expand the icon using the mouse. I can, however, call up the "Workstation Options" menu at any time and I can pop a window to the foreground with the mouse. I notice that this only happens in the morning, after a long period of inactivity. Sometimes all of the mouse clicks I have done will begin to happen one after another 15 to 30 minutes later. It's as if the mouse input was being buffered and suddenly got turned back on. The thing that bugs me is that some of the mouse clicks go through (like calling up the "Workstation Options" menu), but others don't (like moving a window). As far as I can tell, there are no resource allocation problems. I'm not really sure what to look for in this case. A shutdown/reboot solves the problem. Would it be the hardware mouse controller? The display manager? Is it SPR time? Any ideas? ================================================================================ Note 308.1 VS2000 Mouse Problem 1 of 3 EISNER::MAYHEW "Bill Mayhew" 10 lines 23-FEB-1989 11:03 -< Couple things to look at >- -------------------------------------------------------------------------------- Do you get the same behavior when you use the keyboard mouse-equivalent keystrokes? (Ctrl-shift-arrows move around, Ctrl-shift-Select is the same as hitting the left button) (Note: I assume this is still true with VWS 4.0) With the mouse sitting still, and the VS2000 "idling", does a MONITOR SYSTEM show any I/O going on? I am wondering if this is related to the strange behavior I see on my VS2000 when it gets hit with a lot of serial-line activity at once (but mine is at VMS 4.6 and does worse things). It is discussed elsewhere in this conference, but without resolution. ================================================================================ Note 308.2 VS2000 Mouse Problem 2 of 3 EISNER::MCCORMICK_B "Bob McCormick" 7 lines 24-FEB-1989 18:45 -< I saw problem like this: was bad mouse >- -------------------------------------------------------------------------------- I saw a problem like this once ... yes, take Bill's suggestion and check the I/O ... as you (most likely) know, the mouse is a serial device ... I had a flakey mouse once that just rambled on and on ... garbage data ... does the mouse cursor track at all? Have you tried another mouse? When the button events do occur, do they occur where they should have (ie on the objects you wanted them to) or just randomly (or where the cursor currently sits)? ================================================================================ Note 308.3 VS2000 Mouse Problem 3 of 3 EISNER::WATERMAN "Doug Waterman (609)423-1040" 22 lines 27-FEB-1989 09:01 -< More info... >- -------------------------------------------------------------------------------- >> I saw a problem like this once ... yes, take Bill's suggestion and >> check the I/O ... as you (most likely) know, the mouse is a serial >> device ... I checked it. I only see I/O happening when the mouse does respond. Those times when I click and don't get a response -- no I/O. >> I had a flakey mouse once that just rambled on and on >> ... garbage data ... does the mouse cursor track at all? Cursor tracking looks great. >> Have you tried another mouse? Not yet. But good suggestion. I'll see if I can dig one up. >> When the button events do occur, do they occur >> where they should have (ie on the objects you wanted them to) or >> just randomly (or where the cursor currently sits)? They occur at the locations where I clicked the mouse, regardless of the current location of the cursor.  ================================================================================ Note 309.0 Would you buy a used LN03R? 24 replies EISNER::TINKELMAN "Bob Tinkelman - CCA" 17 lines 23-FEB-1989 21:56 -------------------------------------------------------------------------------- We are currently using an LN03R in our New York office. We are considering buying one or two additional LN03Rs - one for our Cambridge office and one for a client's machine. We bought our first LN03R late last year at what seemed like quite a good price from a DEC reseller. [Scherer's - $3950] Before simply "doing it again", I thought I should at least consider the used market. It appears that we can buy a used LN03R for approximately half the price of the discounted new model. However, I'd be interested in feedback on the following: 1. How large a "page counter" should I be willing to accept? Do LN03s die at some point or do they, given regular maintenance, run forever? 2. What's the story with the VAX ScriptPrinter software license with a used LN03R? Did it travel with the printer or stay with the first VAX to which the printer was attached? [I guess this doesn't matter to us with our "second" LN03R, but it does to our client with his first.] ================================================================================ Note 309.1 Would you buy a used LN03R? 1 of 24 EISNER::KENNEDY "Terry Kennedy" 15 lines 24-FEB-1989 05:37 -< It's not the mileage, it's the maintenance >- -------------------------------------------------------------------------------- I *never* buy used items containing moving parts without 'right of refusal'. We've dealt with many vendors in the DEC used market, and few have had any problems with this. We offer to buy item 'X', they freight it to us on their tab, we inspect it and if we're happy, we pay the items cost + freight. Other- wise, they get to fix it or come cart it away (our choice). While that may be overkill in your case, certainly asking for a printout of the test page(s) from the unit in question is reasonable. If you're putting it under contract (anybody's), then they should fix it when it breaks, regard- less of 'mileage'. A possible exception is where the engine is rated for X pages per month, the unit is 1 year old, and there are X*24 pages on the odometer. The point I'm trying to make is that the quality of maintenance (and of op- eration) makes more of a difference than the amount of use... ================================================================================ Note 309.2 Would you buy a used LN03R? 2 of 24 EISNER::MAYHEW "Bill Mayhew" 22 lines 24-FEB-1989 11:18 -< Some LN03 specifics >- -------------------------------------------------------------------------------- I agree with Terry, but... My local Field Service unit manager confided in me that the LN03 is one of the two most trouble-prone pieces of equipment they service (i.e. responsible for the most calls). This is reflected in the high service price, which just got higher (it is now 22% per year of the list price of a base LN03). He also confided that it is common to have to swap out an LN03 when the page counter gets above 100K pages. I have heard this figure elsewhere, too. My first LN03 had to be swapped at about 40K pages, and they swapped it with 3 other units before finally getting one that worked right. I insisted that it have a reasonably-close page count to my original printer (or lower), and they complied. I always try to steer clear of used, maintenance-dependent electro-mechanical equipment unless confronted with a very good price. By my personal standards, that would mean under $1K for a standard LN03, well under $2K for an LN03R. Of course a very good warranty would also be influential. ================================================================================ Note 309.3 Would you buy a used LN03R? 3 of 24 EISNER::MCCORMICK_B "Bob McCormick" 15 lines 24-FEB-1989 18:50 -< Easy to abuse - depreciates FAST = LN03 >- -------------------------------------------------------------------------------- I agree with Bill and Terry's comments... The organization I work for has a well seasoned LN03 ... its one of the few things in the office that's a peripheral thats under an on-site maintenance contract. They have been known to push 10,000 pages a month through it. And even though its under service, it has frequent calls: its mechanical, its got tons of heat, and its an appliance (ie users use it but don't maintain it) Its quality HAS degraded over time: nothing you can do about it. Unless you are making a killing buying the used gear, I'd stick with new laser technology unless you could insist on some level of performance guarantee ... and I don't think you're going to find that many places... ================================================================================ Note 309.4 Would you buy a used LN03R? 4 of 24 EISNER::KOZAM 14 lines 25-FEB-1989 01:02 -< LN03 fades insidiously >- -------------------------------------------------------------------------------- Laser printers (especially the LN03) seem to slowly and insidiously degrade in quality, regardless of how much maintenance you give them. You get little streaks, lines, ever-so-slightly blurry printing, fade outs. So, if you need top quality printing, don't buy a used LN03. These printers WILL still print, however, so if you're happy setting up a two printer system (one for top quality, one for rough drafts), then a used item may be OK. On a slight tangent, I recently visited a client and noticed an old LA34 sitting next to the LN03. The LA34 was happily, albeit slowly, banging out pages. On questioning, they explained that they had recently calculated the cost per page on the LN03 to be about 10 cents, while the LA34 was only about 1 cent. Impack printing isn't dead yet - not even moribund. ================================================================================ Note 309.5 Would you buy a used LN03R? 5 of 24 EISNER::HUNT_AL "Alan B. Hunt (Ford Motor Co.)" 13 lines 27-FEB-1989 18:06 -< Made to throw away >- -------------------------------------------------------------------------------- From what I have heard the problem isn't with LN03's but with all table top laser printers. They are designed to work till they reach something around the 100,000 page limit and then watch out. For their original intent (light work in an office environment) this will suffice and give a few years of output. The basic design used by all the manufacturers is essentially to make throw away hardware. Sounds bad but to make them last longer would probably drive the price so high that the average user couldn't afford them. For this reason I wouldn't want to buy any used laser printers. ================================================================================ Note 309.6 Would you buy a used LN03R? 6 of 24 EISNER::ANGLEY "Kevin Angley" 6 lines 5-MAR-1989 12:20 -< In support of the little guy >- -------------------------------------------------------------------------------- I have two LN03's with approximately 130,000 miles (pages) on them. No quality problems except a very occasional thing that FS fixes right away. 10 cents a page? I haven't done the calculation but that seems awfully high - my rough guess would be 3-4 cents. ================================================================================ Note 309.7 Would you buy a used LN03R? 7 of 24 EISNER::RICE "Gary Rice" 9 lines 6-MAR-1989 07:55 -------------------------------------------------------------------------------- > 10 cents a page? I haven't done the calculation but that seems awfully > high - my rough guess would be 3-4 cents. I DID the calculation for an APPLE Laserwriter IINT. Assuming the lifetime of 100,000 pages, 3000 pages per toner cartridge, and no residual market value (along with the best price that I could find from the local dealer) the price per page came out to $0.08. Gary ================================================================================ Note 309.8 Would you buy a used LN03R? 8 of 24 EISNER::MAYHEW "Bill Mayhew" 26 lines 6-MAR-1989 15:49 -< Price-per-page calculation for an LN03 >- -------------------------------------------------------------------------------- Assume: 3000 pp./toner cartridge, 2 toner cartridges/kit, kit @ $85 DEC list 7500 pp./user maintenance kit, kit @ $202 DEC list 36 mo. printer lifetime, base LN03 @ $2953 DEC list, maintenance @ $54/mo. (DEC Basic service), straight-line depreciation with no salvage value 5000 pp./month (last 'official' rated service level for an LN03 I've seen, though some DEC folks will say 10,000) Paper @ $25/case (10 reams, type "1120" standard white paper) Paper: $0.005/page Toner: 0.014 User Maint. kit: 0.027 Service: 0.011 Depreciation: 0.016 Subtotal $0.073 This doesn't include electricity (probably around $0.001/page), space, and air conditioning costs. An LN03R (ScriptPrinter), at $5605 list, would add about 1.5 cents/page to the above calculation, all other things being equal. ================================================================================ Note 309.9 Would you buy a used LN03R? 9 of 24 EISNER::BURNET "John Burnet" 7 lines 6-MAR-1989 17:07 -< For the Ricoh 4081, the magic number is 600,000 >- -------------------------------------------------------------------------------- The rated engine life of the Ricoh 4081 (used in the LN03, 3+, and 3R) is 600,000 pages. I imagine that it is possible to have the engine reconditioned after that point, but it's probably better just to buy a new printer. The "odometer" setting on a used printer will of course determine in part what the resale value of that printer is. I myself probably wouldn't want to buy an LN03 that had printed more than about 250,000 pages. ================================================================================ Note 309.10 Would you buy a used LN03R? 10 of 24 EISNER::NORTON "Bill Norton" 12 lines 7-MAR-1989 09:21 -< What to do with *old* LN03's? >- -------------------------------------------------------------------------------- If you had 5 LN03's coming off lease, with counters ranging from 140,000 to 355,000, and the lessor offered a buyout at $1500 each, would you: 1. Buy them out, since on DEC maintenance they'll refurbish them as needed. 2. Lease new ones to get fresh hardware. 3. Get something 'better'. If so, what? These are plain LN03's, and only thing they have to do is print *lots* of text, from DSR and WPS+. So who's got a higher rated lifetime and monthly page count, quieter operation, higher speed, lower supply cost. ================================================================================ Note 309.11 Would you buy a used LN03R? 11 of 24 EISNER::MAYHEW "Bill Mayhew" 30 lines 7-MAR-1989 10:03 -< My two cents: Don't go for it >- -------------------------------------------------------------------------------- Easy for me to say (since I don't have to do the negotiating), but at that price I'd be inclined to tell the lessor to pound salt, since I think that's marginally a rip-off price for units of that age. Offer them $950 or $1000, tops. Else, lease new ones to get fresh hardware (and be sure to look into the buy-out costs up-front... unless your monthly payments were *very* low, the lessor is making A Bundle if they sell them to you at $1500.) You should be able to lease 5 new ones at around $130/month each (excluding maintenance), with an end-of-term buy-out of no more than 10% ($295/printer), on a 3-year lease. (Rates as of 12/88, may have gone up a bit since then with the general rise in interest rates in the economy.) DEC will refurbish and, if necessary, replace your current machines to keep them going, but you can lose a *lot* of uptime persuading them to do it, esp. trying to get them to maintain that ever-nebulous "acceptable print quality." LN03s are essentially a full generation old, at this point. I haven't looked recently, but HP LaserJet IIs are an obvious option. The Xerox 4045 is also a very nice, 20% faster, considerably lower supply-cost machine... *if* one of the two companies that has been doing LN03 emulation for them did a sufficiently complete job for you (probably YES, based on your described workload). A colleague figured out that he could buy a 4045 on what he'd save on operating an LN03 in 1-2 years. I don't know if I fully believe that, but... ================================================================================ Note 309.12 Would you buy a used LN03R? 12 of 24 EISNER::BURNET "John Burnet" 28 lines 8-MAR-1989 19:18 -< Talaris 1590 >- -------------------------------------------------------------------------------- > would you: > 1. Buy them out, since on DEC maintenance they'll refurbish them as needed. But will DEC extend this to refurbishing the print engine? Is this covered in the maintenance agreement for LN03s? (I'm asking from ignorance here -- if it really is covered, that would certainly provide another explanation for why LN03 maintenance is so expensive.) > So who's got a higher rated lifetime and monthly > page count, quieter operation, higher speed, lower supply > cost. Talaris 1590: 15 ppm, two paper trays, Ricoh engine (forgot the model number), rated for something like 20K pages/month, 1M-page engine lifetime (not absolutely sure of those numbers, but they're close). Fully LN03+ (plus) compatible; also emulates the HP Laserjet and a few others. The engine uses the *same* toner cartridges as the LN03, and a slightly different maintenance kit. (All maint. kit items except the OPC belt are identical to the corresponding LN03 items.) If you do consider the 1590, don't let them sell you a 1590T (faster graphics processing) or any of the other fancy options (Postscript, hard disk, megabytes of memory, font cartriges) unless you really need the features. The basic 1590 is about $8000 and can be considered an "LN03+ times two" with regard to speed, paper handling capacity, and engine life. It's quite a capable printer. Disclaimer: No, I don't work for Talaris. :-| ================================================================================ Note 309.13 Would you buy a used LN03R? 13 of 24 EISNER::MAYHEW "Bill Mayhew" 23 lines 8-MAR-1989 23:51 -< DEC *seems* to fix ANYTHING that goes wrong >- -------------------------------------------------------------------------------- > But will DEC extend this to refurbishing the print engine? Is this > covered in the maintenance agreement for LN03s? (I'm asking from > ignorance here -- if it really is covered, that would certainly provide > another explanation for why LN03 maintenance is so expensive.) My working assumption is "yes". This question is sort of like the "I don't need a license for compiler X to run my software written in X on a target VMS system, do I?" question posed in DEC_SOFTWARE though... I'm not sure where, if anywhere, it is written. My working assumption is that DEC will do whatever they have to do to keep the system/peripheral running. I have no reason to doubt this. I doubt that they would particularly refurbish the print engine specifically -- they'd either replace the relevant FRU, or replace the whole printer (somewhere in here I mentioned that DEC went through 4 printers in one week getting my LN03 fixed, at one point not long ago). I too assume that this is part of the reason why maintenance on an LN03 is so expensive (and went up by 10% sometime within the past 8 months). ================================================================================ Note 309.14 Would you buy a used LN03R? 14 of 24 EISNER::ROECKEL 17 lines 6-APR-1989 10:07 -< HP LaserJET =\= DEC >- -------------------------------------------------------------------------------- > 3. Get something 'better'. > If so, what? These are plain LN03's, and only thing > they have to do is print *lots* of text, from DSR and > WPS+. So who's got a higher rated lifetime and monthly > page count, quieter operation, higher speed, lower supply > cost. We have some HP LaserJet II's on our network. Extremely reliable, easy to maintain, etc. The only problem is they are *NOT* DEC compatible. What I mean is that all the sequences used to BOLD, UNDERLINE, etc DO NOT WORK. This is really ashame, since the printer is TOPS !! -Bruce ================================================================================ Note 309.15 Would you buy a used LN03R? 15 of 24 EISNER::HAHN "DECUServe Vice Chair" 2 lines 6-APR-1989 17:28 -< HP LaserJET =~ LN03 >- -------------------------------------------------------------------------------- We have just put up a device which allows LN03 emulation on the HP. If you need more information send mail. ================================================================================ Note 309.16 Would you buy a used LN03R? 16 of 24 EISNER::MAYHEW "Bill Mayhew" 20 lines 6-APR-1989 18:56 -< LN03 emulation for LaserJet >- -------------------------------------------------------------------------------- One such family of devices comes from Gold Key Electronics in New Hampshire, 800-325-0150. They have a product line called "Switchmate" which provides printer sharing among multiple incoming data sources (it watches for a data stream), and some models will recognize LN03 data sequences and convert them to HP LJ equivalents, while passing through, unchanged, vanilla HP stuff. I have not used this product, but we have put one of their simpler automatic switches at a client site and it works very nicely. This is a model that talks *to* an LN03. It even allows you to put escape sequences in the text using some combination of backslash characters (I forget the details) in place of the character... so you can put escape sequences in, say, word processing documents where they would otherwise not be allowed. AND, it watches the data stream and when it times out, it sends a formfeed to the printer, eliminating the annoying "6" in the display that forces you to hit the offline/online button on the LN03 to eject the last partial page. ================================================================================ Note 309.17 Would you buy a used LN03R? 17 of 24 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 25 lines 6-APR-1989 19:43 -< what else == LN03R? >- -------------------------------------------------------------------------------- > -< LN03 emulation for LaserJet >- The Laser Connection folks have PS Jet+ (2meg) and PS Jet+M (3meg) kits available claim to convert your HP LaserJet (or any Canon CX based laser printer) to Postscript, and have a JetScript unit for LaserJet IIs. These include 35 Adobe fonts. At first glance this might = an LN03R. But does it? What has DEC done to the LN03R that won't be compatible with other Adobe PostScript devices? The LN03R is a write WHITE machine, and the Canon CX is write BLACK. What does that do to different kinds of font libraries? Has anyone done this conversion and simply dropped it into an existion LN03R based environment? What worked, and what broke? What happens if you use something at the other end of the spectrum like the KodaK Ektaprint #24 that does PostScript and 92 impressions / min (thats only 46 pieces of paper in duplex mode) in place of an LN03R. Does POSTSCRIPT=POSTSCRIPT=POSTSCRIPT, or not? ================================================================================ Note 309.18 Would you buy a used LN03R? 18 of 24 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 8 lines 6-APR-1989 23:25 -< PostScript >- -------------------------------------------------------------------------------- > Does POSTSCRIPT=POSTSCRIPT=POSTSCRIPT, or not? Pretty much. We use QMS PS-800, PS-810, PS-820, ColorScript 100 and the occasional LaserWriter printers on our cluster. They all pretty much work the same and work with applications that generate PostScript (Including the DECwindows print screen stuff). The biggest differences tend be in the size of the imageable area on the page. All in all, PostScript is wonderful. ================================================================================ Note 309.19 Would you buy a used LN03R? 19 of 24 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 7 lines 7-APR-1989 11:05 -< PostScript is not always PostScript >- -------------------------------------------------------------------------------- > Does POSTSCRIPT=POSTSCRIPT=POSTSCRIPT, or not? Not. We have Texas Instruments Omni Lasers. They handle encapsulated PostScript just fine, but we had to wrangle with them to get them to handle non-encapsulated PostScript, and not all non-encapsulated PostScript files work, although many will. In particular, we sometimes have trouble with the files produced by Mentor. ================================================================================ Note 309.20 Would you buy a used LN03R? 20 of 24 EISNER::MAYHEW "Bill Mayhew" 5 lines 7-APR-1989 13:15 -< Enhanced HP doesn't get you the LN03R software >- -------------------------------------------------------------------------------- > At first glance this [enhanced HP LaserJet] might = an LN03R. One thing you would of course NOT get, going this route, is the license for DEC's Scriptprinter Software (the ASCII/ANSI->PS symbiont), which is included with the LN03R. ================================================================================ Note 309.21 Would you buy a used LN03R? 21 of 24 EISNER::TINKELMAN "Bob Tinkelman - CCA" 11 lines 7-APR-1989 16:57 -< The question: NonDEC PS printers with DEC ScriptPrinter Software >- -------------------------------------------------------------------------------- RE: < Note 309.18 by EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" > | > Does POSTSCRIPT=POSTSCRIPT=POSTSCRIPT, or not? | | Pretty much. We use QMS PS-800, PS-810, PS-820, ColorScript 100 | and the occasional LaserWriter printers on our cluster. Do you use DEC's ScriptPrinter software? [You get its license free with an LN03R. Or you can buy it separately.] Has anyone used this with PS printers other than the DEC ones? ================================================================================ Note 309.22 Would you buy a used LN03R? 22 of 24 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 19 lines 7-APR-1989 19:29 -< Not the DEC SW >- -------------------------------------------------------------------------------- > Do you use DEC's ScriptPrinter software? No. We use a 3rd party symbiont from Pacer Software, called Pacer Print. It does pretty much the same things as the ScriptPrinter software. We have been midlin' pleased with it. > They handle encapsulated PostScript just fine, but we had to > wrangle with them to get them to handle non-encapsulated > PostScript Interesting. Encapsulated PostScript (EPS) is really just a comment based structuring convention, so there shouldnt really be any difference. One thing EPS does specify is the maximum record length. Could whatever youre driving the printer with be freaking on the file structure that your application is writing the PostScript out with? We have seen some problems with files that are stream_lf format. An extra carriage return or a truncated record can really ding up a PostScript program. Is TI using Adobe? ================================================================================ Note 309.23 Would you buy a used LN03R? 23 of 24 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 15 lines 9-APR-1989 00:53 -< Fonts should be genned for white vs. black. >- -------------------------------------------------------------------------------- >The LN03R is a write WHITE machine, and the Canon CX is write >BLACK. What does that do to different kinds of font libraries? Don't know in the general case, but when we first got our TeX distribution it came with the standard DVI2LN3 driver, and a set of font files generated for write-black printers. They worked, but the characters were universally too thin -- every page looked as if the printer was running out of toner. We told Metafont (I guess; I'm not the TeXpert here) to generate a set of font files optimized for write-white, and things started looking the way we expected them to. Much better, in other words. Oh, the underscores in the CM fonts, and the thin lines in the line-drawing set, all of which are just one pixel wide, still come out a little spotty, but we can live with that. ================================================================================ Note 309.24 Would you buy a used LN03R? 24 of 24 EISNER::SAUNDERS "John Saunders, Pacer Software, In" 39 lines 15-APR-1989 01:01 -< PostScript almost = PostScript >- -------------------------------------------------------------------------------- I just read an excellent book, "Real World PostScript(tm)", published by Addison-Wesley, Editied by Stephen F. Roth. It is a collection of papers on several aspects of the PostScript language, and it's use (and abuse) in the "Real World" of professional-quality printing. Chapter 3, by Michael Fryd, is on "Writing Device-Independent PostScript". It makes the distinction between Page Description Languages vs. Printer Control Languages. Basically PostScript is mostly a PDL, but has PCL elements in it (so a print symbiont can select paper trays, for instance). He points out that a frequent mistake is to confuse the two categories, and place printer-specific commands in the PostScript document. PostScript documents should describe what to print, not how to print it. Unfortunately, some of the most popular PostScript-generating applications in the world (e.g., Apple's LaserWriter driver) generate device *dependant* PostScript! This causes a flurry of excitement here at Pacer whenever Apple comes out with a new version of the LaserWriter driver and we have to rewrite their AppleDict procset for the LPS40 and LN03R :-(. So, is PostScript = PostScript? It should be, but it isn't always. John Saunders Sr. Software Engineer (Printing and Not-Printing) Pacer Software, Inc. P.S. Any of you out there writing applications which generate PostScript should *please* contact Adobe and get a copy of the documents: "Cooperative Printing, Guidelines for a distributed printing environment", PN LPS5030; "Document Structuring Conventions, Specification", PN LPS5001; and "PostScript Printer Description Files, Specification", PN LPS5003. You can get them through the mail (PO Box 7900, Mountain View, CA 94039-7900), or if you're on the Internet (or uucp), you can get them by sending a message containing the word "help", to decwrl!adobe!ps-file-server. Then stand back and watch the info fly!  ================================================================================ Note 310.0 KMV1A Receive Data LED States No replies EISNER::PEPE "Rich Pepe" 15 lines 24-FEB-1989 02:46 -------------------------------------------------------------------------------- KMV1A Receive Data LED State Question I have attempted to install a KMV1A in a MicroVAX II BA23 system. Currently the software, VAX-LINK for VMS, is not seeing data from the IBM Host. What puzzles me is that the Receive Data LED on the KMV1A is on solid. I would expect to see this indicater blink as the RCV indicater does on the breakout box. The indicater is on whether or not the RS232 cable is installed. > Is this normal to have the RCV indicater in a steady on condition? > If this is not normal, what may be the fault? A second KMV1A exhibited the same characteristic.  ================================================================================ Note 311.0 Info on FCO 11/750 - I008? 2 replies EISNER::BRYANT "Geoff Bryant" 4 lines 24-FEB-1989 14:21 -------------------------------------------------------------------------------- Does anyone have detailed information on FCO 11/750 - I008? Any information would be appreciated, including how/if it relates to VMS version 4.x and 5.x? ================================================================================ Note 311.1 Info on FCO 11/750 - I008? 1 of 2 EISNER::GRANZOW 24 lines 24-FEB-1989 17:04 -< vax 750 pcs >- -------------------------------------------------------------------------------- < Note 311.0 by EISNER::BRYANT "Geoff Bryant" > -< Info on FCO 11/750 - I008? >- < Does anyone have detailed information on FCO 11/750 - I008? < Any information would be appreciated, including how/if it relates < to VMS version 4.x and 5.x? Retrofit all 11750 systems to allow loading of PCS on power fail on all systems. Problem/Symptoms: 1: Patchable control store is not loaded at boot time. 2: UDA50 boot ROM code does not allow enough time for RA60 to spin up after power fail. 3: ROM code for TU58 does not allow enough time for TU58 to initialize. 4: Massbus boot ROM code has wrong hex code for RP07 drive. This FCO changes the boot ROM's on the L0011, L0016, L0022 Memory control modules. Also changes the 11750 Microcode. New console p/n BE-T204*-ME (* MIN REV. M) NOTE: If PCS is loaded in SYSSTARTUP.COM, the commands should be removed so that PCS will not be loaded twice. This FCO should have no bearing on VMS V5.x ================================================================================ Note 311.2 Info on FCO 11/750 - I008? 2 of 2 EISNER::TANNENBAUM "Saul Tannenbaum" 2 lines 25-FEB-1989 17:24 -< See also notes 187 and following in this conference >- -------------------------------------------------------------------------------- < Any information would be appreciated, including how/if it relates < to VMS version 4.x and 5.x?  ================================================================================ Note 312.0 DEC TUC02 TAPE DRIVE CLEANING KITS 6 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 1 line 27-FEB-1989 12:12 -------------------------------------------------------------------------------- What is Trichlorotrifloroethane? ================================================================================ Note 312.1 DEC TUC02 TAPE DRIVE CLEANING KITS 1 of 6 EISNER::KENNEDY "Terry Kennedy" 13 lines 27-FEB-1989 18:16 -------------------------------------------------------------------------------- > What is Trichlorotrifloroethane? Stuff for cleaning tape drives, of course 8-) Seriously, there should be a MSDS (Material Safety Data Sheet) in the box. It is required by law, and will tell you all you need to know about the stuff. Basically, is one of a family of degreasing agents. Others in this family are things like the various Freons (this is one of 'em) as well as some industrial degreasers, many of which have been banned or severe- ly restricted in recent years. This one is actually not that harmful, relatively speaking... ================================================================================ Note 312.2 DEC TUC02 TAPE DRIVE CLEANING KITS 2 of 6 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 9 lines 27-FEB-1989 18:36 -< FREON 113 >- -------------------------------------------------------------------------------- > Seriously, there should be a MSDS (Material Safety Data Sheet) in the I found the sheet buried under the cleaning pads. It is FREON 113. Interesting item on this five page "sheet"..... Other precautions - Freon 113 will attack some types of plastics, rubber and coatings. ================================================================================ Note 312.3 DEC TUC02 TAPE DRIVE CLEANING KITS 3 of 6 EISNER::KENNEDY "Terry Kennedy" 5 lines 27-FEB-1989 18:44 -------------------------------------------------------------------------------- > Other precautions - Freon 113 will attack some > types of plastics, rubber > and coatings. Which means *don't* use it on the capstan roller of the tape drive! ================================================================================ Note 312.4 DEC TUC02 TAPE DRIVE CLEANING KITS 4 of 6 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 13 lines 28-FEB-1989 15:08 -< May be OK on capstans. >- -------------------------------------------------------------------------------- >> > Other precautions - Freon 113 will attack some >> > types of plastics, rubber >> > and coatings. >> >> Which means *don't* use it on the capstan roller of the tape drive! Somehow, I suspect that if the tape drive vendor markets something for cleaning tape drives, the substance should be suitable for that purpose (but then, I always believe what the salespersons say). Do the instructions with the kit include any cautions about cleaning capstans? Possibly the capstans are not one of the materials affected. ================================================================================ Note 312.5 DEC TUC02 TAPE DRIVE CLEANING KITS 5 of 6 EISNER::HARVEY "Jack Harvey" 1 line 28-FEB-1989 21:44 -< OK on TU80 Capstans :-) >- -------------------------------------------------------------------------------- ================================================================================ Note 312.6 DEC TUC02 TAPE DRIVE CLEANING KITS 6 of 6 EISNER::MCCORMICK_B "Bob McCormick" 9 lines 6-MAR-1989 19:32 -< itsa potent drink! >- -------------------------------------------------------------------------------- Jeff: thats 3 chlorines 3 florides and an ethane... itsa potent drink! don't leave the lid off that stuff for long... its one drink that won't stay with you (long)  ================================================================================ Note 313.0 Looking for mouse prints 10 replies EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 9 lines 27-FEB-1989 13:28 -------------------------------------------------------------------------------- 1) Does anyone have the pin-outs for the old style (rectangular) mouse? This is the one with the DB-15 connector (only 10 pins installed). 2) With a suitable DB-15 to Micro-DIN-8 adapter cable, will the old style mouse work with a GPX (VCB02) controller? 3) Will the new style "corporate" mouse work with the older VCB01 controller? ================================================================================ Note 313.1 Looking for mouse prints 1 of 10 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 47 lines 28-FEB-1989 23:32 -< Mouse guts >- -------------------------------------------------------------------------------- DECUServe really has me spoiled. After not receiving a reply in under a day or so, I decided to answer my own questions. The results of the mouse dissection are: >> 1) Does anyone have the pin-outs for the old style >> (rectangular) mouse? This is the one with the DB-15 >> connector (only 10 pins installed). The old style (VS10X-EA) mouse is the Hawley model X063X made by the Mouse House of Berkeley, CA. The outputs are TTL signals directly out of a pair of 74279 quad RS flip flops. They are then decoded by the VCB01 controller. DB-15 pin Function 1 Up/Down signal 1 2 Up/Down signal 2 3 Right/Left signal 1 4 Right/Left signal 2 6 +5 volts 9 Ground/Shield 10 Ground/Shield 12 Right button 13 Center button 14 Left button All the signals are de-bounced by the RS flip flops. If you want a laugh, see what major company name is formed by turning the model number (X063X) upside down. >> 2) With a suitable DB-15 to Micro-DIN-8 adapter cable, will >> the old style mouse work with a GPX (VCB02) controller? No way. The new "corporate" mouse uses RS-232 compatible signals to communicate with the controller. The new mouse has intelligence built into it (sorta like the LK201 keyboard). There are only four wires used by the new mouse (+5, Ground, Serial data to mouse, Serial data from mouse). It runs at 4800 bps. >> 3) Will the new style "corporate" mouse work with the older >> VCB01 controller? It looks like it will, but I won't guarantee it. Does anyone have a corporate mouse I can test? If anyone wants more details, let me know. ================================================================================ Note 313.2 Looking for mouse prints 2 of 10 EISNER::KOZAM 15 lines 1-MAR-1989 00:45 -< 4800 baud? >- -------------------------------------------------------------------------------- > No way. The new "corporate" mouse uses RS-232 compatible > signals to communicate with the controller. The new mouse has > intelligence built into it (sorta like the LK201 keyboard). > There are only four wires used by the new mouse (+5, Ground, > Serial data to mouse, Serial data from mouse). It runs > at 4800 bps. Are you sure about 4800 bps? I thought that the mouse (VSXXX-AA) runs at 9600 baud. On the VAXstation 2000, both the keyboard and mouse appear to run at that speed. (This was discovered by dissecting the pin-outs on the DB-15 connector so that I could temporarily get 2 more serial lines in place of the fancy display and mouse. Both lines ran at 9600 baud.) ================================================================================ Note 313.3 Looking for mouse prints 3 of 10 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 4 lines 1-MAR-1989 00:54 -< A further question >- -------------------------------------------------------------------------------- > Serial data to mouse, Serial data from mouse). It runs ^^ BTW.. what serial data do you send a mouse? ================================================================================ Note 313.4 Looking for mouse prints 4 of 10 EISNER::HARVEY "Jack Harvey" 11 lines 1-MAR-1989 03:12 -------------------------------------------------------------------------------- > (This was discovered by dissecting the pin-outs >on the DB-15 connector so that I could temporarily get 2 more serial >lines in place of the fancy display and mouse. Both lines ran at 9600 I think you simply found the pins (5,13,14,15) that the DEC 423 converter (70-23766) on the MicroVAX 2000 uses, in addition to the VS2000 printer port, to create the three 423 ports in the MV2000 configuration. The DB-15 also has the three color video signals for the fancy display. The speed may depend on whether the box thinks it's a VS or a MV. ================================================================================ Note 313.5 Looking for mouse prints 5 of 10 EISNER::HARVEY "Jack Harvey" 2 lines 1-MAR-1989 03:16 -< Mine take any cereal data >- -------------------------------------------------------------------------------- > BTW.. what serial data do you send a mouse? ================================================================================ Note 313.6 Looking for mouse prints 6 of 10 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 49 lines 1-MAR-1989 13:51 -< Feeding and being fed by a mouse >- -------------------------------------------------------------------------------- > BTW.. what serial data do you send a mouse? I can't fully answer that, but this will give you some idea. From the VAXstation II/GPX Technical Manual, AZ-GNFAA-MN: ---------------------------------------------------------------------- The mouse uses optical-mechanical technology to detect movement of a rubber coated ball with a resolution of .005". The ball movement is translated into X and Y coordinate movement by two perpendicular shafts activated by the ball. The motion of a shaft, sensed by optical decoders, causes the two output bits for that direction to form waves in quadrature. The frequency of the waves is determined by the speed of the ball movement, and the phase (plus or minus 90 degrees) is determined by the direction the ball is traveling. A dedicated microprocessor inside the mouse converts the quadrature X-Y pulses to relative X-Y coordinates which are then transmitted to the VCB02 base module. The X-Y coordinates are transmitted to the VCB02 as a 3-byte position report at a 4800 baud rate. The mouse has two operating modes that determine when and how often the mouse transmits a position report. The operating modes are incremental Stream and Prompt. Prompt mode is the power-up default. In Prompt Mode, the mouse will only generate a report in response to a Request Mouse Position command. The mouse can respond to over 95 position requests per second. In Incremental Stream Mode, the mouse will generate reports at 55 Hz intervals any time there is movement or a change in button position since the last report. If the mouse is motionless and no buttons have changed, no report is generated. The report rate under continuous movement or button change is 55 reports per second. The operating modes are selected by software sending the proper ASCII commands. The mouse also contains self-test and identification logic which is activated by command from the computer or upon mouse power up. Upon completion of the self-test, the mouse transmits a self-test report consisting of a 2-byte identification code plus a 2-byte code describing the status of the mouse electronics and firmware. For further information, refer to the VCB02 Video Subsystem Technical Manual (AZ-GLGAA-MN). ---------------------------------------------------------------------- Unfortunately, the Electronic Store shows AZ-GLGAA-MN (as well as AZ-*) to be an invalid part number. Does anyone have a copy of this manual? ================================================================================ Note 313.7 Looking for mouse prints 7 of 10 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 6 lines 1-MAR-1989 23:37 -< Are there any 'dumb' devices left? >- -------------------------------------------------------------------------------- << -< Mine take any cereal data >- Somehow, I _KNEW_ someone would jump on that straight line! At any rate... based on Alan's findings, guess the ASCII input line makes sense. Amazing where smarts are showing up! ================================================================================ Note 313.8 Looking for mouse prints 8 of 10 EISNER::ERSKINE "Chris Erskine" 1 line 2-MAR-1989 07:13 -< When you see mouse prints - Set a trap to catch it >- -------------------------------------------------------------------------------- ================================================================================ Note 313.9 Looking for mouse prints 9 of 10 EISNER::MERRIMAN "George Merriman - CCA/NY" 1 line 2-MAR-1989 15:20 -< A friendly word from your moderator >- -------------------------------------------------------------------------------- Enough mouse jokes! ================================================================================ Note 313.10 Looking for mouse prints 10 of 10 EISNER::MCCORMICK_B "Bob McCormick" 15 lines 6-MAR-1989 19:43 -< Serious and not so serious reply >- -------------------------------------------------------------------------------- | BTW.. what serial data do you send a mouse? well ... it must be either a "MBn UP EVENT" command or ... a command that generates very high voltage (via one of those tiny power supplies they use in night vision goggles: +few volts to +10,000 volts) ---> sorry, couldn't resist ... seriously now: The mouse doesn't have to be the only pointing device... The corporate tablet has a four button puck and a two button pen (point and small button on the side) that is mouse exchangeable. Serial data OUT to the pointing device can control its tracking rate (how many samples per whatever it generates), etc.  ================================================================================ Note 314.0 VAXstation Video and Mouse Connections 15 replies EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 30 lines 27-FEB-1989 13:44 -------------------------------------------------------------------------------- I need some help with the VR260 (monochrome) monitor as used with VAXstations, both the original (VCB01) and GPX (VCB02) versions. A friend came over yesterday with a VAXstation-II/GPX that he had picked up at an auction, minus the monitor and cables. The documentation shows that the cable (BC18P-10) for a monochrome system is the same as that used on the non-GPX (VCB01) VAXstation, so we hooked my cable and monitor to his system. There is an image on the screen, but there appears to be a video sync problem. The data bits (dots) appear to start while the beam is still in the middle of the horizontal retrace. The vertical sync is OK. This means that the first few characters on a line read from right-to-left and are spread out (while the beam is retracing). The rest of the line reads normally. I attempted to adjust the monitor to accomodate the GPX board, but to no avail. There are no adjustments or jumpers I can find on the GPX board. Substituting another VR260 monitor showed exactly the same problem. By grossly mis-addjusting the VR260 horizontal sync it is possible to get a readable image on the screen and we found nothing functionally wrong with the system (VMS v5.0-2 and VWS v4.0). Is there a special version of the VR260 monitor for GPX systems? Is there a jumper or switch I missed somewhere? What is different between the video signals from the VCB01 and VCB02? ================================================================================ Note 314.1 VAXstation Video and Mouse Connections 1 of 15 EISNER::HARVEY "Jack Harvey" 9 lines 28-FEB-1989 06:25 -< Use Green Color Cable >- -------------------------------------------------------------------------------- My company is in the middle of a location move and the world has been randomized, so I can't check this for you. However, I believe I saw your exact condition while fooling around with a B/W monitor on a color VAXstation 2000. As I recall, the solution is to use a color cable and to connect the GREEN BNC connector to the B/W monitor. This line apparently has the proper sync signal. I think DEC mentions this someplace in the VS2000 setup manual. ================================================================================ Note 314.2 VAXstation Video and Mouse Connections 2 of 15 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 8 lines 28-FEB-1989 12:45 -< I would expect sync to be on the green signal >- -------------------------------------------------------------------------------- >> As I recall, the solution is to use a color cable and >> to connect the GREEN BNC connector to the B/W monitor. I thought that was what the BC18P-10 cable did. Unfortunately, we don't have a color cable available. We are willing to hack up a B/W cable if necessary, but would pefer not to. Does anyone have the pinouts for the BC18Z (color) cable? ================================================================================ Note 314.3 VAXstation Video and Mouse Connections 3 of 15 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 21 lines 28-FEB-1989 23:42 -< Zeroing in on the video signals >- -------------------------------------------------------------------------------- >> As I recall, the solution is to use a color cable and >> to connect the GREEN BNC connector to the B/W monitor. I took a GPX cabinet kit, a BC18P (monochrome) cable, and an ohmmeter and now have an almost complete understanding of what signals are on what pin at both ends of the BC18 cable. The only thing I am lacking is the DB15 pins that the COLOR signals are on. If anyone can spare a few minutes with an ohmmeter and their BC18Z cable, I would like to know the pins used for the RED, BLUE and GREEN signals. They will be on pins 9, 10, and 11 for the center pins of the three BNC connectors. I just want to know which one is which. When I have that, I will post a complete pin/signal list. Oh, one other thing to look for: are pins 10-13 used in the BC18Z cable? They are not in the BC18P. Thanks in advance, Alan ================================================================================ Note 314.4 VAXstation Video and Mouse Connections 4 of 15 EISNER::HARVEY "Jack Harvey" 9 lines 1-MAR-1989 12:48 -< Partial Answer? >- -------------------------------------------------------------------------------- A element of confusion, Alan: my color cable is PC19S, molded in the mounting block. GREEN 10 RED 1 BLUE 11 I'll try to resolve the confusion and check out the rest of the lines in this thing. (I need the info too.) ================================================================================ Note 314.5 VAXstation Video and Mouse Connections 5 of 15 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 16 lines 1-MAR-1989 13:46 -< What workstation is the PC19S cable for? >- -------------------------------------------------------------------------------- >> A element of confusion, Alan: my color cable is PC19S, molded >> in the mounting block. >> >> GREEN 10 >> RED 1 >> BLUE 11 It sounds like your cable is not from a VAXstation-II/GPX. The GPX cabinet kit I have has three pieces of coax going from the controller to the DB15 connector. They are paired as (center/shield) 11/1, 10/2, and 9/3. What is your cable on? When I get the answer to the remaining questions, I will post the definitive pin-outs for the cables, with each wire's function. ================================================================================ Note 314.6 VAXstation Video and Mouse Connections 6 of 15 EISNER::HARVEY "Jack Harvey" 33 lines 1-MAR-1989 19:35 -< BC19S Color Cable Connections >- -------------------------------------------------------------------------------- The confusion is that my BC19S is for the VAXstation 2000 with color board. To add to the confusion, the B/W VS2000 uses the BC18P, same as your box. I will post my data anyway in hopes that some correlation will be found. Readers are warned this is *not* a direct answer to Alan's question. Wiring of the BC19S color cable for the Vaxstation 2000 2,6,7-o o-5 DB-15 connections to mouse 13-o o-12 (DIN female socket view) 4-o o o | 8 |--| |---| |--| | | DB-15 connections to keyboard | | (RG female socket view) | o o o o | | | | 14 2 15 6,7 BNC center pins: Red - 1 Green - 10 All BNC shields - 2,6,7 Blue - 11 ================================================================================ Note 314.7 VAXstation Video and Mouse Connections 7 of 15 EISNER::HARVEY "Jack Harvey" 3 lines 1-MAR-1989 19:42 -< Help! >- -------------------------------------------------------------------------------- We still need someone with a real VAXstation II/GPX to buzz out the cable to the monitor/mouse/keyboard so we can nail this. ================================================================================ Note 314.8 VAXstation Video and Mouse Connections 8 of 15 EISNER::HARVEY "Jack Harvey" 41 lines 2-MAR-1989 02:42 -< VS2000 Video and Mouse Details >- -------------------------------------------------------------------------------- The VAXstation 2000 Hardware Installation Guide (October 1987) page 31: " VR260 Monochrome Monitor with 4-plane Graphics Coprocessor " " 1. Insert ONLY the G video cable connector of the color " video cable assembly (BC19S) into the socket on the back of " the monitor... " " Note: The R and B connectors will hang freely - do not " connect them to anything. Similiar text occurs for the VR150 (14") monochrome monitor. Buzzing the BC18P (monochrome) cable shows that the single B/W BNC center conductor goes to DB-15 pin 9. Scoping the video on this line shows a wimpy and ringing signal. This is what one gets if using the BC18P on a color VS2000. When a BC19S cable is installed, the Blue and Red BNC's have good video signals, but the horizontal sync pulse does not always occur: the scope shows a faint trace at a non-zero level. When the display times out (turns off due to lack of user activity, i.e., CRT saver) the Blue and Red signals stop altogether: zero. The Green signal, however, has a distinct 1.84 us sync pulse that is solid on the scope and continues during the display time out. -------- Mouse Serial Data When the MV/VS2000 motherboard is in MV configuration, the DB-15 connector has RS232 signals for TTA1:. The RD (input to computer) signal is on pin 5; SD (output from computer) is on pin 13. Both of these pins go to the mouse DIN connector. See .6 for pinout details. One can then speculate that the mouse is on "TTA1" when in VS configuration. ================================================================================ Note 314.9 VAXstation Video and Mouse Connections 9 of 15 EISNER::MCCORMICK_B "Bob McCormick" 26 lines 6-MAR-1989 20:01 -< Two things about the VR260 >- -------------------------------------------------------------------------------- Two notes regarding this thread: Its been two years since I've been to the mfg location for the VR260, but if things are still the same, you're better off not attempting to make a lot of adjustments to the VR260 in the field. If you take the plastic skin off, and open the metal flap (door) to get to adjustments you will find that they will change when you close the flap. During manufacture they peform these adjustments in a "faraday field" type arrangement: a large (I'd estimate) 8 foot cube frame with wires wound around it. The CRT and the adjuster are in the center of the cube, and things are set-up for proper compensation. Then again, if you don't care about accuracy... Second: Although they say there is a maximum length to cables, I was VERY successfull in locating a VAXstation (with VCB01 in BA123) A W A Y from my work area. I had a 40 metre cable with the coax and twisted pair made up for the monitor and keyboard. I made two very minor adjustments to the length of the coax for the video (ie I crimped a connector on, tried it out, chopped it off, crimped another connector on... and the 3rd one was fine). I also made an extension cable for the "old" brick style mouse. It was very nice to sit at my desk without the noise and heat of the BA123 box ... and use my VAXstation! ================================================================================ Note 314.10 VAXstation Video and Mouse Connections 10 of 15 EISNER::HAHN "DECUServe Vice Chair" 2 lines 15-MAR-1989 07:00 -< VAX 3500 to VR260 mono long extension >- -------------------------------------------------------------------------------- So if I want to move a VR260 monochrome monitor 75feet from a VAXStation 3500 box, what kind of cable will I need? ================================================================================ Note 314.11 VAXstation Video and Mouse Connections 11 of 15 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 6 lines 16-MAR-1989 12:43 -< Just a reminder: I still need the pin numbers. Please help. >- -------------------------------------------------------------------------------- >> If anyone can spare a few minutes with an ohmmeter and their >> BC18Z cable, I would like to know the pins used for the >> RED, BLUE and GREEN signals. They will be on pins 9, 10, >> and 11 for the center pins of the three BNC connectors. >> I just want to know which one is which. When I have that, >> I will post a complete pin/signal list. ================================================================================ Note 314.12 VAXstation Video and Mouse Connections 12 of 15 EISNER::HAHN "DECUServe Vice Chair" 5 lines 24-MAR-1989 09:18 -< What kind of cable >- -------------------------------------------------------------------------------- For my long extension can I just add a pin to pin extension cable DB15 to DB15 between the VAXStation 3500 and the 10ft cable that connects to the VR260. Or - is/are there one or more coax in the bundle? ================================================================================ Note 314.13 VAXstation Video and Mouse Connections 13 of 15 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 21 lines 24-MAR-1989 15:03 -< Use coax for these pins >- -------------------------------------------------------------------------------- >> For my long extension can I just add a pin to pin extension cable >> DB15 to DB15 between the VAXStation 3500 and the 10ft cable that >> connects to the VR260. Or - is/are there one or more coax in >> the bundle? You should have 75 ohm coax for the following DB15 pins on a BC18Z cable. These were determined by looking at a GPC cabinet kit. Signal Shield 9 3 10 2 11 1 Those are the three video signals (R, G, B). I am still waiting for someone with a GPX system to tell me which color is on which pin. My monochrome (BC19P-10) cable only uses the pin 9/3 pair. I still don't know why my BC19P cable does not seem to properly work for connecting a VR260 to a VAXstation-II/GPX. As soon as I get it figured out, I will post the complete pin-outs for the various cables and connectors. ================================================================================ Note 314.14 VAXstation Video and Mouse Connections 14 of 15 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 53 lines 18-APR-1989 15:44 -< The answer is ... >- -------------------------------------------------------------------------------- Okay I got the info on the BC18Z cable (from a GPX upgrade kit that included a VR290 19" color monitor) !! First, lets define the pins on the connector at the CRT end. When holding the cable with the knob on the top and the keyboard and mouse connectors facing you the "fingers" in the keyboard connector will be on the top and the "notch" in the mouse connector will also be on the top. I define pin 1 of the mouse connector to be directly beneath the "notch" with increasing numbers in a clockwise rotation. I define pin 1 of the keyboard connector to be on the left and increasing number as you move to the right. The DB15 is label (get a magnify glass !!) DB15 Opposite end 1 Case Gnd 2 Case Gnd 3 Case Gnd 4 M 2 5 M 5 6 Case Gnd 7 Case Gnd 8 Signal Gnd 9 RED 10 GREEN 11 BLUE 12 M 6 13 M 3 14 K 4 15 K 1 Keyboard DB15 1 15 2 Case Gnd 3 Signal Gnd 4 14 Mouse DB15 1 Signal Gnd 2 4 3 13 4 Case Gnd 5 5 6 12 7 N/C I hope this helps ! ================================================================================ Note 314.15 VAXstation Video and Mouse Connections 15 of 15 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 15 lines 19-APR-1989 11:45 -< Thanks for the valuable info! >- -------------------------------------------------------------------------------- >> DB15 Opposite end >> 9 RED >> 10 GREEN >> 11 BLUE OK, that explains it! Since the BC18P (monochrome) cable has the video on pin 9, that means that the monitor is trying to sync on the RED signal instead of GREEN. I will combine this with all the cable information I have and post it here in a few days. Nit-pickers note: the micro-DIN pin numbers in the previous note are not the ones used in DEC's documentation. When I post the combined info, I will change to that.  ================================================================================ Note 315.0 PRO-380 to VR201 cable length limits? 1 reply EISNER::SCOPELLITI "Whatsa behind is uva no importan" 7 lines 28-FEB-1989 00:07 -------------------------------------------------------------------------------- Anyone know how long the cable linking a VR201 monitor to a PRO 380 (aka 8530/8550/8700/8800/8810N/8820N console) can be? We'd like to work around the cable length limitation of the RTI cable hooking the PRO-Console to the VAX 8XXX. In particular, we'd like to place the PRO-380 on the system box and the monitor/keyboard about 30 feet away. ================================================================================ Note 315.1 PRO-380 to VR201 cable length limits? 1 of 1 EISNER::CROWELL "Shefth of the Fourth Order" 2 lines 28-FEB-1989 17:31 -< smae only different >- -------------------------------------------------------------------------------- I've run my Pro-350 with as much as 25 ft. of cable linking to a VR241. No problems yet.  ================================================================================ Note 316.0 Fiber Optics / DECServer / Connectivity 4 replies EISNER::BYRNE_C "Charlie Byrne" 165 lines 28-FEB-1989 00:31 -------------------------------------------------------------------------------- <<< EISNER::DUA0:[NOTES$LIBRARY]VMS.NOTE;1 >>> -< VMS and bundled utilities >- ================================================================================ Note 155.25 Need help with Console Concentrator! 25 of 32 EISNER::MCKEE "Randy McKee - Texaco, Inc" 17 lines 24-FEB-1989 10:29 -< Are the FOCAs required for the VCC? >- -------------------------------------------------------------------------------- We're considering getting a VAX Cluster Console for about 8 systems, possible using a VAXStation 3100 as the hardware. (Much cheaper than the uVAXes discussed in this conference previously!) Anyway, DEC is telling us we need to buy all this fiber optics equipment to hook up our computers, even though we have all our computers within about 30 feet of one another. Our feeling is, we ought to be able to forgo the fiber optics connectors (which seem to go from RS-232 to fiber adpator to fiber adaptor back to RS-232) and go directly from RS-232 to RS-232. Note 155.13 makes this look possible. Is this true? Can we just hook up the VAXes we want to manage directly to, say, a DECServer 200's terminal ports and punt the fiber stuff altogether? Randy McKee Texaco, Inc. (We're working to keep your trust!) (713)954-6046 ================================================================================ Note 155.26 Need help with Console Concentrator! 26 of 32 EISNER::MAYHEW "Bill Mayhew" 3 lines 24-FEB-1989 11:29 -< Not sure why, but... >- -------------------------------------------------------------------------------- I remember reading somewhere in the distant past that one of the reasons for specifying a fiber-optic interconnect for this product was to keep the grounds of the various systems electrically isolated. ================================================================================ Note 155.27 Need help with Console Concentrator! 27 of 32 EISNER::WOOD "Jim Wood (609)423-1040" 8 lines 24-FEB-1989 21:33 -< It's possible >- -------------------------------------------------------------------------------- Our people replaced the fiber optics with the funky cable-with-the-switch and have had no problems at all. In fact, our DEC field service rep researched the part number to order extra funky cables and helped install the system during the 8700 installation last year. Unless things have changed in the past 18 months, you'll need to secure an unlisted parts quote for the cables, then order 'em. If you need details, I'll ask the system manager of the system on Monday. ================================================================================ Note 155.28 Need help with Console Concentrator! 28 of 32 EISNER::COY "Dale E. Coy" 42 lines 24-FEB-1989 23:32 -< Not to scare you - just to lay out the risks >- -------------------------------------------------------------------------------- From a technical viewpoint, Bill is right. The reason for the fiber is to keep from connecting multiple system grounds together through the VCS. I would GUESS that, if all of your machines are within 30 feet of each other, and are all fed from the same electrical distribution system, then you'd be OK. But I'm not there to check it, and don't ask me to write an insurance policy. There are actually 2 grounds involved ("frame" and "signal"), but I'll ignore that for the moment. Usually, connecting the grounds together doesn't cause problems. What I'm saying is that there is a LOW probability of problems. If you do have problems, they may or may not show up instantly. If things are just a "little" bad, it could show up as trash on the VCS lines, or even on other terminal lines - or intermittant strangeness anywhere on the connected systems. VERY hard to diagnose. If you have big-time problems with differences in ground potential, then the connecting cables may be carrying large amounts of current - or could burn up - or you might "smoke" some hardware. Even if everything is OK now, it's possible for a component failure in the future to cause one or more of these to crop up. Now - that sounds like a horror story. As I said, it's LOW probability. Your qualified electrician or technician can make some measurements and tell you how things stand right now - but make sure he/she understands the question. And it doesn't guarantee the future. Would I do it? Probably, for most installations. But that's MY opinion, and I have NO idea what the reliability/availability or value/risk/benefit situation looks like at YOUR place. BTW, if you decide to do it, have a talk with your maintenance outfit. Make sure you know whether they will continue to respond to calls about flakey communications, whether they might suddenly refuse to replace a smoked component, etc. ================================================================================ Note 155.29 Need help with Console Concentrator! 29 of 32 EISNER::TINKELMAN "Bob Tinkelman - CCA" 13 lines 25-FEB-1989 00:11 -< Console-ports to DS200 to VCC works just fine >- -------------------------------------------------------------------------------- RE: < Note 155.25 by EISNER::MCKEE "Randy McKee - Texaco, Inc" > > We're considering getting a VAX Cluster Console for about 8 systems, >... > Can we just > hook up the VAXes we want to manage directly to, say, a DECServer > 200's terminal ports and punt the fiber stuff altogether? Yup. We did that for one client. It was last summer so some of the details are a little fuzzy. We had seven microVAX IIs that had been managed using a long table with 7 VT200s lined up (keyboards at an angle so they'd all fit). We ran the console cables to ports on a DS200 and used VCC on a VS200 to manage all 7 systems. ================================================================================ Note 155.30 Need help with Console Concentrator! 30 of 32 EISNER::MERRIMAN "George Merriman - CCA/NY" 13 lines 25-FEB-1989 17:03 -< a real-life horror story >- -------------------------------------------------------------------------------- > If you have big-time problems with differences in ground potential, > then the connecting cables may be carrying large amounts of current - > or could burn up - or you might "smoke" some hardware. Once upon a time in a diode factory I worked for the transformer in the power supply of a diode tester developed a primary to secondary short which superimposed the 120v AC line voltage on the 5v logic supply. This caused considerable damage to the tester in general, but the most spectacular damage was to the interface card in the tester and its mate in the HP computer that ran it. When I opened up the cabinet I could see the IC packages on the interface board glowing like little coals -- same thing in the computer. In this case optical isolation would have saved us a LOT of money. ================================================================================ Note 155.31 Need help with Console Concentrator! 31 of 32 EISNER::KOZAM 11 lines 25-FEB-1989 22:54 -< How about fuses? >- -------------------------------------------------------------------------------- > If you have big-time problems with differences in ground potential, > then the connecting cables may be carrying large amounts of current - > or could burn up - or you might "smoke" some hardware. How about putting some fuses into the RS-232 line. They'd probably be VERY low current so that they would blow quickly in case some mishap occurred. I haven't tried this - it's just an idea. I too have seen flakey terminals and printers take out their respective computer ports. Fortunately, only the driver chips are usually zapped and replacing them is a snap. ================================================================================ Note 155.32 Need help with Console Concentrator! 32 of 32 EISNER::TRAYSER "VMS Conference Moderator" 8 lines 27-FEB-1989 23:02 -< Lets get back on track. Power discussion moved to HW conference >- -------------------------------------------------------------------------------- The topic police have decided that this discussion is no longer related to VMS, but rather to Hardware and 'connectivity'. As such, I suggest the small sub-discussion concerning Power problems be redirected to the Hardware_Help conference and we get back on the track of discussing the VAX Console Software. Thanks! $ ================================================================================ Note 316.1 Fiber Optics / DECServer / Connectivity 1 of 4 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 28 lines 28-FEB-1989 15:50 -< VCS with ModTap + DS100 >- -------------------------------------------------------------------------------- (Skip 2 paragraphs to avoid the nattering) First off, I want to bitch about transplanted notes. Maybe it is just me, but I find them MUCH harder to read through and make sense out of, as the usual controls don't work when they are all mushed together. I didn't see any note as to why this was done here, but I, for one, would prefer to NOT see this happen much at all, and would prefer simple pointers to the other stream. There isn't even any mention of where this all was uprooted from. >Yup. We did that for one client. It was last summer so some of the >details are a little fuzzy. We had seven microVAX IIs that had been >managed using a long table with 7 VT200s lined up (keyboards at an angle >so they'd all fit). We ran the console cables to ports on a DS200 and >used VCC on a VS200 to manage all 7 systems. It was done with ModTap adapters, no less, and the modular cords were then collected into harmonicas so 8 ports could be plugged together with a 25 pair cable. This allowed connecting ALL systems to the DecServer for VCS use, and moving one 25 pair plug a couple of inches put all cpus onto their own private CRTs. Dec's fiber system has a switch per cpu to reconnect a normal console, or switch to VCS. The all-or-none mode was fine for our application, and actually the other side of the harmonica where we were moving the 25 pair cable, had the individual modular cords that would allow any individual reconfiguration desired. ================================================================================ Note 316.2 Fiber Optics / DECServer / Connectivity 2 of 4 EISNER::BYRNE_C "Charlie Byrne" 13 lines 1-MAR-1989 21:11 -< A Brief Aside, now back to the subject >- -------------------------------------------------------------------------------- I agree transplanted notes are a pain. > < Note 316.1 by EISNER::BRUCE "Barton F. Bruce - CCA/MA" > > There isn't even any mention of where this all was > uprooted from. ||| VVV > <<< EISNER::DUA0:[NOTES$LIBRARY]VMS.NOTE;1 >>> > -< VMS and bundled utilities >- >Note 155.25 Need help with Console Concentrator! 25 of 32 ^^^^^^ |||||| ================================================================================ Note 316.3 Fiber Optics / DECServer / Connectivity 3 of 4 EISNER::HARVEY "Jack Harvey" 20 lines 2-MAR-1989 03:17 -< Simple Optical Isolators? >- -------------------------------------------------------------------------------- > I agree transplanted notes are a pain. We all agree. VAXnotes needs to support this, but that's off topic. This thread was clearly getting into hardware issues and now that it's here, I can reply. I always thought use of fiber optics to provide power ground isolation was world class over-kill. We could have a direct lightning strike on one node without bothering the other nodes via the cluster console. Of course the star coupler and/or Ethernet will flash over anyway... Low voltage isolation is, of course, a desirable goal. There must be simple optical isolators available that will do this. I made a quick scan of Black Box and didn't find a direct answer, however. One could probably use two of their RS232/Current loop converters back to back. (See bottom of page 120, Jan 89 BB catalog.) This may provide adequate power ground isolation, but only RD and SD signals. Are other RS232 control leads required? ================================================================================ Note 316.4 Fiber Optics / DECServer / Connectivity 4 of 4 EISNER::COY "Dale E. Coy" 15 lines 2-MAR-1989 23:28 -< Universal Rule: There are no universal rules >- -------------------------------------------------------------------------------- > I always thought use of fiber optics to provide power ground > isolation was world class over-kill. Of course it is - in most systems, the probability of a problem like this is VERY low. However, it's NOT ZERO. Some people have world-class systems, and a few thousand dollars for this is money well spent. I once worked on a system where continuous operation was worth millions of dollars. You do "strange" things in that case. For instance, contrary to usual practice, the fire protection system did NOT shut down the computer. If I had used a VCS, it would have been optically isolated. As you say, though, for most systems it's overkill.  ================================================================================ Note 317.0 Where to find it? 2 replies EISNER::MAYHEW "Bill Mayhew" 2 lines 4-MAR-1989 18:22 -------------------------------------------------------------------------------- This topic is related to where to find obscure DEC-computer-related parts. Commercialism guidelines apply, of course. ================================================================================ Note 317.1 Where to find it? 1 of 2 EISNER::MAYHEW "Bill Mayhew" 21 lines 4-MAR-1989 18:25 -< Keycaps >- -------------------------------------------------------------------------------- Prompted by a comment in the DECUS_SYMPOSIA conference... < Note 21.12 by EISNER::HARVEY "Jack Harvey" > -< Practical Giveaways >- > o Keyboard key caps. RECYCLE at the Boston Children's Museum often has bins with hundreds or thousands of LK201 keycaps. Some of these appear to me to have minor printing flaws (i.e. they're "irregulars") though not nearly as irregular as some I've seen on actual LK201s. RECYCLE collects industrial scrap and sells it by the bag (a few bucks for a grocery bag) primarily to teachers and other such folks for use in arts and crafts projects, or what have you. There are similar endeavors at other museums (esp. children's museums) across the country. One of (or is it *the*) Digital's LK201 plants is located nearby, in Roxbury. I have no connection with RECYCLE or the Boston Children's Museum, though I used to work there. ================================================================================ Note 317.2 Where to find it? 2 of 2 EISNER::MERRIMAN "George Merriman - CCA/NY" 4 lines 7-MAR-1989 17:31 -< look in HARDWARE_HELP 267 for more information >- -------------------------------------------------------------------------------- Note 267 in this conference relates to hard-to-find parts. Please post your replies there. I have set this note nowrite. George Merriman -- HARDWARE_HELP co-moderator  ================================================================================ Note 318.0 Q-bus interfaces on an 8250 5 replies EISNER::DRIESMAN 4 lines 6-MAR-1989 09:33 -------------------------------------------------------------------------------- Has anyone tried to install Q-bus interfaces on an 8250? The interfaces in question are the KMV1A and IEQ11. From the handbooks I have read, it appears that both a UNIBUS adapter and a Q-bus adapter are required. ================================================================================ Note 318.1 Q-bus interfaces on an 8250 1 of 5 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 9 lines 6-MAR-1989 11:18 -< Two bus converters is probably not the answer >- -------------------------------------------------------------------------------- >> Has anyone tried to install Q-bus interfaces on an 8250? >> The interfaces in question are the KMV1A and IEQ11. >> From the handbooks I have read, it appears that both a UNIBUS adapter and >> a Q-bus adapter are required. I have severe doubts about this working successfully. The handshake delays through two bus converters would probably cause bus time-outs. You would probably find that it would work some of the time, but fail when activity was high. ================================================================================ Note 318.2 Q-bus interfaces on an 8250 2 of 5 EISNER::PROVOST 21 lines 7-MAR-1989 10:31 -< QBUS timeout worse than UNIBUS timeout >- -------------------------------------------------------------------------------- >> Has anyone tried to install Q-bus interfaces on an 8250? >> The interfaces in question are the KMV1A and IEQ11. >> From the handbooks I have read, it appears that both a UNIBUS adapter and >> a Q-bus adapter are required. The arbitration on the UNIBUS and QBUS require a response from the processor within a short period of time after an address is sent. Since the timeout is specified in the device, the BI machine cannot override it. The BI UNIBUS adapter wins the race by "stalling" the BI so that it is guaranteed a slot for the return signal. This causes BI performance to be downgraded when UNIBUS activity is occurring. The QBUS has a shorter timeout, so the problem gets worse. This, in theory, is why a QBUS adapter is not available for the BI. Adding a QBUS adapter to a UNIBUS which is attached through a BI UNIBUS adapter adds the delays but still must meet the timeout. Result is not likely to work. Tom ================================================================================ Note 318.3 Q-bus interfaces on an 8250 3 of 5 EISNER::CROWELL "Shefth of the Fourth Order" 3 lines 12-MAR-1989 10:55 -< BI-Q >- -------------------------------------------------------------------------------- I don't know whether they've announced it as a product, but AEON Systems in Abluquerque was working on a BI-Q adapter. (They've already done BI-VME, BI-CAMAC, etc.) They'll no doubt be at DEXPO. ================================================================================ Note 318.4 Q-bus interfaces on an 8250 4 of 5 EISNER::HASSINGER "Bob Hassinger" 1 line 13-MAR-1989 14:49 -< Is their BI interface blessed (licensed) by DEC? >- -------------------------------------------------------------------------------- ================================================================================ Note 318.5 Q-bus interfaces on an 8250 5 of 5 EISNER::CROWELL "Shefth of the Fourth Order" 5 lines 16-MAR-1989 09:59 -< A Happy BI licensee >- -------------------------------------------------------------------------------- Is their BI interface blessed (licensed) by DEC? Yes, They are one of the happier DEC CMPs. >  ================================================================================ Note 319.0 TA78 => TEU78??? 8 replies EISNER::PROVOST 9 lines 8-MAR-1989 11:21 -------------------------------------------------------------------------------- We have a TEU77-AB on a MASSBUS adapter on a VAX-11/780. We can get "free" a TA-78 with three TU78's. We have no HSC. Can I downgrade a TA-78 to a TEU78 using parts from TEU77 cheaply? Anyone have a TEU78 they want to swap for one which was upgraded at great expense to a TA-78? ================================================================================ Note 319.1 TA78 => TEU78??? 1 of 8 EISNER::KENNEDY "Terry Kennedy" 13 lines 8-MAR-1989 14:30 -< Good news / Bad news >- -------------------------------------------------------------------------------- > We can get "free" a TA-78 with three TU78's. See comments at end. > Can I downgrade a TA-78 to a TEU78 using parts from TEU77 cheaply? Nope, the bad news is that the formatter in a TU78 is a TM78, while that in a TU77 is a TM03. However, you can pick up a whole used TM78 formatter for about $800 these days. Does that qualify as cheaply? Be careful with TU78's - you can go broke applying the FCO kits to make them work at 6250. If you want to know what the quick checks are to see if the drives are up to rev, drop me a Mail message... ================================================================================ Note 319.2 TA78 => TEU78??? 2 of 8 EISNER::PROVOST 21 lines 9-MAR-1989 10:57 -< TA has formatter? >- -------------------------------------------------------------------------------- re: -< Good news / Bad news >- >> Can I downgrade a TA-78 to a TEU78 using parts from TEU77 cheaply? > Nope, the bad news is that the formatter in a TU78 is a TM78, while that >in a TU77 is a TM03. However, you can pick up a whole used TM78 formatter >for about $800 these days. Does that qualify as cheaply? TA78 should have a GCR Formatter. TU78's are slaves. What I need from TEU77 is interface to MASSBUS. ... or am I missing something... > Be careful with TU78's - you can go broke applying the FCO kits to make >them work at 6250. If you want to know what the quick checks are to see >if the drives are up to rev, drop me a Mail message... TA78 and TU78's all recently rebuilt from ground up by 3rd party maintenance vendor. TEU77 out of rev, so interface to MASSBUS may need upgrade. MASSBUS adapters we have in profusion. ================================================================================ Note 319.4 TA78 => TEU78??? 4 of 8 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 14 lines 9-MAR-1989 13:26 -< More data on TA/TU78 >- -------------------------------------------------------------------------------- The TU78-AB(AD) and the TA78-BF(BJ) MASTER drives differ only in that the TU78 formatter connects to the Massbus, while the TA78 formatter connects to an HSC40/50/70. The TU78-AF(AJ) SLAVE drives can be connected to EITHER a TU78 or TA78 master. The TEU78-FB(FD) is a TU78-AB(AD) master drive with an 11/780 Massbus adaptor included as a kit. We have had out TU78 for about two years now with very few problems, it has been MUCH less of a hassle than out TU77. It worked well at 6250 from day 1. (Of course, our DEC hands have upgraded it to TU79 specs according to the FCO's, but it gave little difficulty even before the intestine transplant.) ================================================================================ Note 319.5 TA78 => TEU78??? 5 of 8 EISNER::KENNEDY "Terry Kennedy" 15 lines 9-MAR-1989 21:08 -< More info >- -------------------------------------------------------------------------------- > TA78 should have a GCR Formatter. TU78's are slaves. > What I need from TEU77 is interface to MASSBUS. > ... or am I missing something... That's what I get for not using full model numbers... The base drive is the same for TU78 slave, TU78 master, TA78 master. The only differ- ence is the formatter in the TU78 master speaks Massbus and the TA78 master speaks SDI. The TU77 master uses a completely different set of cards. What you need is the complete TM78 formatter assembly, as I said in .1. Well, to see what sort of a job your 3rd-party refurb house did, look for a brown gasket at the bottom edge of the supply reel housing. Also, open the transport and look at the read preamp card. Does it have a metal shield completely covering the read preamps? ================================================================================ Note 319.6 TA78 => TEU78??? 6 of 8 EISNER::COY "Dale E. Coy" 5 lines 9-MAR-1989 21:44 -< Brown gasketitis >- -------------------------------------------------------------------------------- > for a brown gasket at the bottom edge of the supply reel housing. Of course, I spent 3 years trying to get our DEC FS folks to put in the brown gasket - and even gave them the FCO numbers - and still didn't succeed. ================================================================================ Note 319.7 TA78 => TEU78??? 7 of 8 EISNER::KENNEDY "Terry Kennedy" 7 lines 10-MAR-1989 02:03 -< FCO has been cancelled >- -------------------------------------------------------------------------------- > Of course, I spent 3 years trying to get our DEC FS folks to put > in the brown gasket - and even gave them the FCO numbers - and still > didn't succeed. The FCO has been cancelled - "all drives have been retrofitted". Uh-huh. I can give you the Pertec number for the part, but you'll have to order $100 worth (and that's quite a few...) ================================================================================ Note 319.8 TA78 => TEU78??? 8 of 8 EISNER::COY "Dale E. Coy" 6 lines 10-MAR-1989 21:01 -< Situation OK, it's the thought that counts. >- -------------------------------------------------------------------------------- > I can give you the Pertec number for the part, but you'll have to order > $100 worth (and that's quite a few...) Thanks for the offer, but the drive was working fine without the FCO. It was just the principle of the thing. I was "entitled" to it. Sigh........  ================================================================================ Note 320.0 READ/WRITE LASER DISKS 21 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 22 lines 8-MAR-1989 23:45 -------------------------------------------------------------------------------- Can you find a flaw in this line of thinking........ I am about to propose to one of my customers they place a Trimarchi Laserase disk on their system instead of a TK70. The primary use of the drive is for backup. TK70 Laserase ------------- ----------------- List price w/ctl $6,800 $5,500 Capacity 300MB 600MB Media costs (list) $46 $175 Random access No Yes The write transfer rates are about the same. The read transfer rate of the Laserase is much higher. Comments? ================================================================================ Note 320.1 READ/WRITE LASER DISKS 1 of 21 EISNER::MAYHEW "Bill Mayhew" 1 line 8-MAR-1989 23:52 -< I've had same line of thinking >- -------------------------------------------------------------------------------- ================================================================================ Note 320.2 READ/WRITE LASER DISKS 2 of 21 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 31 lines 9-MAR-1989 02:19 -< R/W optical vs 8MM for BACkup >- -------------------------------------------------------------------------------- We have been doing the same sort of comparison, but to 8mm! not TKx0. The density ISN'T there, but one gets BOOTABILITY, and reliability. The drives we would be backing up are WAY too big, though (1.23 gig CDC Sabre), but the controller (Dilog DQ256) allows partitioning a disk into 1,2,3, or 4 equal size pieces. When partitioned, a whole 'disk' can be totally backed up to optical. Each night a different one, and those that change less frequently could be done less frequently. The Alphatronics we were looking at is almost double that price and these optical companies DON'T have a lot of product out there yet. Is paying more getting more, or is it simply the same Sony hardware priced by different companies? I know Trimarchi has good prices on RD54 look-alikes, but when I look at their removable winchester packaging at trade shows, it seems to remind me of prototypes, rather that finished production products. Not that that may necessarily indicate anything bad, but... Are there any standards, yet, that would allow interchangability? A NYC CDROM on PCs company has a R/W optical for about $5.5K, and one of the BIG CDROM mastering houses will take input in the form of these folks R/W disk, and make a RO master from it. If what is available from Trimarchi, or Alphatronics is standardised, maybe we could bypass mag tape in DECUS CDROM efforts if someone gets one of these soon enough. Other than wearing a FREE Trimarchi (& Clearpoint combined) tee shirt today I have NO connection to any of these companys. ================================================================================ Note 320.3 READ/WRITE LASER DISKS 3 of 21 EISNER::HARVEY "Jack Harvey" 13 lines 9-MAR-1989 04:14 -< Secondary Uses? >- -------------------------------------------------------------------------------- > The primary use of the drive is for backup. ^^^^^^^ A few years ago (before TK50's were available), I gave a customer demanding low price a system with a TK25 as the only removable medium. Within a year the TK25 was an orphan and the customer wanted an interchange capability. Trimarchi's Laserase may follow the same path. In my case, I warned the customer in advance that the TK25 would probably isolate him, but he insisted. So I got to add a 9 track drive to his system when the light dawned. ================================================================================ Note 320.4 READ/WRITE LASER DISKS 4 of 21 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 2 lines 9-MAR-1989 05:08 -< YUP - I LOVE CAPTIVE CUSTOMERS >- -------------------------------------------------------------------------------- > In my case, I warned the customer in advance that the > TK25 would probably isolate him, but he insisted. So ================================================================================ Note 320.5 READ/WRITE LASER DISKS 5 of 21 EISNER::KILGALLEN "Larry Kilgallen" 4 lines 9-MAR-1989 07:16 -< A discouraging word >- -------------------------------------------------------------------------------- Gee, Jeff, I get nervous when people start discussing the latest technology in the same topic with a discussion of their system backups. I presume you have figured out a way to get software onto the system. ================================================================================ Note 320.6 READ/WRITE LASER DISKS 6 of 21 EISNER::NAGY "Frank J. Nagy, VAX Guru" 16 lines 9-MAR-1989 08:52 -< 8mm tape is working as backup solution >- -------------------------------------------------------------------------------- Our examining of this issue has lead us to 8mm tape drives. The rough costs are about $2000 for controller (QBus->SCSI, handles disks and tapes) and about $3500 for the tape drive. Media is cheap, from $10 for Sony Video-8 to $30 for "certified" media. Performance is good; a VAXStation-II/GPX can backup about 120 MB of data (/NOBACKUP files excluded from count) in about 10 minutes from a CDC Wren-V disk to one of these. We have had/are having/working on reliability problems. Several drives have gone back to vendor and been replaced. We watching this as we add more systems with more 8mm drives. Oh BTW, our vendor for the controller and tape drive is Summus Computers. I have no relation to Summus other than as a customer. ================================================================================ Note 320.7 READ/WRITE LASER DISKS 7 of 21 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 12 lines 9-MAR-1989 15:09 -------------------------------------------------------------------------------- > Gee, Jeff, I get nervous when people start discussing the latest > technology in the same topic with a discussion of their system backups. Being one of those people who can't type the word BACKUP without also typing /VERIFY I am not too worried. Plus we do test our backup sets from time to time. > I presume you have figured out a way to get software onto the system. For a small fee we will transfer the distro tapes to disk for our customers. We do the same thing now for our TK25 and CDC 9448 sites. ================================================================================ Note 320.8 READ/WRITE LASER DISKS 8 of 21 EISNER::KILGALLEN "Larry Kilgallen" 4 lines 9-MAR-1989 17:06 -< One other consideration >- -------------------------------------------------------------------------------- Make *sure* you can count on being able to boot from the strange disk for all present and future versions of VMB. That way when the magnetic disk seeks the southern pole, you will be able to (have them) get stand-alone backup running. ================================================================================ Note 320.9 READ/WRITE LASER DISKS 9 of 21 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 8 lines 9-MAR-1989 17:27 -------------------------------------------------------------------------------- > Make *sure* you can count on being able to boot from the strange > disk for all present and future versions of VMB. That way when the > magnetic disk seeks the southern pole, you will be able to (have > them) get stand-alone backup running. Good point. In theory the drives look like a standard MSCP disk to the hardware and software but I will check it out. ================================================================================ Note 320.10 READ/WRITE LASER DISKS 10 of 21 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 19 lines 7-APR-1989 23:41 -< FIRST LOOK >- -------------------------------------------------------------------------------- Well I just got my LASERase disk drives from Trimarchi. I was all set to install them when I realized I didn't have any media. It has been so long since I have purchased a removable drive I forgot you have to order the packs also. Looks like FED-EX on Monday. Anyway having read through the manual they are "130mm Magneto-Optical disk drives model RO-5030E". They are built for OEM's since the manufacture's name appears nowhere. However the lawyers got them - the manual (which also avoids the use of names) has a disclaimer from RICOH corp. They are SCSI drives. The media format is WORM compatiable - so it looks like you could make packs that could be read by WORM drives. There is tons of ECC stuff in the disk format - a lot more than I have seen in magnetic drives. One surprise - they advertise these as 600MB drive *BUT* it is only 300MB per side. That means only 300MB on line. ================================================================================ Note 320.11 READ/WRITE LASER DISKS 11 of 21 EISNER::KENNEDY "Terry Kennedy" 16 lines 8-APR-1989 16:28 -< What am I missing? >- -------------------------------------------------------------------------------- > That means only 300MB on line. What is the benefit/justification of 300 Mb optical over 300 Mb of conventional disk? Reasons I could think of, but don't apply to this medium are: o More storage capacity o Smaller drive o Faster access time o Cheaper o Longevity of media in a storage environment o Not erasable for audit trails But none of these apply to this type of drive. What am I missing that makes it attractive? ================================================================================ Note 320.12 READ/WRITE LASER DISKS 12 of 21 EISNER::MAYHEW "Bill Mayhew" 6 lines 8-APR-1989 18:27 -< Removability >- -------------------------------------------------------------------------------- > What am I missing that makes it attractive? Removable media at a cost that's a lot lower than the other "removable" 5.25" disks (the ones that use the entire drive as the removable unit). Thus while 300Mb is the instantaneous limit, the effective storage capacity is a lot higher (if your application fits the model). ================================================================================ Note 320.13 READ/WRITE LASER DISKS 13 of 21 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 2 lines 8-APR-1989 19:12 -< REMOVABLE BACKUP SETS >- -------------------------------------------------------------------------------- > What is the benefit/justification of 300 Mb optical over 300 Mb of > conventional disk? ================================================================================ Note 320.14 READ/WRITE LASER DISKS 14 of 21 EISNER::KENNEDY "Terry Kennedy" 12 lines 9-APR-1989 01:58 -< More questions >- -------------------------------------------------------------------------------- > REMOVABLE BACKUP SETS Hmmm. If you're backing up as well as using them for on-line disks, then you need 2 of the beasties. Plus, the long-term retention integrity for these things is anybody's guess. Even the manufacturers don't know - if you ask the right people the right questions, they will admit that aren't sure, but when extrapolating from an ongoing one-year test, mumble mumble... If you're just looking for backup, there are conventional technologies which offer 300Mb per "thing", where "thing" is a single piece of whatever backup medium you are using. ================================================================================ Note 320.15 READ/WRITE LASER DISKS 15 of 21 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 26 lines 9-APR-1989 11:46 -< THE SPEED IS WHAT WE ARE LOOKING AT >- -------------------------------------------------------------------------------- > If you're just looking for backup, there are conventional technologies > which offer 300Mb per "thing", where "thing" is a single piece of whatever > backup medium you are using. We looked at three solutions... DRIVE LIST ------------------------------------------ ----------- TK70 $8,000 LASERase $5,500 Dual 150MB tape drives $4,000 All prices include controllers and are list. However the LASERase uses a SCSI controller that can also drive the WREN drives. Now Terry you are going to say you can get the Wangtec drives for less than $1150 a piece but to date the written quotes I have gotten all are between $1150 and $1250. If the speed of drives turns out to be the same as tape they will go back. If the drives turn out to be 2 to 3 times faster they are worth the added cost. ================================================================================ Note 320.16 READ/WRITE LASER DISKS 16 of 21 EISNER::MAYHEW "Bill Mayhew" 10 lines 9-APR-1989 17:26 -< Prices changed on TK70 very recently >- -------------------------------------------------------------------------------- > TK70... $8000 Note that the current price list (i.e. the Estore) lists TK70 drives at $4600 and controllers at $1000, bringing the total to just $6000. On the other hand, I still think the optical drive is at least a theoretically superior solution because of the improved random-accessness as compared to a TK... and it may be a better solution from the standpoint of the user interface (i.e. media insertion, etc.) as well. ================================================================================ Note 320.17 READ/WRITE LASER DISKS 17 of 21 EISNER::MAYHEW "Bill Mayhew" 1 line 9-APR-1989 17:27 -< Typo correction: TQK70s are $1400, not $1000 >- -------------------------------------------------------------------------------- ================================================================================ Note 320.18 READ/WRITE LASER DISKS 18 of 21 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 5 lines 9-APR-1989 18:47 -< WHO IN THEIR RIGHT MIND WOULD BUY A TK50 ANYMORE? >- -------------------------------------------------------------------------------- > Note that the current price list (i.e. the Estore) lists TK70 drives > at $4600 and controllers at $1400, bringing the total to just $6000. Interesting - there is now less than a $1700 dollar list price difference between the TK70 and the TK50. ================================================================================ Note 320.19 READ/WRITE LASER DISKS 19 of 21 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 7 lines 9-APR-1989 19:38 -< Ans: Software vendors. >- -------------------------------------------------------------------------------- > < Note 320.18 by EISNER::KILLEEN "Jeff Killeen DECUServe Chair" > > -< WHO IN THEIR RIGHT MIND WOULD BUY A TK50 ANYMORE? >- Answer: Anyone who sells software and who needs to ship it to sites that have TK50s. If they'd just given us a true "TK50 mode" on the TK70, this wouldn't be a concern. But noooooo.... ================================================================================ Note 320.20 READ/WRITE LASER DISKS 20 of 21 EISNER::HARVEY "Jack Harvey" 3 lines 9-APR-1989 21:27 -< Chance >- -------------------------------------------------------------------------------- When DEC introduced the TK50, they proposed it as a new industry standard. Ha! Only Big Blue can do that. For all others, it's a dice game. ================================================================================ Note 320.21 READ/WRITE LASER DISKS 21 of 21 EISNER::CONROY "Alan Conroy" 5 lines 22-APR-1989 03:56 -< You can convince most of the people most of... >- -------------------------------------------------------------------------------- > When DEC introduced the TK50, they proposed it as a new > industry standard. Ha! Only Big Blue can do that. For all > others, it's a dice game. And not even THEY can do it all the time.  ================================================================================ Note 321.0 Request Info on Market Share of Laser Printers 4 replies EISNER::OSBRINK 10 lines 9-MAR-1989 15:57 -------------------------------------------------------------------------------- Request Information on Market Share of VAX Laser Printers We are interested in purchasing a laser printer for graphics and text output in an engineering environment. Does anyone have information on the relative market share of laser printers used on VAX systems? Also any experiences in using Tektronix graphics, Postscript output, and output thru multiple VAXes would be helpful. We have SAS, GKS, Tektronix 4014 software and DECservers. ================================================================================ Note 321.1 Request Info on Market Share of Laser Printers 1 of 4 EISNER::PROVOST 2 lines 10-MAR-1989 13:34 -< TEKtronix hardcopy >- -------------------------------------------------------------------------------- We use LN03+ for TEKtronix output. It works well. We have 4 LN03+ printers on an ETHERNET. ================================================================================ Note 321.2 Request Info on Market Share of Laser Printers 2 of 4 EISNER::BURNET "John Burnet" 24 lines 11-MAR-1989 16:24 -< Take a look at Talaris >- -------------------------------------------------------------------------------- We too have a bunch of LN03+ printers for text and graphics output, but lately we have switched to the Talaris 1590 for higher-volume use. We have two 1590s so far. I can highly recommend this printer; see my remarks in note 309.12 in this conference. The 1590 is 99% compatible with the LN03-plus; as far as I know, the only area of incompatibility is that the 1590 doesn't yet handle soft (downloaded) LN03 fonts. A firmware upgrade that will introduce this capability is supposed to be in the works. If you will want to use fonts in ROM cartridges, you should be aware that the cartridges are (of course) different between the DEC printers and the Talaris printers. The 1590 stores all of its fonts in the compact "PK" format that was developed for use with TeX font files. Talaris will sell you a single font cartridge that contains *all* of the fonts that DEC sells in its eight or ten different LN03 cartridges. Postscript is available as an option for the 1590; as far as I know it's not (yet?) an Adobe Postscript, but the Compugraphic interpreter re-worked for use with the Talaris Printstation controller. As for market share, if you really want hard numbers, you might consider buying reports from market-research firms. The name Dataquest comes to mind; also the "Data Sources" books published by Ziff-Davis. Also, both Digital Review and Digital News publish comprehensive laser-printer "roundups" every year. ================================================================================ Note 321.3 Request Info on Market Share of Laser Printers 3 of 4 EISNER::ROECKEL 26 lines 11-APR-1989 08:24 -< Multiple VAXen a snap! >- -------------------------------------------------------------------------------- > Also any experiences in using Tektronix graphics, Postscript output, >and output thru multiple VAXes would be helpful. We have SAS, GKS, >Tektronix 4014 software and DECservers. We also have an LN03+ we use for Tektronix works. Seems to work fine. We have many HP LaserJet II's, but they are not DEC compatible, so lots of software DOES NOT work with them. As far as multiple VAXen accessing a single printer, this is fairly straight forward using a DECserver. Basically, you plug the printer into the DECserver, then define VMS Print Queues on ALL your VAXen that point to this DECserver. The DECserver will handle the contention problems (more that one VAX wants to access printer at same time) on its own. The biggest trick is making sure you have the latest version of LAT software and DECserver software to make it successful. I know there is a thread in here somewhere that defines all the details about configuring DECservers with printers. If you can't find it, and still need more detail, feel free to call me. -Bruce W. Roeckel Florida Power Corporation (813) 384-7851 ================================================================================ Note 321.4 Request Info on Market Share of Laser Printers 4 of 4 EISNER::TIHOR "Stephen Tihor;SITE SIG,VAX SIG" 3 lines 3-JUL-1989 16:33 -< Imagens in the TCP/IP world >- -------------------------------------------------------------------------------- At NYU we have had very good overal experience using IMAGEN (now partof QMS) networked printers from both VMS and J. Random Unix boxes.  ================================================================================ Note 322.0 VAXstation 2000 SCSI Details 1 reply EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 17 lines 10-MAR-1989 09:58 -------------------------------------------------------------------------------- Not sure where to post this so we'll try dropping it here. There is an interesting article in the March 1989 DEC Professional (for a change) entitled "SCSI Port on the VAXStation 2000". It is written by Victor R. Gold Jr. of Peritek Corp, Oakland California. Although it is very short on details it does explain, in very basic terms, how the NCR 5380 SCSI Controller chip works on this system. Two "keys" on this implementation are the 16 Kb Disk/Tape(SCSI) RAM buffer and the location (in memory) of the 5380's control registers (IO410$AB_SCTLS). I think that a couple of "our" hardware experts (Alan, Jamie, we are looking for volunteers) should kidnap this guy and "sweat" him for all the gory details which would then of course be publish by DECUS (or maybe we should just ask him nicely ?) ;-) ================================================================================ Note 322.1 VAXstation 2000 SCSI Details 1 of 1 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 11 lines 10-MAR-1989 13:29 -< You bring the hot lights, and I'll bring the cuffs! >- -------------------------------------------------------------------------------- > I think that a couple of "our" hardware experts (Alan, Jamie, we are > looking for volunteers) should kidnap this guy and "sweat" him for all > the gory details which would then of course be publish by DECUS (or > maybe we should just ask him nicely ?) ;-) I saw that article. I *hate* articles like that! Well, at least, I hate the "dangle the carrot and then snatch it away at the end" aspect. I even have a sample project in mind -- an Exabyte 8mm tape drive with a SCSI port. Hmmm.  ================================================================================ Note 323.0 Carrier Detect on DMF32 and ??? No replies EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 6 lines 10-MAR-1989 13:31 -------------------------------------------------------------------------------- It is well known that the "full modem control" ports on the DMF32 will not accept data until the Data Carrier Detect line is asserted. Are there any OTHER DEC muxes that behave this way???  ================================================================================ Note 324.0 RTS/CTS and DEC terminal muxes 1 reply EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 79 lines 10-MAR-1989 13:49 -------------------------------------------------------------------------------- I posted this on Pageswapper, meaning to cross-post here, but it took me a few days. Oh well. -------------------------------------------------------------------------------- In an SIR for Fall '88 (I believe), someone wanted the DEC terminal muxes and drivers to honor "DTR flow control", so that printers that dropped DTR when their buffer was full (or nearly so) could be used without resorting to xon/xoff or other "inband" mechanisms. Ensuing discussion on Pageswapper pointed out that DTR is an OUTPUT from a terminal mux as well as from a printer and therefore not a candidate for this purpose. BUT... experimentation has revealed that many of DEC's muxes and drivers DO pay attention to CTS, "Clear To Send". If CTS is deasserted, data will stop coming out of the port until whatever's attached to the port reasserts CTS. This fact is not mentioned anywhere in the DEC manuals that I'm aware of, but it has been verified on the DHV11, ports 0 and 1 of the DMF32, and even on the DB25 serial port on a MicroVAX 2000 (this despite the fact that the terminal driver book says that the uV2000 doesn't provide modem control! It works just fine with modems. Even under V5. So there.). It therefore seems possible to use such muxes to drive printers that use DTR flow control. Just use a null modem cable that connects the printer's DTR to the mux's CTS. I believe that the VAX port in question has to be set to /MODEM. If you have a different type of mux, you can try it easily. Hook a terminal to a port through a null modem cable and a breakout box (put the box between the mux and the null modem cable). Set the terminal port to /modem and start some lengthy output to the tube. Then use the breakout box to open CTS (pin 5) and see if the output doesn't stop. There will be a brief delay between opening the circuit and the cessation of output, due to the fifos (silos) in both the terminal and the mux and to software delays (the mux only seems to tell the VAX about the signal state change, it's up to the VAX to stop the output). This feature is handy not only for dealing with the aforementioned printers but also with "smart" modems, such as those which run various levels of MNP, not to mention the Telebit Trailblazer. You can tell the modem to use CTS flow control to throttle data from the VAX and all will be well, with a fully transparent data path. With the Trailblazer this means that you can lock the local DTE interface speed to 19200 for incoming calls, and turn off /autobaud on the line in question; this provides much more reliable connections and again gives you a transparent data path. (In particular, assuming that the far end is set up similarly, XOFFs sent from the VAX get all the way to the far end, just as is intended, so no means is needed whereby the VAX can tell the local modem to stop sending.) It would be most useful if people with other muxes could try this on their hardware and post the results here! (And what about terminal servers?) (US175084 "Harry Herman" replied:) -< Seems to work just fine for modems connected to DS-200s >- -------------------------------------------------------------------------------- I have our incoming/outgoing combination modems all hooked up to the DECserver 200s with 25-pin straight through cables and using RTS/CTS for flow control. It seems to work just fine. Our modems are US Robotics Courier HSTs that are Hayes compatiable and do MNP level 5 error correction and data compression. They are set up to do fixed 19,200 baud to the terminal server, and negotiate line speed to the remote modem. ================================================================================ Note 324.1 RTS/CTS and DEC terminal muxes 1 of 1 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 5 lines 10-MAR-1989 14:32 -< make sense >- -------------------------------------------------------------------------------- I knew (in theory at least) that CTS/RTS were used during half duplex communication for flow control. I a little suprised that it seems to work still in full duplex mode. We had better not tell DEC, they might just "fix" this !  ================================================================================ Note 325.0 Looking for 63xx new-owner advice No replies EISNER::NORTON "Bill Norton" 16 lines 10-MAR-1989 19:24 -------------------------------------------------------------------------------- What can anyone tell us about goodies in the 63xx console subsystem? After installing ours myself I'm curious about what else might be lurking in the RBD's (ROM-Based Diagnostics). The installation guide has you run one of these in order to test the load path TK70, but just says "...performs a read/write test". The command was Z/BI:6 E to connect to BI node 6 on XMI adapter E, then you get a RBD6> prompt, and enter D2/TR/T=6/C. Do all BI cards have RBD's? What do the parameters/switches mean? What other tests could/should be run? The manual recommends a console baud rate of 1200 or less. Has anyone run a higher speed without trouble?  ================================================================================ Note 326.0 RRD50 and the VS3100 No replies EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 5 lines 15-MAR-1989 11:37 -------------------------------------------------------------------------------- In some of the announcement stuff I have for the RRD40 CD drive, it says "Current owners of RRD50 can use the RRD40 controller". Does anyone know if this means I can plug an RRD50 drive into a VAXstation 3100?  ================================================================================ Note 327.0 MicroVAX II to MicroVAX 3600 Upgrade!! 36 replies EISNER::WOOD "Jim Wood (609)423-1040" 56 lines 15-MAR-1989 15:14 -------------------------------------------------------------------------------- Just in from my DECrep: MicroVAX II to MicroVAX 3600 upgrade ------------------------------------------------------------- System Upgrade Kit @ 28,770 consists of 1 KA650-AA CPU 1 MS650-BA 16MB memory 1 VMS 1-20 user license (QL-001AB-BJ) 1 DECnet end node license (QL-D04AB-AA) Workstation Upgrade Kit @ 10,745 consists of 1 KA650-BA CPU 1 MS650-AA 8MB memory 1 VMS 1-2 user license (QL-001AC-BB) 1 DECnet end node license (QL-D04AC-AA) (Both include doc/diag, manual, and upgrade kit) Boards not usable with the MVIII are DHQ11 all revisions shipped to date RQDX1 must be replaced RQDX2 must be replaced Min rev levels for other boards are M3104 DHV11-M B1 M8053 DMV11-M D M7164 SDI controller F3,F4 M7165 KDA50-QA B14 M7168 QDSS 4-plane mem A1 M7169 QDSS Base mod D2 M7206 TSV05 controller B1 M7504 DEQNA-M K4 M7516 DELQA-M C3 M7546 TK50 controller J2 M7552 KRQ50 B1 M7555 RQDX3-M C1 M7620-AA KA650-AA B1 M7620-BA KA650-BA B1 M7621-AA MS650-AA A1 M7740 KLESI-QA TU81E C4 M7957 DZV11-M C RA60-A SDI disk S3 RA81-A SDI disk D2 RA82-A SDI disk C4 RD53-A ST506 disk A1 RD54-A ST506 disk B1 TK50-A DECtape III B2 TS05-A tape drive D TU81E tape drive A1 -------------------------------------------------------- Customer installable. You return MVII CPU and DEC memory. Layered products need to be upgraded. Requires special quote. More details as I get 'em. ================================================================================ Note 327.1 MicroVAX II to MicroVAX 3600 Upgrade!! 1 of 36 EISNER::MERRIMAN "George Merriman - CCA/NY" 7 lines 15-MAR-1989 15:33 -< What happens to your old MVII licenses? >- -------------------------------------------------------------------------------- > System Upgrade Kit @ 28,770 consists of > 1 KA650-AA CPU > 1 MS650-BA 16MB memory > 1 VMS 1-20 user license (QL-001AB-BJ) > ^^^^^^^^^^^^^^^^^^^^^ > 1 DECnet end node license (QL-D04AB-AA) > ^^^^^^^^^^^^^^^^^^^^^^^ ================================================================================ Note 327.2 MicroVAX II to MicroVAX 3600 Upgrade!! 2 of 36 EISNER::MAYHEW "Bill Mayhew" 1 line 15-MAR-1989 16:54 -< Please provide part number for kit, when you have it >- -------------------------------------------------------------------------------- ================================================================================ Note 327.3 MicroVAX II to MicroVAX 3600 Upgrade!! 3 of 36 EISNER::WOOD "Jim Wood (609)423-1040" 8 lines 15-MAR-1989 22:21 -< VMS & DECnet upgrade included >- -------------------------------------------------------------------------------- VMS and DECnet licenses are "upgraded" as part of the entire package. All other licenses must be upgraded separately. (The QL part numbers for VMS and DECnet are apparently special upgrade-only part numbers) Other aspects of this program are still very fuzzy, but we're interested, and our DECrep is checking on a number of things for us. I'll post data as I receive it. ================================================================================ Note 327.4 MicroVAX II to MicroVAX 3600 Upgrade!! 4 of 36 EISNER::CROWELL "Shefth of the Fourth Order" 9 lines 16-MAR-1989 10:04 -< License Upgrade? >- -------------------------------------------------------------------------------- > VMS and DECnet licenses are "upgraded" as part of the entire package. > All other licenses must be upgraded separately. For my workstation this would be a down grad: From: VMS 16-user DECNET routing node TO: VMS 1-2 user DECNET end-node A bit costly. ================================================================================ Note 327.5 MicroVAX II to MicroVAX 3600 Upgrade!! 5 of 36 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 4 lines 16-MAR-1989 12:29 -< Why won't the DHQ11 work with MicroVAX 3600? >- -------------------------------------------------------------------------------- >> Boards not usable with the MVIII are >> DHQ11 all revisions shipped to date Does anyone know why there is a problem with the DHQ11? ================================================================================ Note 327.6 MicroVAX II to MicroVAX 3600 Upgrade!! 6 of 36 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 25 lines 16-MAR-1989 16:51 -< I don't know, but here are the possibilities. >- -------------------------------------------------------------------------------- >Does anyone know why there is a problem with the DHQ11? There are three issues which can cause Qbus devices to be unusable on the MV3xxx Qbus. If the driver for the device writes to Q-bus map registers (bypassing the system-supplied routines), the driver must be modified to not perform multiple successive writes thereto; these writes must be interleaved with either I/O space accesses or writes to physical memory. If the device can drop an interrupt request before the interrupt is serviced (passive release), there is a problem; this is not supported on the MicroVAX 3xxx. If the device can do this in response to action by the driver, the driver can be modified to avoid the problem. (Either the driver must only clear the interrupt enable when there are no additional interrupts pending, or the driver must do at least one device register read access after clearing the interrupt enable bit). If the device can drop the interrupt request all by itself, it cannot be used on a MV3xxx Qbus, and there is no software fix that i know of. Finally, DMA DATIO cycles are not supported on the MV 3xxx. If the device can perform such, the driver must be modified so as to preclude them when running on a MV 3xxx. ================================================================================ Note 327.7 MicroVAX II to MicroVAX 3600 Upgrade!! 7 of 36 EISNER::GARONE 21 lines 17-MAR-1989 16:24 -< MicroVAX II to 3xxx upgrade - follow up . . . >- -------------------------------------------------------------------------------- < Note 327.0 by EISNER::WOOD "Jim Wood (609)423-1040" > -< MicroVAX II to MicroVAX 3600 Upgrade!! >- Just in from my DECrep: MicroVAX II to MicroVAX 3600 upgrade ------------------------------------------------------------- System Upgrade Kit @ 28,770 consists of 1 KA650-AA CPU 1 MS650-BA 16MB memory 1 VMS 1-20 user license (QL-001AB-BJ) 1 DECnet end node license (QL-D04AB-AA) etc. etc. etc. >>> Looks like you got some valuable information. I'd like to follow up with your DECrep on this. Can you tell me who he is?? Tx, /Steve Garone ================================================================================ Note 327.8 MicroVAX II to MicroVAX 3600 Upgrade!! 8 of 36 EISNER::SHANNON "Terry C. Shannon, IDC" 6 lines 18-MAR-1989 11:04 -< The Lurid Details. >- -------------------------------------------------------------------------------- The official word on the MicroVAX II upgrade is that it is available on a case-by-case basis, hence the need for a special quote from your DECrep. DEC claims that the majority of its customers will find it more cost-effective to purchase a brand new MicroVAX 3300/3400. ================================================================================ Note 327.9 MicroVAX II to MicroVAX 3600 Upgrade!! 9 of 36 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 9 lines 18-MAR-1989 22:29 -< I sure hope I'm wrong about this... >- -------------------------------------------------------------------------------- || Looks like you got some valuable information. I'd like to follow || up with your DECrep on this. Can you tell me who he is?? >> The official word on the MicroVAX II upgrade is that it is available >> on a case-by-case basis, hence the need for a special quote from your >> DECrep. Oops! It looks like some DECrep is going to get a talking-to for talking too much. ================================================================================ Note 327.10 MicroVAX II to MicroVAX 3600 Upgrade!! 10 of 36 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 4 lines 19-MAR-1989 02:24 -< no NEW SYSTEM sale can be jeopardised >- -------------------------------------------------------------------------------- One customer of ours has said that his DEC rep had to internally produce a document SWEARING in his own blood that NO sale was being jeopardised by the upgrade, and that was in DEC's best interest to offer that particular upgrade... ================================================================================ Note 327.11 MicroVAX II to MicroVAX 3600 Upgrade!! 11 of 36 EISNER::WOOD "Jim Wood (609)423-1040" 3 lines 19-MAR-1989 20:27 -< Upgrade Details Secret!!?? >- -------------------------------------------------------------------------------- My DECrep said nothing about keeping details secret. I've sent him mail with questions that I have and some that were posted here. When (if?) he gets back to me, I'll post it. ================================================================================ Note 327.12 MicroVAX II to MicroVAX 3600 Upgrade!! 12 of 36 EISNER::SHANNON "Terry C. Shannon, IDC" 3 lines 20-MAR-1989 08:08 -< Rumour Has It. >- -------------------------------------------------------------------------------- My DECrep said nothing about keeping details secret. Nevertheless, I haven't caused such a stir at DEC since January 18, 1988! ================================================================================ Note 327.13 MicroVAX II to MicroVAX 3600 Upgrade!! 13 of 36 EISNER::WOOD "Jim Wood (609)423-1040" 10 lines 21-MAR-1989 09:44 -< Contact CSS Support for Details >- -------------------------------------------------------------------------------- Latest from my DECrep: He knows nothing (so to speak) about keeping the upgrade secret and he's not concerned that I've posted the info. He says that your DECrep needs to contact the local/district CSS support person for details on the upgrade. His CSS rep is away this week, so it'll be a week before he's able to answer my (and your) questions posed so far. ================================================================================ Note 327.14 MicroVAX II to MicroVAX 3600 Upgrade!! 14 of 36 EISNER::PRIGOT "Jonathan M. Prigot" 7 lines 21-MAR-1989 16:22 -< Redirection >- -------------------------------------------------------------------------------- Although this discussion is a hardware thread, it is probably more appropriate in the MICROVAX_SYSTEMS conference. Please resume it under Note 62.* there. Thanks, Jonathan M. Prigot Co-moderator MICROVAX_SYSTEMS ================================================================================ Note 327.15 MicroVAX II to MicroVAX 3600 Upgrade!! 15 of 36 EISNER::KOZAM 17 lines 21-MAR-1989 20:39 -< HARDWARE/MICROVAX/VMS & Moving Threads >- -------------------------------------------------------------------------------- > Although this discussion is a hardware thread, it is probably more > appropriate in the MICROVAX_SYSTEMS conference. Please resume it > under Note 62.* there. > > Jonathan M. Prigot > Co-moderator MICROVAX_SYSTEMS 1. There is great overlap between HARDWARE_HELP, MICROVAX_SYSTEMS, VMS, and DEC_SOFTWARE. With uVMS gone, and uVAX hardware now running neck-and-neck with 6XXX or 8XXX systems, it seems that MICROVAX_SYSTEMS no longer fills a distinct need. Notice how few notes are posted there. 2. I suspect that many people, like myself, read all of these news groups. By reposting articles, it just adds to the clutter. Unless a thread is CLEARLY in the wrong place, I vote that it's best not to move it, although a cross-reference to this thread might have been placed in other conferences. ================================================================================ Note 327.16 MicroVAX II to MicroVAX 3600 Upgrade!! 16 of 36 EISNER::HARVEY "Jack Harvey" 15 lines 22-MAR-1989 03:48 -< MicroVMS is Obsolete >- -------------------------------------------------------------------------------- > With uVMS gone, and uVAX hardware now running neck-and-neck > with 6XXX or 8XXX systems, it seems that MICROVAX_SYSTEMS no > longer fills a distinct need. > By reposting articles, it just adds to the clutter. I concur. I now find it difficult to differentiate between a MicroVAX and a VAX *except by hardware characteristics*, such as bus, packaging and hardware options. These are not topics called out in MICROVAX_SYSTEMS 1.0. To really reduce the clutter, I think it's time to close out the MICROVAX_SYSTEMS conference. However, until that's done, I suggest the moderator directions be followed. ================================================================================ Note 327.17 MicroVAX II to MicroVAX 3600 Upgrade!! 17 of 36 EISNER::PRIGOT "Jonathan M. Prigot" 5 lines 22-MAR-1989 11:53 -< Reasoning >- -------------------------------------------------------------------------------- The thought is that since this thread relates specificly to the MicroVAX it belongs in the specific (MICROVAX_SYSTEMS) rather than the general (HARDWARE_HELP). Discussions over whether the MicroVAX line is/should be distinct from the rest of the VAX world should be done over there as well. ================================================================================ Note 327.18 MicroVAX II to MicroVAX 3600 Upgrade!! 18 of 36 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 9 lines 23-MAR-1989 00:33 -< We're talking something that needs hand tools. >- -------------------------------------------------------------------------------- I concur with .15 and .16 To quote note 1.0: The purpose of this conference is to discuss hardware problems, glitches, ECOs, diagnostics, how-to questions, etc. ^^^^^^^^^^^^^^^^ If this thread doesn't answer "How do I upgrade a microVAX II to a microVAX II, then we're gonna have to look at all the DECserver questions, PDP 11 questions, etc. and redirect them. ================================================================================ Note 327.19 MicroVAX II to MicroVAX 3600 Upgrade!! 19 of 36 EISNER::ERSKINE "Chris Erskine" 3 lines 23-MAR-1989 07:25 -< Calling local moderators >- -------------------------------------------------------------------------------- Since the request for the move came from the moderator of uVAX what do the moderators of HARDWARE HELP feel. I think that they should really make the call since they police this conference. ================================================================================ Note 327.20 MicroVAX II to MicroVAX 3600 Upgrade!! 20 of 36 EISNER::BYRNE_C "Charlie Byrne" 8 lines 23-MAR-1989 11:09 -< This was not done in haste >- -------------------------------------------------------------------------------- I (coincidentally) had originally started the MVII-MVIII upgrade question in MICROVAX several months ago. When I saw the same topic starting up in HARDWARE it occured to me that something was wrong. The moderators of both conferences discussed it (one option was to do NOTHING). What you see is what we decided upon. -Charlie Byrne, CoMOderator, HARDWARE_HELP ================================================================================ Note 327.21 MicroVAX II to MicroVAX 3600 Upgrade!! 21 of 36 EISNER::COY "Dale E. Coy" 23 lines 24-MAR-1989 23:34 -< If you've gotta discuss it, don't do it >- -------------------------------------------------------------------------------- > The moderators of both conferences discussed it (one > option was to do NOTHING). What you see is what we decided > upon. The moderation philosophy will evolve as experience dictates. The result, in this case, was apparently to "kill" an active thread. There have been no replies to the TECHNICAL issue for 3 days, and it was red-hot before it was moved. For your consideration, I'd like to suggest that tampering with a thread (moving it, or even saying "back on track") be viewed as something to do when there is a CLEAR BENEFIT, and a PRESSING NEED to do it. IMHO, if several moderators have to DISCUSS whether to do something with a thread, the answer should (almost) ALWAYS be to do NOTHING. It's particularly drastic to move a thread that's well-established and had (as I recall) 14 notes. Please note that I'm not objecting to "strong moderation" - and you folks are doing a fine job. Just suggesting that examining results may influence future decisions. ================================================================================ Note 327.22 MicroVAX II to MicroVAX 3600 Upgrade!! 22 of 36 EISNER::HARVEY "Jack Harvey" 10 lines 25-MAR-1989 13:06 -< The Problem is MICROVAX_SYSTEMS >- -------------------------------------------------------------------------------- I quite agree, Dale, the move has had a stunning effect on the technical issue. It certainly has not resumed in the redirected location. We may have lost it. I disagree, however, that the moderators' judgment was wrong. I think the real problem is the existence of the MICROVAX_SYSTEMS conference. To address this, I have started topic 90 in NEW_CONFERENCE_IDEAS, wherein I propose closing it and give reasons and statistics supporting the action. ================================================================================ Note 327.23 MicroVAX II to MicroVAX 3600 Upgrade!! 23 of 36 EISNER::BYRNE_C "Charlie Byrne" 14 lines 26-MAR-1989 03:07 -< Lets move to NEW_CONF >- -------------------------------------------------------------------------------- > < Note 327.22 by EISNER::HARVEY "Jack Harvey" > > -< The Problem is MICROVAX_SYSTEMS >- > I think the real problem is the existence of the > MICROVAX_SYSTEMS conference. To address this, I have started > topic 90 in NEW_CONFERENCE_IDEAS, wherein I propose closing > it and give reasons and statistics supporting the action. I haven't even read your proposal yet, but judging from the controversial thread currently being discussed elsewhere (something like "What is a micro/mini/super-mini/mainframe") this should be interesting. I just wrote some suggestions, but I realized they don't belong here and will move them to NEW_CONF. In any event, I'm think the thread will become very active very soon. ================================================================================ Note 327.24 MicroVAX II to MicroVAX 3600 Upgrade!! 24 of 36 EISNER::WOOD "Jim Wood (609)423-1040" 21 lines 31-MAR-1989 21:28 -< Latest on MVII to MV3600 upgrade >- -------------------------------------------------------------------------------- The latest from my DECrep: As Terry Shannon wrote in a previous reply, the MVII to MV3600 upgrade is a sort-of secret. It is apparently meant for large customers and VARs would can order 100 or more kits at a time. My DECrep said someone in Maynard saw an article in ComputerWorld that told of the upgrade and mentioned that the news broke on DECUServe. My DECrep didn't get in hot water, but his CSS rep who copied the announcement and sent it down to us is, well, in potential trouble. The bottom line is that he's not sure that he can sell us an upgrade unless we buy 100 or so. Just before my DECrep came in today, a VAR from Edison NJ named Total Tec called me and told me of the two upgrade paths. He said that Total Tec had kits on order and will have them available Real Soon Now. If Total Tec is going to sell the upgrade kits, I imagine other large outfits like Hamilton Avnet and others will also have them for sale. ================================================================================ Note 327.25 MicroVAX II to MicroVAX 3600 Upgrade!! 25 of 36 EISNER::SHANNON "Terry C. Shannon, IDC" 4 lines 2-APR-1989 15:20 -< C'mon DEC, How's About A Wee Bit O' Glasnost? >- -------------------------------------------------------------------------------- I see no reason why *anyone* should get in trouble over the upgrade policy. I also suspect that the xx April announcement of more powerful MicroVAXes and the repricing/repositioning of existing MicroVAX systems may impact the upgrade program . . . ================================================================================ Note 327.26 MicroVAX II to MicroVAX 3600 Upgrade!! 26 of 36 EISNER::MCCORMICK_B "Bob McCormick" 3 lines 2-APR-1989 17:48 -< Has there been formal announcement yet? >- -------------------------------------------------------------------------------- A rumour I heard had the number peged at 25 systems versys 100 Has there actually been a formal release of this info, or are we all talking "futures" here ?! ================================================================================ Note 327.27 MicroVAX II to MicroVAX 3600 Upgrade!! 27 of 36 EISNER::WOOD "Jim Wood (609)423-1040" 6 lines 2-APR-1989 22:31 -< 100 or 25? Not sure >- -------------------------------------------------------------------------------- The 100 came from my DECrep. In his defense, it sounded more he was told a large number to impress on him that the upgrade program was for large customers and the like. I haven't seem anything printed on the upgrade program other than the info I posted in .0. ================================================================================ Note 327.28 MicroVAX II to MicroVAX 3600 Upgrade!! 28 of 36 EISNER::BONE "George Bone, MINSY" 8 lines 3-APR-1989 09:55 -< Well I asked MY DECrep, and MY DECrep said... >- -------------------------------------------------------------------------------- I asked our DECrep about the upgrade, and SHE said (after checking it out with her sources) that it was only for sites with 24 or more microVAXen. Personally, it sounds like bull ... er, uh... like a lot of hogwash to me, but if I continued, I'd have to move this over to SOAPBOX. Incidentally, I'm glad that this thread is continuing in HARDWARE, where it belongs rather than MICROVAX. ================================================================================ Note 327.29 MicroVAX II to MicroVAX 3600 Upgrade!! 29 of 36 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 5 lines 3-APR-1989 10:34 -< I heard ... >- -------------------------------------------------------------------------------- The rep I talked to said nothing about quantity but did say that the upgrade requests would be reviewed and approved "on a case by case basis". One way to achieve success is to say you going to switch brands (sort of like saying "I'm going to hold my breath untill I turn blue ..." and just a childish). ================================================================================ Note 327.30 MicroVAX II to MicroVAX 3600 Upgrade!! 30 of 36 EISNER::SHANNON "Terry C. Shannon, IDC" 4 lines 3-APR-1989 20:27 -< Speak Softly But Carry A Big Blue? >- -------------------------------------------------------------------------------- How to convince your DECrep to sell you an upgrade in 25 words or less: Just say "AS/400." ================================================================================ Note 327.31 MicroVAX II to MicroVAX 3600 Upgrade!! 31 of 36 EISNER::SMITHJR "Charles T. Smith, Jr." 5 lines 3-APR-1989 22:08 -< Then, there's the subtle approach >- -------------------------------------------------------------------------------- A card from your local IBM rep, placed correctly on or about ones desk (for example, sticking out shyly from under the phone) can also do wonders if you have the local DECrep over to your office. ================================================================================ Note 327.32 MicroVAX II to MicroVAX 3600 Upgrade!! 32 of 36 EISNER::SHANNON "Terry C. Shannon, IDC" 4 lines 3-APR-1989 23:02 -< Help Is On The Way >- -------------------------------------------------------------------------------- It stands to reason. Your DECrep has a pathological fear of the AS/400, AS/400 advertisements (and possibly even M*A*S*H reruns because they also feature Alan Alda!). Perhaps next week's MicroVAX announcement will help restore processor parity to the midrange and serenity to your DECrep! ================================================================================ Note 327.33 MicroVAX II to MicroVAX 3600 Upgrade!! 33 of 36 EISNER::MAYHEW "Bill Mayhew" 16 lines 3-APR-1989 23:47 -< DEC found the neighborhood... now where's the ballpark? >- -------------------------------------------------------------------------------- > Your DECrep has a pathological fear of the AS/400... As well he/she should... > Perhaps next week's MicroVAX announcement will help restore > processor parity to the midrange and serenity to your DECrep! Not unless DEC does something to straighten out its "support" mess, too! It's taken DEC nine months to wake up to the AS/400 threat... now how long do you think it will take them to realize where the threat really is?? Hint: It ain't in the hardware, and little of it is in the software... ================================================================================ Note 327.34 MicroVAX II to MicroVAX 3600 Upgrade!! 34 of 36 EISNER::MERRIMAN "George Merriman - CCA/NY" 4 lines 4-APR-1989 13:30 -< Remember, this is HARDWARE_HELP >- -------------------------------------------------------------------------------- This thread is clearly headed for SOAPBOX-land. Please take your flames and snide comments (however well founded) to that conference. George Merriman, HARDWARE_HELP co-moderator ================================================================================ Note 327.35 MicroVAX II to MicroVAX 3600 Upgrade!! 35 of 36 EISNER::MAYHEW "Bill Mayhew" 5 lines 4-APR-1989 15:39 -< Alleged part numbers, not verified >- -------------------------------------------------------------------------------- I have heard a rumor that the part numbers for the upgrades are 2T-KA650-UA and -UB (for the MV and VS versions, respectively). These show up on the Estore as CSS (presumably Computer Special Systems) products. There also exist -UC and -UD variants. ================================================================================ Note 327.36 MicroVAX II to MicroVAX 3600 Upgrade!! 36 of 36 EISNER::JELKE 9 lines 30-AUG-1989 10:41 -< Has anyone upgraded yet? >- -------------------------------------------------------------------------------- Has anyone actually performed the upgrade? Were there any problems? Is it really customer installable? We have 30 to 50 MicorVAXes at our site but I'm the only one here to get a quote for the upgrade. I've never found DEC people reluctant to talk about it or restrict sales to any quantity. Perhaps time has solved those non-functional problems.  ================================================================================ Note 328.0 Looking for LPS40 speed 5 replies EISNER::SCOPELLITI "Whatsa behind is uva no importa" 18 lines 15-MAR-1989 19:00 -------------------------------------------------------------------------------- We have a Printserver 40 that is not running as fast as we'd like. The job selects paper from both the main tray and one of the smaller trays. These selections are done pseudo-randomly... 5 sheets from main, then 2 from small tray, then 3 from main, etc. What we are seeing is the printserver, upon getting the change tray command, waits until the paper path is clear before selecting the other tray. Needless to say, this turns a 40 page/minute printer into about a 12 page/minute unit. I have a suspicion that we can speed things up by alternating between the two small trays, rather than using the main tray at all. Anybody have any experience? We're waiting for a second tray to be shipped, so we can't test our hypothesis just yet. (Application is written in Postscript). ================================================================================ Note 328.1 Looking for LPS40 speed 1 of 5 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 9 lines 16-MAR-1989 16:30 -------------------------------------------------------------------------------- You may call it "specsmenship" but the 40 page/min is for plain ASCII text sent through the ANSI translator to the LPS40 using any one tray ! I have timed my printer and find this rating to be "accurate" (a busy 11/780 can not feed it fast enough!) I have seen such "delays" when switching trays. The question is why are you pulling from different trays within one job ? ================================================================================ Note 328.2 Looking for LPS40 speed 2 of 5 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 36 lines 17-MAR-1989 00:35 -< More info.. lot's more info. >- -------------------------------------------------------------------------------- Well.. I guess a description is in order: The LPS 40 is printing monthly credit union statements. The "raw" data is converted into a postscript file, then sent to the LPS40. The raw data consists of each person's monthly statement consisting of the person's monthly activity (deposits/withdrawals) and a Visa statement, if the person has a Visa account. The Visa statements are in the small top tray, the regular statements in the big, lowest, bin. The credit union statement(s) and the Visa statement(s) are required to be printed on two different forms and the whole process requires everything to be in-line... i.e. each person's statements must be together. There is an envelope stuffer that keys off some marks printed by the LPS40. It thus knows when to fold/seal/get next envelope. As in most such cases, I can't change the process steps. Some history: When first called-in, the LPS40 was running at an agregate speed of about 7 pages / minute. Investigation showed that the postscript generator was generating redundant "SETPAPERTRAY" statements; i.e. even if 3 SETPAPERTRAY was in effect, there would be another at the start of the next statement. Removing the redundant paper tray selections resulted in a speed of about 12 pages/minute (got a smile out of the user!). (Removing all paper tray selections resulted in a speed of 44 pages/minute! Unfortunately, the output was useless since the Visa statements were not on the right forms. However, this did prove that we did not have a postscript processing bottleneck or etherneck bottleneck.) The user was shown an LPS40 running at high speed selecting different trays, so, I'm trying to locate the apparent bottleneck. Hence, I'm theorizing that switching between the big bin and either of the small trays requires the LPS40 to wait for the paper paths to be empty. Perhaps, there is less delay when switching between the two small trays??? ================================================================================ Note 328.3 Looking for LPS40 speed 3 of 5 EISNER::COY "Dale E. Coy" 6 lines 17-MAR-1989 00:45 -< Can we determine "why"? >- -------------------------------------------------------------------------------- This is a silly question, but what does DEC say? Not-quite-so-silly: have you tried asking the print-engine folks if this behavior is inherent in the engine (and MUST be that way), or if it's an "artifact" in the controller (hardware or software) that could be fixed? ================================================================================ Note 328.4 Looking for LPS40 speed 4 of 5 EISNER::MAYHEW "Bill Mayhew" 14 lines 17-MAR-1989 09:00 -< Can the form design be changed? >- -------------------------------------------------------------------------------- One of the things I have seen several major companies doing over the past couple of years is go to a *very* standard -- basically, blank with some colored backgrounds -- single form, on top of which all form-specific stuff is printed at the time the data is printed on the form. Clearest example I can think of is NETel bills, which used to have a distinct first page and multiple backup pages with call detail information on different paper. Now they're all on the same paper, and the laser printer structures the data differently and makes the pages *look* different (within the technological limits) to the recipient. Can something like this be done in your case? ================================================================================ Note 328.5 Looking for LPS40 speed 5 of 5 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 10 lines 17-MAR-1989 10:08 -< Nope.. can't touch it. >- -------------------------------------------------------------------------------- We're waiting for our DECrep 100 to locate a resource within DEC who knows the LPS40 in depth. The demo site where they saw the LPS40 selecting multiple trays at speed was a DEC office. My feeling is that we are lacking some bit of knowledge about the interaction of the LPS40 hardware and postscript. I can just hear it now... "Oh, yeah.. don't select the tray like that; use mumble instead because that avoids the mumble problem in the LPS40!" As for the forms, good idea, but we can't change them.. bummer...  ================================================================================ Note 329.0 DR780 help desired. 12 replies EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 17 lines 16-MAR-1989 12:24 -------------------------------------------------------------------------------- We have a potential application coming up for which we would need to transfer *really* large amounts of data *really* rapidly (at least from our normal viewpoint -- we're projecting the need to move 4GBytes in each task from an external source into disk). The VAX environment would be an 8650, with multiple HSC50's and several RA90's (or third-party equivalents). The DR780 looks interesting, with its 6.67MByte/sec transfer rate, but I would like a little more info if anyone know about it. It is not shown as a current product; is the DEC-supplied software still upgraded or otherwise suitable for current VMS releases? Is it smart enough to transfer data, for example, directly from an external device to an HSC connected to a CI on the same SBI as the DR780 without having to involve the SBIA and 8650 CPU? Assuming I can find one of these suckers on the used market, how reliable is it? ================================================================================ Note 329.1 DR780 help desired. 1 of 12 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 8 lines 16-MAR-1989 12:37 -< The DR780 cannot talk directly to CI >- -------------------------------------------------------------------------------- >> Is it smart enough to transfer data, for example, directly from an >> external device to an HSC connected to a CI on the same SBI as the >> DR780 without having to involve the SBIA and 8650 CPU? I have not used the DR780 myself, but am certain that it cannot do that. After all, the DR780 was designed looong before the CI, and the CI is not a trivial device to talk to. ================================================================================ Note 329.2 DR780 help desired. 2 of 12 EISNER::KENNEDY "Terry Kennedy" 7 lines 16-MAR-1989 15:28 -< Just an SBI version of DR11-W >- -------------------------------------------------------------------------------- >> Is it smart enough to transfer data, for example, directly from an >> external device to an HSC connected to a CI on the same SBI ... Think of it as a SBI version of a DR11-W and you'll get the idea. You can program it to snarf the data into main memory. What you do with it at that point is up to you. It should still be supported, but I don't have my VMS SYSGEN manual home - just look in the device tables... ================================================================================ Note 329.3 DR780 help desired. 3 of 12 EISNER::KILGALLEN "Larry Kilgallen" 1 line 16-MAR-1989 17:41 -< Have you considered a DR32? >- -------------------------------------------------------------------------------- ================================================================================ Note 329.4 DR780 help desired. 4 of 12 EISNER::BATES "Ken Bates" 18 lines 16-MAR-1989 20:14 -------------------------------------------------------------------------------- > We have a potential application coming up for which we would need > to transfer *really* large amounts of data *really* rapidly > (at least from our normal viewpoint -- we're projecting the need > to move 4GBytes in each task from an external source into disk). > The VAX environment would be an 8650, with multiple HSC50's and > several RA90's (or third-party equivalents). Don't forget that you'll have other limitations. The k.ci in the HSC will limit you to about 4 MB/sec, and the CI adaptor will reduce that down to about 1.5 MB/sec (maybe a bit more, I don't recall exactly). At 1.5 MB/sec over the CI, you won't be using the full capability of the RA90, since it spirals at 1.72 MB/sec. Third party vendors usually don't tell you what their spiral rates are, so I don't know how they would compare. If I recall, there's a session in Atlanta DECUS called "Striping" or some such thing which you may be interested in... - Ken ================================================================================ Note 329.5 DR780 help desired. 5 of 12 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 7 lines 17-MAR-1989 10:33 -< DR780 = DR32 for 78n/86nn? >- -------------------------------------------------------------------------------- >> -< Have you considered a DR32? >- The 1988 "Vax Systems Hardware Handbook -- UNIBUS Systems" book leaves me with the distinct impression that the DR780 is the DR32 packaged for 78x and 86xx systems (and the DR750 for 750's). I admit that the book is a little vague, but have I got it wrong again? ================================================================================ Note 329.6 DR780 help desired. 6 of 12 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 8 lines 17-MAR-1989 10:48 -< CI780 limit recognized. >- -------------------------------------------------------------------------------- >> ... and the CI adaptor will reduce that down to about >> 1.5 MB/sec (maybe a bit more, I don't recall exactly). That's why I was hoping that the DR780 would talk directly to a CI780; I had visions of putting a second SBIA in the 8650 and using it solely for DR780-to-CI780-to-HSC transfers. The DR780 might still be significantly faster than the old faithful DR11 in actual use, but it looks like it will cost a lot more $$. ================================================================================ Note 329.7 DR780 help desired. 7 of 12 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 14 lines 17-MAR-1989 11:29 -< DR780 to CI780 transfers unlikely. >- -------------------------------------------------------------------------------- As a rather straightforward DMA device, the DR780 (which is, indeed, an implementation of the DR32 architecture for the 780) can transfer data to or from successive locations in memory. If you have another adapter on the SBI that will respond like "memory", requiring little other programming, the DR780 might be able to transfer data directly to or from that device. The CI780 does not fit this description. (Historical note: I seem to recall that in the early days of the PDP11, one of the claims made for the Unibus was that peripherals would be able to talk directly to each other without requiring the data to go through intermediate buffers in memory. While this capability certainly exists on the Unibus for certain combinations of controllers, no operating system ever written for the PDP11 EVER took advantage of this, to my knowledge.) ================================================================================ Note 329.8 DR780 help desired. 8 of 12 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 15 lines 17-MAR-1989 12:37 -< The capability exists; has anyone used it? >- -------------------------------------------------------------------------------- >> (Historical note: I seem to recall that in the early days of the PDP11, >> one of the claims made for the Unibus was that peripherals would be able >> to talk directly to each other without requiring the data to go through >> intermediate buffers in memory. While this capability certainly exists >> on the Unibus for certain combinations of controllers... This is why some controllers have a bit that says "Don't increment the DMA memory address". This allows a DMA controller to shove its data to a single location (in this case, a data register in another controller). I have thought of applications where this might be useful, but have never seen it done in practice. Has anyone ever done it? (If you reply to this query, perhaps it should be moved to a separate topic) ================================================================================ Note 329.9 DR780 help desired. 9 of 12 EISNER::KENNEDY "Terry Kennedy" 9 lines 18-MAR-1989 15:43 -< It is done, may even help here >- -------------------------------------------------------------------------------- >> (Historical note: I seem to recall that in the early days of the PDP11, >> one of the claims made for the Unibus was that peripherals would be ... Well, various Unix implementations use one of the KMC processor modules to give DZ11's DMA capability. Kind of clever, really... In better relation to this topic, you might be able to get similar 'assist' functionality with a DR11-W and a KMC, but you'll have to write some really weird software to do it... ================================================================================ Note 329.10 DR780 help desired. 10 of 12 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 30 lines 19-MAR-1989 02:47 -< RSX has KMCs doing it too >- -------------------------------------------------------------------------------- > Well, various Unix implementations use one of the KMC processor modules >to give DZ11's DMA capability. Kind of clever, really... Terry has a RSTS shop, so I will stick up for RSX. RSX has it several times over! (or is it just M+, anyway:) These first 2 are layered products you had to pay more for, and were (I think) the first projects of Dave Mitton (of DECNET-DOS fame) when he joined DEC. These first 4 examples ONLY use the -A KMC processor. COMMIOP-DZ KMC nurse-maiding several DZs COMMIOP-LP KMC nurse-maiding several LPs Under DECNET: (the microcode is included in your Decnet kit) KDZ KMC nurse-maiding several DZs KDP KMC nurse-maiding several DUPs Those were over the bus connections. Just to finish KMC's usage, the following is not over the bus, but is a cable over the top connection, and is NOT the -A 1k processor but on the -B 4k processor. The KMS-1P is actually the 4k KMC-B processor and the line card half robbed from a 2 card DMR! The microcode to run this is INCLUDED in your 11M+ PSI kit. It is MUCH less expensive to BUY a used DMR and KMC-B and throw away the unneeded DMR's CPU than to buy a used KMS-1P. Some dealers know the trick, but some don't... ================================================================================ Note 329.11 DR780 help desired. 11 of 12 EISNER::BORCHARDT "RODNEY P. BORCHARDT" 5 lines 7-JUL-1989 13:36 -< DR780 >- -------------------------------------------------------------------------------- < Note 329.0 by EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" > -< DR780 help desired. >- I far as I know the DR780 is no longer avaiable. The micro code has not changed in 7 years, as far as I know. We have one on a 780 running v5.1-1 of vms. ================================================================================ Note 329.12 DR780 help desired. 12 of 12 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 4 lines 18-JUL-1989 16:30 -< Thanks for the info. >- -------------------------------------------------------------------------------- Thank you for the info. The application in which this would have been used, if we could get one (the used market response was "DR-What??") ended up on a Gould/IGD image processor, so we don't need the I/O speed we expected.  ================================================================================ Note 330.0 Power fail on RA disks? 12 replies EISNER::SCOPELLITI "Whatsa behind is uva no importa" 14 lines 18-MAR-1989 00:20 -------------------------------------------------------------------------------- We've got the following configuration: VAX 8530, KDB50s, RA81s, RA82s If we take a power hit (i.e. lights out for a while), when power comes back, the RA disks do not spin up, but rather have to be manually restarted. Our field service man, who is quite good, claims this is the way life is. Has all the sequencing cables plugged in and everything. My comment is "Well, then what good is power fail recovery?" Am I missing some real obvious point? Or am I losing what's left of my mind? ================================================================================ Note 330.1 Power fail on RA disks? 1 of 12 EISNER::COY "Dale E. Coy" 4 lines 18-MAR-1989 01:02 -< Seems like a tall tale >- -------------------------------------------------------------------------------- RAs on an HSC normally/generally/usually/almost always spin back up correctly, in sequence, on a power fail reboot. Can't comment on the KDB, but "it oughta". ================================================================================ Note 330.2 Power fail on RA disks? 2 of 12 EISNER::RICE "Gary Rice" 5 lines 18-MAR-1989 09:06 -< Tall Tales >- -------------------------------------------------------------------------------- We have several "unscheduled" blackouts in the machine room every year. The only drives that routinely fail to power back up are the SI drives. Our Ra60s, 81s and 82s do it just fine. Gary ================================================================================ Note 330.3 Power fail on RA disks? 3 of 12 EISNER::KENNEDY "Terry Kennedy" 5 lines 18-MAR-1989 15:47 -< Works here >- -------------------------------------------------------------------------------- We have 4 RA81's on a UDA50 on a 780, and 1 RA81 on a UDA50 on a PDP-11, all of which spin up just fine. As I recall, it is a driver-level issue. Are you running VMS? What version? Also, the first drive in the chain needs to have a power sequence terminator installed (but if it was missing, your drives should *never* start...) ================================================================================ Note 330.4 Power fail on RA disks? 4 of 12 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 9 lines 18-MAR-1989 19:38 -< Are we talking about autoreboot or pfail recovery? >- -------------------------------------------------------------------------------- Hmm. I think the difference here is "power fail recovery" vs. "automatic reboot after powerfail". When our RA-series drives (two on a KDB50, one on a KDA) are powered on with the RUN button pushed in, they just sit there. When VMS startup issues a MOUNT command, they spin up. This works fine during a VMS boot (whether after powerfail or not), but I can see it failing on power fail recovery without special support in the driver (since no mount command is issued in this case, things are expected to just keep right on from where they were). ================================================================================ Note 330.5 Power fail on RA disks? 5 of 12 EISNER::KENNEDY "Terry Kennedy" 9 lines 18-MAR-1989 20:49 -< On powerfail restart they wait for first access >- -------------------------------------------------------------------------------- > ...but > I can see it failing on power fail recovery without special support in > the driver (since no mount command is issued in this case, things are > expected to just keep right on from where they were). You are correct, they will not *automatically* spin back up right away. However, the first time someone accesses the disk in question, you get a OPCOM message saying 'volume offline, mount verification in progress' and they spin right back up... ================================================================================ Note 330.6 Power fail on RA disks? 6 of 12 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 22 lines 19-MAR-1989 12:03 -< That's the way it works! >- -------------------------------------------------------------------------------- re: .0 What your Field Service person told is correct: that's the way they work. If the RUN/STOP button is depressed and the drive is powered up, it will just "sit there". When the drive is first accessed during a boot sequence, the host will issue an ONLINE command to the drive, causing it to spin up. When the drive spins up and loads heads (and runs any self tests, etc), the host may then access the drive. If you have a power fail, upon power restoration, the drive will not spin up. The host, upon completing its power fail restart will attempt to restart any I/O pending at the power fail time. The driver will detect that the disk drives are offline, and report this via the mount verification means. The mount verification will then handle bringing the drive ONLINE. This only happens when an I/O is requested for that particular drive. The drive will not be spun up until it is accessed. Stu ================================================================================ Note 330.7 Power fail on RA disks? 7 of 12 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 5 lines 20-MAR-1989 00:06 -< Thanks again.. mind back in working order! >- -------------------------------------------------------------------------------- Thanks for the thoughts... Guess what I've seen makes sense when you think about it! Now, to test it on my system. Let's see.. where _IS_ the building's main power switch located?.... ================================================================================ Note 330.8 Power fail on RA disks? 8 of 12 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 8 lines 18-APR-1989 23:51 -< The continuing quest for power-fail/reboot... >- -------------------------------------------------------------------------------- Well.. at a non-disclosure meeting today at one of DEC's ACTs, I overheard someone asking one of the DEC presenters the exact same question! "My 8530 does _NOT_ spin up the system disk on power fail." So.. I'm not losing my mind! The DEC person was one of the high-end storage folks from Colorado Springs. He was most interested in the problem and promised us an answer. I'll post results as they happen. ================================================================================ Note 330.9 Power fail on RA disks? 9 of 12 EISNER::MERRIMAN "George Merriman - CCA/NY" 2 lines 19-APR-1989 04:12 -< What's ACT? >- -------------------------------------------------------------------------------- Well.. at a non-disclosure meeting today at one of DEC's ACTs, I ^^^ ================================================================================ Note 330.10 Power fail on RA disks? 10 of 12 EISNER::WICHMAN "Jon Wichman" 4 lines 19-APR-1989 09:34 -< ACT = Demonstration Center? >- -------------------------------------------------------------------------------- -< What's ACT? >- It's what they call their demo centers. I forget what the "ACT" actually stands for at the moment... ================================================================================ Note 330.11 Power fail on RA disks? 11 of 12 EISNER::CHADWICK "Keith Chadwick, Fermilab" 2 lines 19-APR-1989 10:36 -< ACT :== Application Centre for Technology >- -------------------------------------------------------------------------------- > I forget what the "ACT" actually stands for at the moment... ================================================================================ Note 330.12 Power fail on RA disks? 12 of 12 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 1 line 19-APR-1989 10:37 -< Application Center for Technology >- -------------------------------------------------------------------------------- >> -< What's ACT? >-  ================================================================================ Note 331.0 Hygrothermograph charts needed 2 replies EISNER::HAHN "DECUServe Vice Chair" 22 lines 20-MAR-1989 12:46 -------------------------------------------------------------------------------- I have a humidity/temperature recorder which has been in our computer room for a while. The plate in the back of the instrument has the following: Weather Measure Corporation PO Box 41257 Sacremento, CA 95841 916/481-7565 Model H-302 I have tried to reach them at the above and through information and have had no luck. What I need is a source of charts to put on the drum of this recorder. The identifiers on the chart are: HYGROTHERMOGRAPH C 302-W-HF C 313-W-HF The size is 12 5/8 by 7 inches Any help will be appreciated. ================================================================================ Note 331.1 Hygrothermograph charts needed 1 of 2 EISNER::MERRIMAN "George Merriman - CCA/NY" 10 lines 20-MAR-1989 20:10 -------------------------------------------------------------------------------- Try: Nashua Graphic Products Graphic Controls Div. (or something like that) 189 Van Rensselear Street Box 1272 Buffalo, NY 14240 716-847-7597 800-828-1535 ================================================================================ Note 331.2 Hygrothermograph charts needed 2 of 2 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 11 lines 22-MAR-1989 00:26 -< Another source for Hygrothermograph charts >- -------------------------------------------------------------------------------- >> What I need is a source of charts to put on the drum of this recorder. I buy mine (for a different model instrument, however) from: Belfort Instruments 727 S. Wolfe St. Baltimore, MD 21231 (301) 342-2626 They have both the charts and the ink (in various colors).  ================================================================================ Note 332.0 Info on serial to parallel converters for LP25 8 replies EISNER::STAMERJOHN "RW Stamerjohn" 2 lines 21-MAR-1989 13:22 -------------------------------------------------------------------------------- I need to hook an LP25 up to a serial port. Does anyone have info on RS-232 to parallel converters which will do the trick. ================================================================================ Note 332.1 Info on serial to parallel converters for LP25 1 of 8 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 9 lines 21-MAR-1989 14:08 -< Black Box converters work fine for us. >- -------------------------------------------------------------------------------- >> I need to hook an LP25 up to a serial port. Does anyone have info >> on RS-232 to parallel converters which will do the trick. We have used three different serial/parallel converters from Black Box for this purpose, for Centronics and Dataproducts printer interfaces, to serial ports on DEC and Honeywell boxes. They have all three worked correctly out of the box, run 24 hours/7 days without any difficulty (one of them has been in use for six years). ================================================================================ Note 332.2 Info on serial to parallel converters for LP25 2 of 8 EISNER::KOZAM 7 lines 21-MAR-1989 20:47 -< Be careful about flow control >- -------------------------------------------------------------------------------- The cheapest serial->parallel boxes don't support XON/XOFF flow control (they do it by asserting/dropping one of the RS-232 modem control lines - either CTS or DTR - I don't recall which one right now). Since not many DEC print drivers support this (although SOME do), you'll probably need to get a more expensive one that can translate the appropriate parallel signals into XON/XOFF. ================================================================================ Note 332.3 Info on serial to parallel converters for LP25 3 of 8 EISNER::SMITHJR "Charles T. Smith, Jr." 2 lines 21-MAR-1989 22:05 -< We like black box, too... >- -------------------------------------------------------------------------------- We've been using a black box on an LP26 for 4 years. Unpacked it, plugged it in, and have never worried about since. ================================================================================ Note 332.4 Info on serial to parallel converters for LP25 4 of 8 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 18 lines 22-MAR-1989 09:49 -< And yet another ... >- -------------------------------------------------------------------------------- Black Box makes several different serial to parallel converters. The one we have used is called a Mini Print Spooler and has up to 1 Mbyte of RAM buffer. These work very well and properly handle XON/XOFF flow control. The input is a female DB25 and the output is a male Centronics. 64K buffer TD-PIC90B $ 299 256K buffer TD-PIE90B $ 525 1M buffer TD-PIF90B $1350 Black Box is also now selling a new parallel extender, that I think is much more than a parallel/RS-232/parallel converter. They claim it can go up to 1000 ft and use 4 pair of phone wires. Call Black Box at 412-746-5565. Just another satisfied customer. (and if you don't like their stuff you can send it back within 30 days for a refund !) ================================================================================ Note 332.5 Info on serial to parallel converters for LP25 5 of 8 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 7 lines 22-MAR-1989 13:01 -< Mediator >- -------------------------------------------------------------------------------- We've used a product called Mediator from Integrated Applications, Inc, 8600 Harvard Ave., Cleveland, OH 44105, (216)341-6700. It was very inexpensive, could do EIA flow to XON/XOFF flow conversion, can handle 110 to 38.4K bps, and has worked without a single glitch for several years. I have no affiliation with the above named company. ================================================================================ Note 332.6 Info on serial to parallel converters for LP25 6 of 8 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 15 lines 23-MAR-1989 00:20 -< Dataproduct's conversion kits. >- -------------------------------------------------------------------------------- We went a different route. We converted the printers internally by buying the conversion kit. We contacted Dataproducts (They're the guys who actually build most of DEC's big printers), gave them the model of the printer as well as the actual board numbers of the current interface and the CPU board in the printer. They gave us the proper conversion kit numbers to order, then we called Printer Warehouse and bought the kits at about 1/2 Dataproduct's list price. All is working well. I can get phone numbers for both if anyone wants them.. I'm home right now. Standard disclaimer: I have no affiliation with either Dataproducts or Printer Warehouse. ================================================================================ Note 332.7 Info on serial to parallel converters for LP25 7 of 8 EISNER::KENNEDY "Terry Kennedy" 23 lines 23-MAR-1989 01:26 -< Comments on changing the interface; other technical stuff >- -------------------------------------------------------------------------------- > We went a different route. We converted the printers internally > by buying the conversion kit. Handy if you can get the kits for less than the cost of the external buffer. Otherwise, the DP internal 'buffer' on the serial board is something like 8K, and you will *definitely* weird out DEC Field Service by changing the inter- face. Next, some rather weird technical tidbits (if you don't understand, don't worry)... 1) Watch out for the tiny serial-parallel buffer boxes on the market which *say* they do DataProducts. We bought the long-line printer extender from Inmac (runs parallel-parallel over 4-wire phone line up to 7,000 feet). It claimed to do DataProducts, however the printer was being driven directly from the PIP bits of an 8048 microprocessor. We had to clip out all the terminators in the LP25 connector PCB to get it to work (standard DP/LP25 termination is 220/330, which is a *bit* stiff). 2) Provided you have the drive current, all you need to convert Centronics to DataProducts is a 7406 to invert the strobe - the rest is passive. I once built such a converter into the DB-50 hood, powered off the DP +5v test line on the DB-50. ================================================================================ Note 332.8 Info on serial to parallel converters for LP25 8 of 8 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 6 lines 24-MAR-1989 23:36 -< Get field service involved. >- -------------------------------------------------------------------------------- << and you will *definitely* weird out DEC Field Service by changing the << interface. We actually enlisted the aid of our field service person! Now there's no doubt in his mind as to what's on the printer.  ================================================================================ Note 333.0 Comparing VAX Architecture Implementations: 8x00, 6x00, 12 replies EISNER::DEJORDY "Richard DeJordy" 29 lines 21-MAR-1989 16:06 -------------------------------------------------------------------------------- Well, As many other probably have recently, my bosses (and their bosses have asked me to justify why a 6 MIPS 8700 is more valueable than two 3 MIPS 3100s. Now they don't care about the Internals of VMS, but they would be willing to accept things like the bandwith of the bus that connects the memory to the cpu is higher and faster.... They want objective, quantitative measures only - qulatitative justifications need not apply. So, does anyone have a good source for a way to explain to non-engineering type the concepts behind the architecture associated with a Mini versus a workstation? I'd like a book that would help me explain concepts that are quite technical to upper management - or maybe that I could give them as required reading. (If I personally was more familiar with the hardware architecture, it would be much easier.) Any advice - or suggestions - are MOST welcome. Thanks Rich DeJordy American Mathematical Society RAD@MATH.AMS.COM DEJORDY on DECUServe and DCS. ================================================================================ Note 333.1 Comparing VAX Architecture Implementations: 8x00, 6x00, and 1 of 12 EISNER::SMITHJR "Charles T. Smith, Jr." 18 lines 21-MAR-1989 22:08 -< One way... >- -------------------------------------------------------------------------------- > As many other probably have recently, my bosses (and their bosses have >asked me to justify why a 6 MIPS 8700 is more valueable than two 3 >MIPS 3100s. Now they don't care about the Internals of VMS, but they >would be willing to accept things like the bandwith of the bus that >connects the memory to the cpu is higher and faster.... They want >objective, quantitative measures only - qulatitative justifications >need not apply. I personally would be hard pressed to justify an 87 or 8800 machine these days with the 6360 here and the 64xx series coming. Of course Digital is apparently offering a real fire sale on these processsors so maybe you could hack a good deal... One thing you might work from is the number of users each will support, in terms of disk space, connected terminals, support devices like printers, tapes, etc. ================================================================================ Note 333.2 Comparing VAX Architecture Implementations: 8x00, 6x00, and 2 of 12 EISNER::KENNEDY "Terry Kennedy" 8 lines 21-MAR-1989 22:57 -< Depends on your type of computation >- -------------------------------------------------------------------------------- > As many other probably have recently, my bosses (and their bosses have >asked me to justify why a 6 MIPS 8700 is more valueable than two 3 >MIPS 3100s. If you occasionally need to do *one* thing fast, 6 mips is twice as fast as 2*3 mips. Unless you want to spend a lot of time looking for parallelism in your programs, a uniprocessor will run them faster than N CPU's, as long as the uniprocessor is faster than the fastest *single* other CPU. ================================================================================ Note 333.3 Comparing VAX Architecture Implementations: 8x00, 6x00, and 3 of 12 EISNER::NAGY "Frank J. Nagy, VAX Guru" 25 lines 22-MAR-1989 09:07 -< I/O bandwidth, # of peripherals, licensing costs >- -------------------------------------------------------------------------------- Aside from the single stream throughput of an 8700 vs a 63xx, the major differences between the big machines (63xx, 88xx) and the 3100 are: (1) I/O bandwidth. The big systems support higher speed (transfer rate) peripherals such as 9-track and 3480-style tape drives. (2) Number of peripherals. The big systems can usually support greater numbers of disks/tapes, etc. (especially via the CI with HSCs). (3) Expensive peripherals. Ues, one can probably buy a 9-track 6250 bpi tape drive connected by SCSI to a 3100, but the cost of the tape drive will likely exceed the total cost of the 3100 AND its disks. The costs of a high-speed 6250 bpi tape (or even the $$$$ 3480 tape) is a fraction of your big machine cost. This strange situation is often easier for the bean counters to understand. On the downside: (4) The same software will cost much, much more for your big VAX than for two 3100s. ================================================================================ Note 333.4 Comparing VAX Architecture Implementations: 8x00, 6x00, and 4 of 12 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 18 lines 22-MAR-1989 10:16 -< A couple of more thoughts >- -------------------------------------------------------------------------------- As mentioned before, stay away from the the whole 8000 series. They are dead. The only possible reason to buy one is if you need to bring the maximum number of MIPS to bear on one problem. And then you should wait for the Aquarius/Ardius machine. The 6300 (soon to be 6400) series is much more bang/$. Frank gave a very good run down on the major difference but here are a couple of more. Memory size/bandwidth - the 3100's are very limited in this area Future expansion - no matter what your boss says your system is going to grow sooner or later. Paying extra $$$ now may save lots later. Number of simultaneous users - VAXstation can only have 2 ! ================================================================================ Note 333.5 Comparing VAX Architecture Implementations: 8x00, 6x00, and 5 of 12 EISNER::BYRNE_C "Charlie Byrne" 1 line 22-MAR-1989 15:04 -< I'll bet DEC will be glad to help you >- -------------------------------------------------------------------------------- ================================================================================ Note 333.6 Comparing VAX Architecture Implementations: 8x00, 6x00, and 6 of 12 EISNER::SMITHJR "Charles T. Smith, Jr." 10 lines 22-MAR-1989 20:25 -< 88xx fire sale in progress >- -------------------------------------------------------------------------------- > As mentioned before, stay away from the the whole 8000 series. > They are dead. Not quite dead... our decrep's are offering some real fire sale deals on the 8800 series right now.... "And not only do you get this 8820 processor, but these disks, and this tape, and a bamboo steamer, and a ginsu knife, how much would you pay now?" ================================================================================ Note 333.7 Comparing VAX Architecture Implementations: 8x00, 6x00, and 7 of 12 EISNER::ERSKINE "Chris Erskine" 11 lines 23-MAR-1989 07:32 -< OS overhead adds up >- -------------------------------------------------------------------------------- One thing which has not been brought up is VMS overhead. With 2 systems, you are running 2 copies of VMS with all of its overhead. With 1 CPU, you only have 1 copy so the over head is less. Since you are running move users on the single machine, the overhead will be greater than 1/2 of the 2 copies but still not as much as the 2. The other areas that you did not bring up is would the 3100's be clustered. If they are and the 8xxx is not, then you have just added more overhead due to lock manager and communications. This whole thread can be continued for your environment. ================================================================================ Note 333.8 Comparing VAX Architecture Implementations: 8x00, 6x00, and 8 of 12 EISNER::HARVEY "Jack Harvey" 8 lines 24-MAR-1989 05:04 -< Hopeless Quest >- -------------------------------------------------------------------------------- > So, does anyone have a good source for a way to explain to >non-engineering type the concepts behind the architecture associated >with a Mini versus a workstation? I'd like a book that would help me >explain concepts that are quite technical to upper management - ... I sympathize, but I think this is really quite hopeless. Try lying. They're used to that and can believe it. It's what marketing people do all the time. :-) ================================================================================ Note 333.9 Comparing VAX Architecture Implementations: 8x00, 6x00, and 9 of 12 EISNER::HARVEY "Jack Harvey" 10 lines 24-MAR-1989 05:08 -< Reliability/Availability Issue >- -------------------------------------------------------------------------------- One should consider the initial computer center to which the single or pair of machines are to be added. If it's important to always have a VAX of some sort available, perhaps to run some super critical time dependent jobs, then the two smaller machines may be strongly preferred for this reason alone. However, if a number of VAXen already exist, such that availability is no longer an issue in the particular shop, this argument fades. ================================================================================ Note 333.10 Comparing VAX Architecture Implementations: 8x00, 6x00, a 10 of 12 EISNER::HARVEY "Jack Harvey" 25 lines 24-MAR-1989 05:09 -< Grosh's Law Broken >- -------------------------------------------------------------------------------- Back in the sixties, when the IBM 360 appeared, Herb Grosh took a look at the performance/price ratios for each machine of the series. He noticed that the larger the machine, the better the performance/price ratio. Moreover, he was able to show performance increased at some exponent greater than one of the price. Somewhat tongue in cheek, he published this and a great furor broke out in letters to the editor of JACM. This continued for years, with scholarly papers examining what came to be called "Grosh's Law" and attempting to evaluate the exponent. Grosh's Law, simply stated, was that the larger the machine, the more bang for the buck. (There was a corollary to the Law: if you couldn't keep the big machine busy, the smaller machine was the better buy. This lead, in part, to the minicomputer, et al.) Grosh himself finally pointed out that the pricing of computers was a marketing decision and not tightly coupled to their manufacturing cost. And DEC marketing has established pretty well that if you don't like the current prices of their product line, just wait a month or two. I suspect they use an AI program, which deep inside has a random number generator. ================================================================================ Note 333.11 Comparing VAX Architecture Implementations: 8x00, 6x00, a 11 of 12 EISNER::COY "Dale E. Coy" 8 lines 24-MAR-1989 22:12 -< Would a simple analogy help? >- -------------------------------------------------------------------------------- It occurs to me that a simple, non-technical analogy might be helpful to your bosses: If 4 people need to go somewhere, is it better to get one 4-passenger car or two 2-passenger cars? That's not a simple question, either - but you can draw lots of parallels with the 8700 vx 2x3100 ... ================================================================================ Note 333.12 Comparing VAX Architecture Implementations: 8x00, 6x00, a 12 of 12 EISNER::AGRAWAL 12 lines 3-APR-1989 12:48 -< answer is match with your needs >- -------------------------------------------------------------------------------- I think the real answer is along the lines of what Frank Nagy had to say. You have to determine your needs and match them with the configurations solutions. If you need a system for a few (2-3) very power hungary users, VAXstations may do the job. If you need the system for lots of users who do lots of different things, you would need the minis. It is dictated by the number of users (licensing issues), memory on the system and number of ports (which probably will not be an issue if you use LATs.)  ================================================================================ Note 334.0 Microcom / Ven-Tel / (Hayes) dial-out problems 11 replies EISNER::LEDERMAN "Bart Z. Lederman" 32 lines 22-MAR-1989 10:57 -------------------------------------------------------------------------------- There are some threads here and in VMS about hooking up various modems for dial-in and dial-out under VMS V5.0 (or other vrersions) dealing with the periodic droppinhg of DTR and modem hang-up. I have gone through all of them and have not been able to solve this problem. I've just moved to a new site which has two dial-in / dial-out modems on a Micro-VAX II. One is a Microcom AX/2400, the other is a Ven-Tel 1200-31. I'm told that both of these used to allow dial-out, and think I even used it myself a few weeks ago (when visiting). But this week they don't work. When you attempt to dial out, the vax periodically drops DTR or something and disconnects. VAXnet says it's getting invalid responses from the modems. You can see it start to go out to the modem, but before the canll can be completed the VAX disconnects. This is probably related to the modem control signal problem (not all of the signals getting through) so I've replaced the cables between the modems and the VAX (total length about 6 feet) with no change. I've tried the suggested Hayes modem settings for the Microcom, and it doesn't work (i.e., setting C1). If I set CTS and DTR to be always on, then I can get the VAX to connect to the modem and not disconnect: however, when the modem dials out and gets a response from the remote modem, you can hear the remote modem respond yet the local modem says there's no carrier. Can anyone using either of these two modems tell me how they have the DIP switches and other options set? This problem is really beginning to bug me: I've never had this much trouble with any other modem set on any other computer I've used in the last 12 years (including other VAXes). ================================================================================ Note 334.1 Microcom / Ven-Tel / (Hayes) dial-out problems 1 of 11 EISNER::PRIGOT "Jonathan M. Prigot" 10 lines 22-MAR-1989 11:49 -< Microcom switches >- -------------------------------------------------------------------------------- Speaking for the Microcom, we have the following switches set: F 1 2 3 4 5 6 7 8 9 10 R 1 2 3 4 5 6 7 8 U U U D U U D U U U U U U D U D U D In addition we have SX1 set to pass XON/XOFF characters through to the VAX. I believe a similar discussion exists on this conference. I would suggest that you do a title SEARCH. ================================================================================ Note 334.2 Microcom / Ven-Tel / (Hayes) dial-out problems 2 of 11 EISNER::MAYHEW "Bill Mayhew" 17 lines 22-MAR-1989 13:10 -< Maybe... >- -------------------------------------------------------------------------------- Bart, What version of VMS are you running? And, can you identify anything that was done between "last week" (when the modems worked) and "today" to the system environment? I have seen the 30-sec DTR drop phenomenon, but I did something and it went away. Foggy memory says it was a SYSGEN parameter but I don't remember which bit... the CSC pointed me in the right direction. (This was back in 4.5 days...) Re: hearing the other modem respond, but your modem says no carrier: Knowing that you are a CompuServe subscriber, I want to put up a flag about this problem. I have had the same problem with some Boston CompuServe modems and my own (you're now in Boston, right?), and am convinced it is one or more of their modems. Be sure you try some other service before accepting this problem as your own. ================================================================================ Note 334.3 Microcom / Ven-Tel / (Hayes) dial-out problems 3 of 11 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 4 lines 22-MAR-1989 13:11 -< It differs depending on what you want >- -------------------------------------------------------------------------------- We use settings UUUDUUUDUU for the front and UUUUUDDD for the rear. In particular, with rear switches one through four up (note .1 shows switch four down), the AX won't forget any register settings or stored phone numbers if one pushes the rear RESET switch. ================================================================================ Note 334.4 Microcom / Ven-Tel / (Hayes) dial-out problems 4 of 11 EISNER::BEYER "Dennis Beyer - Tektronix" 6 lines 22-MAR-1989 19:35 -< 30-sec DTR drop is controlled by TTY_DIALTYPE >- -------------------------------------------------------------------------------- The 30-sec DTR drop phenomenon is controlled by the SYSGEN parameter TTY_DIALTYPE. The default for this parameter is 0 which will cause the 30-sec DTR drop. Set TTY_DIALTYPE to 4 to disable this "feature". The VMS I/O User's Reference Manual: Part I (VMS V5) has a discussion of this parameter on page 8-12 and 8-13. ================================================================================ Note 334.5 Microcom / Ven-Tel / (Hayes) dial-out problems 5 of 11 EISNER::LEDERMAN "Bart Z. Lederman" 18 lines 23-MAR-1989 08:42 -< Making some progress. >- -------------------------------------------------------------------------------- I haven't tried the modem settings given here, but I will. I have gotten the Microcom to dial out after doing a reset to factory settings, but only using VAXnet: SET HOST/DTE and KERMIT give the "connect but no response" situation. However: this modem was set (I -think-) to do speed conversion, and the possibility just occurred to me this morning that it's setting the host line to be the speed it answered at rather than the 9600 Baud the connection originates at (is this how Microcom works?) which VAXnet can handle but the others can't. And, I see this situation going to several different services. Dial-in isn't working at the moment. The Ven-Tel will work on dial-in, and will dial-out only using VAXnet and MODEM_CHECK :== OFF. This is good enough for what people will be using it for here. TTY_DIALTYPE = 0, will have to try to dig up the manuals (there're around here somewhere...). Thanks for the suggestions. ================================================================================ Note 334.6 Microcom / Ven-Tel / (Hayes) dial-out problems 6 of 11 EISNER::CHADWICK "Keith Chadwick, Fermilab" 18 lines 23-MAR-1989 10:29 -< Another modem setting to try >- -------------------------------------------------------------------------------- > I have gotten the Microcom to dial out after doing a reset to factory > settings, but only using VAXnet: SET HOST/DTE and KERMIT give the > "connect but no response" situation. I have a Microcom QX/12K (MNP class 7) modem connected to the CSA0: port on the back of "my" VAXstation, and use SET HOST/DTE to dial out. I was observing the same problem that you were (Connect 2400 reported, but I was unable to log into either DECUServe or the Electronic Store), until I configuted the modem to use "normal mode". The command for doing this is: AT\N0 (AT mode) SMNOR (SX mode) After making this change, I have been able to repeatedly dial out to both DECUServe and the Electronic Store without any problems. -Keith. ================================================================================ Note 334.7 Microcom / Ven-Tel / (Hayes) dial-out problems 7 of 11 EISNER::LEDERMAN "Bart Z. Lederman" 32 lines 30-MAR-1989 12:56 -< TTY_DIALTYPE and switches didn't help. >- -------------------------------------------------------------------------------- I thought I was making progress, but things are really still the same. VMS: it was at VMS 5.0-1, I upgraded to 5.0-2 and saw no change. I upgraded to VMS 5.1, and set TTY_DIALTYPE = 4: this has NOT helped. I still get the 30 second cycling of DTR, even though both modems are set to have modem control signals on all the time. VAXNET won't dial out on the Microcom modem at all. It tries to do a modem check even though MODEM_CHECK is set to "N". And the 30 second DTR cycle knocks down everything anyway. VAXNET will sometimes dial out on the VenTel modem, if it can get a connection fast enough to avoid the 30 second DTR killer. The Microcom switch settings only helped a little as they were set for SX mode and the modem here is set for AT mode. I've tried turning off command echo and status echo: this seems to make VAXNET give up on the Microcom modem altogether. We don't have a full VMS doc set here, so I can't get to the I/O Users' manual. But TTY_DIALTYPE = 4 does not appear to make any difference whatsoever in the way the system is talking to the modems. (I can get to the Microcom modem if I run from a privileged account, set the line to be /NOMODEM/NODIAL/NOHANG, and then Kermit or SET HOST/DTE to the modem. I have to be very careful to set the characteristics back when I'm done, otherwise the modem won't work for dial-in, or will work and not disconnect on logoff. This solution is one which I cannot give to non-privileged users of the system.) ================================================================================ Note 334.8 Microcom / Ven-Tel / (Hayes) dial-out problems 8 of 11 EISNER::CHADWICK "Keith Chadwick, Fermilab" 29 lines 31-MAR-1989 11:08 -< More modem settings to experiment with >- -------------------------------------------------------------------------------- The Microcom QX/12K (MNP class 7) modem which I previously had connected to the CSA0: port on the back of "my" VAXstation, has been loaned to another department in order that they may evaluate it 8-(. In the meantime, I have gone back to a Microcom AX/2400C. After connecting it in place of the QX/12K, I was also unable to success- fully dial into DECUServe. I made the following changes to the default power-up configuration: Front panel switches: U D U D U U U U U U Rear panel switches: U U U U U U U U Disabled the BPS rate adjust feature: AT\J0 (AT mode) BAOFF (SX mode) Selected 'normal' operating mode: AT\N0 (AT mode) SMNOR (SX mode) Stored the current configuration: AT&W (AT mode) With the above configuration, I am able to dial out to DECUServe without any problems (well, almost 8-). I'll capture the full configuration of the modem via AT\S and post it when I next log into DECUServe. -Keith. ================================================================================ Note 334.9 Microcom / Ven-Tel / (Hayes) dial-out problems 9 of 11 EISNER::CHADWICK "Keith Chadwick, Fermilab" 61 lines 31-MAR-1989 11:20 -< Microcom AX/2400c configuration >- -------------------------------------------------------------------------------- > I'll capture the full configuration of the modem via AT\S and > post it when I next log into DECUServe. As promised here it is: at\s IDLE 000:41:31 LAST DIAL T9,1-508-485-2574,,, ID: MODEM BPS 2400 AT MODEM FLOW OFF AT\G0 MODEM MODE NOR AT\N0 AUTO ANS. OFF ATS0=0 SERIAL BPS 9600 AT BPS ADJUST OFF AT\J0 SERIAL FLOW OFF AT\Q0 PASS XON/XOFF OFF AT\X0 - STRIKE ANY KEY TO CONTINUE - PARITY 8N AT BREAK 5 AT\K5 EXIT CHAR 043 ATS2=43 CMD ECHO ON ATE1 RESULTS ON ATQ0 RESULT TYPE MNPL ATV1\V1 DATA ECHO OFF AT\E0 INACT TIMER 00 AT\T0 AUTO RETRAIN ON AT%E1 COMPRESSION ON AT%C1 MAX BLK SIZE 256 AT\A3 AUTO BUFF 0 AT\C0 - STRIKE ANY KEY TO CONTINUE - AUTO CHAR 000 AT%A0 MNP BLOCK OFF AT\L0 IP PROTO OFF AT\I0 EMULATING HP OFF AT\H0 PAUSE TIME 002 ATS8=2 DTR 0 AT&D0 CARR DET 0 AT&C0 DSR 0 AT\D0 RING IND 1 AT\R1 SPKR CTRL 1 ATM1 LEASE LINE OFF AT&L0 ASYNC/SYNC 0 AT&M0 - STRIKE ANY KEY TO CONTINUE - CTS/RTS ON AT&R0 DISC DELAY 000 AT%D0 REM CHAR 042 AT*S42 REM ENABLE OFF AT*E0 REM SEC OFF AT*R0 RDLB ENABLE ON AT&T4 DIAL MODE 4 ATX4 PULSE DIAL US AT&P0 GUARD TONE 0 AT&G0 BELL ON ATB1 OK ================================================================================ Note 334.10 Microcom / Ven-Tel / (Hayes) dial-out problems 10 of 11 EISNER::LEDERMAN "Bart Z. Lederman" 34 lines 4-APR-1989 11:01 -< Working! >- -------------------------------------------------------------------------------- I have them working dial-in/dial-out. The problem turned out to be modem control signals (no surprise) which I found when I finally got a break-out box. The important settings were: ATS0=2 (to allow dial-in as well as dial-out) AT\J0 (so the VAX is always 2400 baud regardless of line speed) AT\Q1 (flow control on the serial line, may not be vital) AT&D0 (ignore DTR, so the modem will respond when you send commands to it even if the VAX thinks the modem is off-line) AT&C1 (Carrier detect follows the remote system. Otherwise, the VAX thinks the modem is always connected to something even if it isn't.) AT\D2 (DSR and RTS follow CD: this one is important. If it's set to AT\D1 the modem will raise one of the signals when you give it a dial command, and in 30 seconds the VAX will hang you up unless the other signals have come up. With AT\D2, none of the modem control signals come up until CD (carrier) come up). The last three did the trick. Now I can dial out (manually or with VAXNET) and dial in. The modems always run at 2400 baud between the modem and the VAX and auto-speed to the remote system. Switches: 1 1 2 3 4 5 6 7 8 9 0 D D D U U U U D U U front U D D U U U U D rear It's only been up a few hours, but I'm fairly certain this has done it. Other settings are factory default. ================================================================================ Note 334.11 Microcom / Ven-Tel / (Hayes) dial-out problems 11 of 11 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 2 lines 4-APR-1989 12:18 -< Microcom will forget >- -------------------------------------------------------------------------------- Of course with rear switches two and three down, if you ever reset your modem, you'll have to reconfigure it all over again.  ================================================================================ Note 335.0 Offline Storage Alternatives / Backup 22 replies EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 14 lines 23-MAR-1989 12:50 -------------------------------------------------------------------------------- Our cluster has a disk farm consisting of 44 RA81s and RA82s. We are going to be getting some RA90s in the near future. We already need most of our second-shift time to do backups. With the new disks, we are looking at 2 1/2 more man-days of operator time to do backups, plus getting another roomful of magtapes and a dedicated TA78. What can people tell me about other options for high-speed, high- density storage media (8mm tape drives, optical, etc.)? We have two different needs, for 'rotating' erasable storage and infinite-retention archiving. We are also cramped for space, budget, and operator time. Any thoughts, guidelines, war-stories, comparisions are welcome. ================================================================================ Note 335.1 Offline Storage Alternatives / Backup 1 of 22 EISNER::KILGALLEN "Larry Kilgallen" 5 lines 23-MAR-1989 14:10 -< High-end solutions for High-end sites >- -------------------------------------------------------------------------------- With all those disk drives, you ought to be able to buy a TA90 cartridge tape drive (pair) from DEC (hardware built by IBM). The stacker option takes away a lot of the time which is "wasted" waiting for the operator to put up another tape (even if you are devoting 2 drives to each backup job, this can be a problem). ================================================================================ Note 335.2 Offline Storage Alternatives / Backup 2 of 22 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 1 line 23-MAR-1989 15:33 -< 8mm = high density >- -------------------------------------------------------------------------------- 8mm <> high speed !!! ================================================================================ Note 335.3 Offline Storage Alternatives / Backup 3 of 22 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 25 lines 23-MAR-1989 19:06 -< backing a BIG disk farm >- -------------------------------------------------------------------------------- > With all those disk drives, you ought to be able to buy a TA90 > cartridge tape drive (pair) from DEC (hardware built by IBM). The slave drives you can add to a DEC master pair are IDENTICAL to IBM's. The DEC guy who told our LUG that admitted it as he sees it as a plus when selling into a company decomissioning IBM CPUs. It might sway someone into at least buying A master drive from him. He said you can roll it up and plug it in. He did mumble something about impacting his sales projections... The COLOR will be different. I havn't looked to see what secondary market prices are on the IBM slaves, but that might help your decision. The NEW card you MUST use in your HSCx for a TA90 is like the old DISK or TAPE one but has complete proms that make it a DISK AND TAPE one (less spares). Ask ALL you friends to 'HONESTLY' tell you how many times their 8mm drives have been back to the manufacturer in the first year, and also ask how many hours they log each day, and what ELSE they do for backup they CARE about. Depending on your own requirments, you might look at rewritable optical drives in a juke box config. Probably BEST as DISKS, not as tape emulators (as DEC does), but that is just my bias. ================================================================================ Note 335.4 Offline Storage Alternatives / Backup 4 of 22 EISNER::RICE "Gary Rice" 13 lines 24-MAR-1989 10:16 -------------------------------------------------------------------------------- > Depending on your own requirments, you might look at rewritable > optical drives in a juke box config. Probably BEST as DISKS, not > as tape emulators (as DEC does), but that is just my bias. We have a similar backup "problem". The optical approach is looking very attractive, but we feel that the marketplace/vendor offerings have not stablized enough yet to jump on the bandwagon. However, I expect that by September we will have selected a vendor and ordered equipment. The likely configuration will be a 50 (or more) platter jukebox that supports 5 1/4" platters. Gary ================================================================================ Note 335.5 Offline Storage Alternatives / Backup 5 of 22 EISNER::CHADWICK "Keith Chadwick, Fermilab" 46 lines 24-MAR-1989 13:40 -< Some suggestions >- -------------------------------------------------------------------------------- Hi John. The cluster that I help manage currently has 48 RA82's, 12 RA90's, and 12 TA79's. Our backup strategy is to perform a full backup once per week (which occupies all of Saturday and part of Sunday), and twice daily incrementals for all of the disks that we back up. Fortunately for us, 20 of the RA82's and 4 of the RA90's are only used as an intermediate data spooling area and thus do not require backup. I would make the following recommendations: Make sure that you are optimizing your existing backups. If your are performing image backups, I would suggest /BUFF=5 /IMAGE/BLOCK=32000, the question of /[NO]CRC is a personal choice: I use CRC, but I know of several people who's opinion I respect (but not with *MY* backups... 8-) that do not. Avoid backing up bound volume sets (if you can). We have observed that the time to backup an "n" disk bound volume set requires more than the time to back up "n" individual disks. Even using the /VOLUME=n qualifier does not help much. I would make an even stronger statement: Don't use bound volume sets unless you have absolutely no alternative. Wait for the long promised (at least at DECUS symposia) faster backup (perhaps in VMS V5.2 ?). The nice thing about the Exabyte 8mm tape drives is that you can have your operator just mount one tape and go away for at least a couple of hours. We are using Exabytes for data logging, and also for non-attended backup on several NI clusters on site. One drawback is that Exabyte drives which attach to the HSC are considerably more expensive than those that attach to either unibus or qbus. You could consider an alternate backup strategy: On Monday do full backups of disks 1-8, and incrementals of all the other disks, on Tuesday to full backups of disks 9-16, and incrementals of all the other disks. Thus over the course of a week, all disks would have a full backup, and daily incrementals. If you have volume shadowing and some spare disks, you could consider a backup strategy which would form shadow sets of two members, wait for the catchup to finish, dismount the second member and perform a physical backup of the second disk. This will be *MUCH* faster than an image backup (but be sure to issue roller skates to the operators first... 8-). -Keith. ================================================================================ Note 335.6 Offline Storage Alternatives / Backup 6 of 22 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 15 lines 24-MAR-1989 23:55 -< The search for adequate backup continues. >- -------------------------------------------------------------------------------- Funny how things happen.. I just dialed-in with the intent of entering a topic on this exact issue! It seems that the capacity of hard disks is growing much faster than the capability of any backup medium. Everytime DEC has shown us a new, bigger disk, I've asked "Great, now how do I back it up in a reasonable time?" This is typically answered with lot's of shuffling of feet and mumbles. Optical disks look like the eventual good choice, but I haven't seen the speed or cost effectiveness yet. As I recall, RV20 platters are $400 (MLP) each. Looked at Andromeda (?) erasable optical disks at DEXPO, but those folks didn't have any encouraging speed figures. The 8mm tapes look good, but again, haven't seen the speed. ================================================================================ Note 335.7 Offline Storage Alternatives / Backup 7 of 22 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 79 lines 25-MAR-1989 06:09 -< some random thoughts. >- -------------------------------------------------------------------------------- > The 8mm tapes look good, but again, haven't seen the speed. I have just one data point to contribute here. Our uV3600 drives an 8mm unit (through a Dilog controller) at an average of a bit more than 200 Kb/sec. This is using BACKUP/BLOCK=65534/CRC/GROUP=5, on a reasonably-fragmented set of disks (no defragmenters are used and the disks haven't been compressed, ever; now that we have the 8mm drive we'll do some full restores). This is calculated by doing a show dev/full to find out how many blocks are in use, subtracting the sizes of the files that are tagged /nobackup, and dividing by the total time to do the backup for the volume, then multiplying the result by two as the BACKUP command includes /verify; the 200 Kbyte/sec rate is the average of the two passes. We are quite satisfied with this; it seems possible to fill a cartridge in around three hours. (I elieve that the max burst rate through the beasts is only around 240 kbyte/sec, so there is little cause to complain.) To be sure, 200 Kbytes/sec would be unacceptable for bulk backups if writing to, say, 6250 bpi tapes. (Indeed, 200 Kbytes/sec is about as fast as this machine seems to be able to BACKUP to 6250 bpi tape, also.) The operator would have to hang around to do a reel change every 150-170 Mbytes or so. But with 2.2 GB on a cart, it's a matter of making sure there's a cart in the drive and going home. The next morning we remove the cart and put it on the shelf. But tens of RA81s/82s are in a whole different league. As someone else has mentioned, a "rolling" strategy might make sense: Each night, do a full backup on some subset of the disk farm, and do incrementals on the rest. By the way, if you expect to be doing lots of recoveries, a nice trick for making recovery from incrementals more tolerable is NOT to write them with /RECORD. Then the incrementals are cumulative; you need only the latest one, plus the most recent image. Of course, the incrementals get longer and longer each night, but you need only keep the most recent and the one before that around... Oh, and if you like having double protection, do incrementals on the set that you're going to do the image on, just before doing the image. Then if the most recent image backup goes bad you can go back to previous image + last incremental. You get the benefit of having a "spare" image backup without having to do the whole thing twice. One of the neatest things about the 8mm package is the form factor: Three 700+ Mbyte 5-1/4 drives, plus the tape unit to back them all up, can fit in one rackmount unit, half the height of an RA81/RA82/etc. Total cost (excluding controllers), about $15K. Is DEC really unaware of why the Qbus refuses to die??? In view of the problems reported by others (none seen here so far) with 8mm units, we take a little precaution. We recycle the cartridges on a two-week basis (except for weeklies which are kept longer). We write a tiny saveset containing just a test file as the first AND last things on the cart; and when the backup procedure mounts the cart, if it recognizes the label as an old backup cart it does a backup/compare on these two savesets. No problems after a month or so of doing this every weeknight. Admittedly this is early in the game. I'll post an update after six months or so. Nor have we had troubles using 8mm carts bought at Tower Video. But then, we are buying top-quality Sony carts. To judge by the mfg stamps on these they are coming from the same LOTS (not just the same factory) as the carts that Exabyte wants four times as much for. The mind boggles at the number of 1600 bpi reels of tape (about 50) that can be put on one of these things, for about $8 in media and three hours of computer time! Digital data drives based on the DAT standards (4 mm wide tape) are rumoured to be just around the corner. Capacity will be smaller (1 Gbyte per cart) but the carts are smaller, file-skip times may be better, and throughput and data reliability may be better. An interesting experiment: Wanna find out just how fast you can do a backup, assuming the fastest possible output device? Try doing a backup to nla0:/save , see how long it takes, and calculate the data rates. (Don't try doing /verify, though! :-) ================================================================================ Note 335.8 Offline Storage Alternatives / Backup 8 of 22 EISNER::KILGALLEN "Larry Kilgallen" 26 lines 25-MAR-1989 10:04 -< You knew that, didn't you? >- -------------------------------------------------------------------------------- > has mentioned, a "rolling" strategy might make sense: Each night, do a > full backup on some subset of the disk farm, and do incrementals on the > rest. > making recovery from incrementals more tolerable is NOT to write them with > /RECORD. Then the incrementals are cumulative; you need only the latest > one, plus the most recent image. Of course, the incrementals get longer > protection, do incrementals on the set that you're going to do the image > on, just before doing the image. Then if the most recent image backup > goes bad you can go back to previous image + last incremental. You get These are all good reminders from Jamie, but I thought they were *basic* techniques that DECUS should be teaching in Introduction to Backup SITE SIG sessions at Symposia. I realize that not everybody goes to Symposia, DECUServe is a different medium, etc., but if there is anyone who *does* go to Symposia and has *not* heard of these techniques, then we (DECUS) are not doing our job in structuring sessions to properly spread knowledge between production sites. I am shocked to visit sites which try to do all their full backups on the same night, but as I have indicated before, sites most in need of outside consultants are those *without* active DECUS members (I am not sure which way the causal relationship runs). ================================================================================ Note 335.9 Offline Storage Alternatives / Backup 9 of 22 EISNER::KILGALLEN "Larry Kilgallen" 2 lines 25-MAR-1989 10:05 -< RV20 is not the answer >- -------------------------------------------------------------------------------- Somebody mentioned something about speed. RV20's specialty is reliability (and non-repudiation). ================================================================================ Note 335.10 Offline Storage Alternatives / Backup 10 of 22 EISNER::NAGY "Frank J. Nagy, VAX Guru" 26 lines 25-MAR-1989 11:02 -< Large disk farm - buy several 8mm drives >- -------------------------------------------------------------------------------- >>But tens of RA81s/82s are in a whole different league. As someone else >>has mentioned, a "rolling" strategy might make sense: Each night, do a >>full backup on some subset of the disk farm, and do incrementals on the >>rest. Another alternative would be to just buy several 8mm drives. Since large disk farms usually mean a VAXCluster, several CPUs can be at work at once saving different sets of the disks to different tape drives. This is the scheme we use for our LAVC. Each MicroVAX-III boot node has an 8mm drive on it; we run two backup batch jobs nightly to backup separate disk sets on each tape. We'll be able to almost triple our disk farm before we need additional tape drives. On factor helping here is the (relatively) low cost of the 8mm tape drives. >>Digital data drives based on the DAT standards (4 mm wide tape) are rumoured >>to be just around the corner. Capacity will be smaller (1 Gbyte per cart) but >>the carts are smaller, file-skip times may be better, and throughput and >>data reliability may be better. On the other hand, the file-skip times are too bad on the 8mm drives. Also, the rumor is that the density of the 8mm drives will like quadruple this year! Yep, that 9+ GB on a single 8mm cassette probably with a somewhat similar increase in the data transfer rates. It looks like 8mm tapes are capturing the "midrange" systems; Digital DAT will likely be a major factor in lower-end systems (consider a 3100 with a DAT tape and a 600 MB disk drive). ================================================================================ Note 335.11 Offline Storage Alternatives / Backup 11 of 22 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 15 lines 25-MAR-1989 18:01 -< Basic != widely-known, apparently. >- -------------------------------------------------------------------------------- > > making recovery from incrementals more tolerable is NOT to write them with > > /RECORD. Then the incrementals are cumulative; you need only the latest > > one, plus the most recent image. Of course, the incrementals get longer > > protection, do incrementals on the set that you're going to do the image > > on, just before doing the image. Then if the most recent image backup > > goes bad you can go back to previous image + last incremental. You get > These are all good reminders from Jamie, but I thought they were > *basic* techniques that DECUS should be teaching in Introduction > to Backup SITE SIG sessions at Symposia. That's how I think of them -- basic. But I mentioned them because I run into a surprising number of people who've never heard of them. ================================================================================ Note 335.12 Offline Storage Alternatives / Backup 12 of 22 EISNER::SCOPELLITI "Whatsa behind is uva no importa" 10 lines 27-MAR-1989 22:43 -< Got a big off-site backup site? >- -------------------------------------------------------------------------------- One comment: Something that seems to be overlooked is the fact that a rolling backup sorta assumes that each disk is independent of the other. In an ALL-IN-1 site, each user's document database can contain pointers to files on several disks. Given this, what is your off-site backup set? You can't just use the last full backup pass, but rather have to also include the incrementals of the older backups. May not be easy, but if you use 10 1/2 inch reels, your operators will have some serious muscles. ================================================================================ Note 335.13 Offline Storage Alternatives / Backup 13 of 22 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 23 lines 28-MAR-1989 07:06 -< Adding fuel to the fire... >- -------------------------------------------------------------------------------- I came back from the weekend hoping to find one or two comments. TWELVE! I'm keeping a listing of this topic to show the beancounters when they ask me to justify DECUServe for another year... Anyway, thanks for the tips on getting more efficiency out of our standard backups. We do most of it already; BACKUP/IMAGE/BLOCK=32768 - /BUFFER=5/DENSITY=6520 is in our operator command procs. We do a standalone on the system disk Monday AMs, weekly full and daily incrementals on the code disks, and weekly fulls on most of the data disks. Some stage their data so fast there's no reason to back them up, but then we also have shelves of RA60 packs that get backed up once a week if they get used. One major thrust towards the alternative media is the storage space involved. We have no possibility of increasing our computer space for another few years at best, and every time a new project starts up, someone wants to dump another CPU, disk stack, printer, and tape unit in the middle of an already crowded room. Room for another tape rack is just not there. That's one of the reason the 8mm tapes are looking good. How many 10.5 inch magtapes equal one TA90 cartridge? I'm still hoping for some good words about optical, too. Are there any? :-} ================================================================================ Note 335.14 Offline Storage Alternatives / Backup 14 of 22 EISNER::BONE "George Bone, MINSY" 9 lines 28-MAR-1989 19:18 -< Think about what you just said. >- -------------------------------------------------------------------------------- < few years at best, and every time a new project starts up, someone wants to - -------------------------------------------------------------------------------- << of the reason the 8mm tapes are looking good. How many 10.5 inch magtapes << equal one TA90 cartridge? As I recall, one TA90 cartridge holds 400MB, while a 10.5 inch magtape holds about 140 MB, so we're talking about 3:1 in capacity. ================================================================================ Note 335.16 Offline Storage Alternatives / Backup 16 of 22 EISNER::NAGY "Frank J. Nagy, VAX Guru" 7 lines 29-MAR-1989 08:02 -< TA90s is only about 280 MB >- -------------------------------------------------------------------------------- >> As I recall, one TA90 cartridge holds 400MB, while a 10.5 inch magtape >> holds about 140 MB, so we're talking about 3:1 in capacity. I thought that the IBM 3480 (or is it 3840? :-), aka TA90 in DECland, had a capacity of either 270 or 290 MBs (my memory, not dual-capacities) at this time. This makes the TA90 to 10.5 reel only 2:1 in per unit storage. ================================================================================ Note 335.17 Offline Storage Alternatives / Backup 17 of 22 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 11 lines 29-MAR-1989 09:51 -< Where is a good place to store tapes? >- -------------------------------------------------------------------------------- >> Do you keep your tapes in the computer room? Our ADP security heavies >> told us to get ALL of our tapes not currently in use out of there, >> which is keeping us busy trying to find vault space. It's a good >> idea, though. If you have a fire in the computer room, how will >> you recover if your tapes are in ther, too? This is an excellent point of contention. Security considerations favor locating tapes outside the computer room, and for cogent reasons. OTOH,H, how many sites have more than one location with the temperature and humidity controls ideally suited to tape storage? ================================================================================ Note 335.18 Offline Storage Alternatives / Backup 18 of 22 EISNER::MAYHEW "Bill Mayhew" 21 lines 29-MAR-1989 10:19 -< Rent-a-vault >- -------------------------------------------------------------------------------- > OTOH,H, how many sites have more than one location with > the temperature and humidity controls ideally suited to tape > storage? Not that I particularly endorse any of them, but there *are* service companies that will come and pick up your tapes and store them for you in their climate-controlled vaults. (Disk packs, too, for that matter.) DEC even had an offering like this at one point, and built a pricey facility in the 'burbs of Boston to house it. Haven't heard anything about it recently, though. Look at the facility before signing, of course. I read a report somewhere (here, possibly!) from someone who had gone to look at such a facility and discovered that the vaults were water-tight -- but sprinkler-equipped, so that if the sprinklers went off, the room would fill with water! "Instar" is the name of another Boston-area company that does, or did, provide this type of service. I know nothing about them except that they have white trucks... ================================================================================ Note 335.19 Offline Storage Alternatives / Backup 19 of 22 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 13 lines 29-MAR-1989 14:29 -< Offsite not suitable for currently-used tapes. >- -------------------------------------------------------------------------------- >> < Rent-a-vault>- This is an excellent approach for storage of archival and/or backup media which will allow you to recover your business-critical data in case your entire facility burns down, but it is not particularly convenient for storage of tapes that are in use -- such as the most recent full backup and the current week's incrementals, or the daily tapes that one user writes and then accesses the next three days. The most desirable approach from a security viewpoint is to keep even these out of the computer room, but it is impractical for most users to put such tapes offsite. An in-house location which has a good environment and reasonable separation from the computer room is what is needed, and most sites don't have it. ================================================================================ Note 335.20 Offline Storage Alternatives / Backup 20 of 22 EISNER::MAYHEW "Bill Mayhew" 5 lines 29-MAR-1989 15:01 -< Cipher 3000? >- -------------------------------------------------------------------------------- Has anyone looked at the new Cipher 3000 series drives in any detail? These use 3480 cartridges, offer up to 570Mb/cartridge, and a transfer rate of "nearly" 1 megabyte/sec., with greater reliability than the 8mm format -- all of this according to the ads, of course. Are there any DEC-compatible subsystems based on this drive, yet? ================================================================================ Note 335.21 Offline Storage Alternatives / Backup 21 of 22 EISNER::COY "Dale E. Coy" 8 lines 29-MAR-1989 21:20 -< Tapes aren't computers (and vice-versa) >- -------------------------------------------------------------------------------- > ...favor locating tapes outside the computer room.... > OTOH,H, how many sites have more than one location with > the temperature and humidity controls ideally suited to tape storage? The "ideal" computer room environment is NOT necessarily ideal for storage of tapes. [OTOH, it's better than the average bank vault] ================================================================================ Note 335.22 Offline Storage Alternatives / Backup 22 of 22 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 15 lines 31-MAR-1989 00:26 -< what tapes want is constancy. >- -------------------------------------------------------------------------------- > The "ideal" computer room environment is NOT necessarily ideal for > storage of tapes. True, if you want really long-term storage for tapes you want to store them under relatively dry conditions (too dry in computer room =leads to static buildup), and I believe that you also want to keep the tapes at a somewhat cooler temp. than the 70 or so that computers live at. (Not that the machines wouldn't be happy at 50 F, but people have to be happy in the computer room too.) But the main thing to do for tapes, given that conditions aren't BAD for them, is to make sure that the conditions don't change too rapidly or too often. And if they're stored under different conditions than the computer room, give them an hour or two to adapt before trying to mount them.  ================================================================================ Note 336.0 VSII monitor (VR260) power failure problems 16 replies EISNER::PROVOST 7 lines 23-MAR-1989 13:43 -------------------------------------------------------------------------------- We have 6 VAXstation II mono workstations. They are extremely reliable, BUT often when we undergo power failure we lose at least one monitor. Is this a common problem? Is there a fix? Powerless Tom ================================================================================ Note 336.1 VSII monitor (VR260) power failure problems 1 of 16 EISNER::MADISON "Matt Madison" 10 lines 23-MAR-1989 19:12 -< Appears to be common problem >- -------------------------------------------------------------------------------- We've certainly been hit hard by this apparent design problem in the VR260. After every power outage we have had to fix at least one or two, and once we lost 12 in one fell swoop. The part runs about $100 through DECmailer, and according to our h/w person, DEC has had quite a time keeping up with demand. I wish I could remember what it's called -- it's the board that's just inside the monitor's case on the left. Anyway, the replacement parts we've received in the last month or so appear to be different from the old ones (according to our tech), so perhaps they've improved the design. Only time will tell. ================================================================================ Note 336.2 VSII monitor (VR260) power failure problems 2 of 16 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 18 lines 23-MAR-1989 21:25 -< Another victim of failed monitors >- -------------------------------------------------------------------------------- >> We've certainly been hit hard by this apparent design problem >> in the VR260. After every power outage we have had to fix at >> least one or two... I have used two different VR260 monitors on my system and both of them have died. The symptoms were a collapsed raster; one partially and the other completely. >> The part runs about $100 through DECmailer, and according to >> our h/w person, DEC has had quite a time keeping up with >> demand. I wish I could remember what it's called -- it's the >> board that's just inside the monitor's case on the left. 54-15624-01 VR260 Deflection Board. My current one has been working OK for about 10 months now, but I haven't turned the power off more than about a dozen times. This seems to be the best argument yet for leaving the monitors on all the time. ================================================================================ Note 336.3 VSII monitor (VR260) power failure problems 3 of 16 EISNER::NAGY "Frank J. Nagy, VAX Guru" 6 lines 24-MAR-1989 08:45 -< Yes, VR260s like to dead on power off >- -------------------------------------------------------------------------------- Yes, we too have run into the problem of VR260s dead with (usually) collapsed raster after either a power failure or when moving a system to another office. Luckily we had some "spare" VR260s lying about and could swap'em quickly. Alternatively, we have had no problems with the VR290 color monitors. ================================================================================ Note 336.4 VSII monitor (VR260) power failure problems 4 of 16 EISNER::MAYHEW "Bill Mayhew" 8 lines 24-MAR-1989 10:28 -< VR260 OK after two years, but... >- -------------------------------------------------------------------------------- I have one VR260 that has been here for almost two years, and gets power-cycled daily. Still works fine, though the raster is not stable any longer, a problem that began about two weeks ago and which I have been too lazy to get fixed. (Every now and again the screen image starts jittering, shrinking and growing, as a garden-variety TV would if its power source became very unstable.) Perhaps my problem is the forerunner of the more serious failure. Methinks I will get off my butt and get it fixed... ================================================================================ Note 336.5 VSII monitor (VR260) power failure problems 5 of 16 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 10 lines 24-MAR-1989 14:49 -< VR260 may have some failings, but I still love it >- -------------------------------------------------------------------------------- >> (Every now and again the screen image starts jittering, >> shrinking and growing, as a garden-variety TV would if its >> power source became very unstable.) Perhaps my problem is the >> forerunner of the more serious failure. Mine had that problem from day one until it died. After replacing the deflection card, it only happens rarely. My VR260 is now 2.5 years old and I love it! Having a 55 line screen has made me much more productive than using a VT-100. ================================================================================ Note 336.6 VSII monitor (VR260) power failure problems 6 of 16 EISNER::MAYHEW "Bill Mayhew" 7 lines 24-MAR-1989 15:32 -< Gotta have my VAXstation >- -------------------------------------------------------------------------------- > Having a 55 line screen has made me much more productive than > using a VT-100. Well, you do need something to drive it with :-}, but I WHOLEHEARTEDLY agree. I really think there is no excuse for a VMS programmer NOT to have a VAXstation. The productivity gains will easily pay for a base VS2000 within 2 years, probably 1. ================================================================================ Note 336.7 VSII monitor (VR260) power failure problems 7 of 16 EISNER::HARVEY "Jack Harvey" 23 lines 24-MAR-1989 18:30 -< Power Failure vs Powering Off >- -------------------------------------------------------------------------------- Tom's original question had to do with *power failure*. There seems to be a consensus this causes VR260 failure. But power failure ain't necessarily the same as turning the unit off with its power switch. I think we have a note indicating they seldom turn the VR260 off (possibly implying that protects it), and another indicating it's turned off daily with no problem. I'd like to get this resolved if possible. Does anyone have any clear cut experience of the on/off switch on the unit causing failures? How about unplugging the power cord with it turned on? There are three different situations here: 1. Real power line failures are frequently proceeded by voltage surge. 2. The power switch in some equipment sometimes does a bit more than just open the circuit, to protect stuff. 3. Pulling the power cord out is like a power line failure but without the surge. ================================================================================ Note 336.8 VSII monitor (VR260) power failure problems 8 of 16 EISNER::KOZAM 18 lines 24-MAR-1989 20:52 -< I've seen this before... >- -------------------------------------------------------------------------------- I turn on/off my VR260 every day without problems, with on-time approximately 6-10 hours/day. No problems. I have clients who turn leave them on continuously (and have very stable power), but have had about half of their monitors fail at one time or another. Even though the screen blanks, the high voltage circuits are still going. It has been my experience that the total time on correlates best with component failure. This was also true with the VT1XX series, which had a habit of blowing their video boards if they were left on. Since the VT1XX failures were so common, I always repaired them myself. It was ALWAYS the same electrolytic capacitor (usually visually fried) plus occasionally an associated diode. For about $ 4 at Radio Shack, I had enough parts to repair about 10 terminals. Has anyone identified the vulnerable component in the VR260 deflection board? Does anyone have the printset? ================================================================================ Note 336.9 VSII monitor (VR260) power failure problems 9 of 16 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 9 lines 24-MAR-1989 21:11 -< Only failure to report... >- -------------------------------------------------------------------------------- >> Has anyone identified the vulnerable component in the VR260 >> deflection board? I tried, but had no success. >> Does anyone have the printset? I asked the same question back in note 88.0, but never got a response. I suspect that DEC doesn't sell it. ================================================================================ Note 336.10 VSII monitor (VR260) power failure problems 10 of 16 EISNER::NAGY "Frank J. Nagy, VAX Guru" 9 lines 25-MAR-1989 11:08 -< Power failure AND poweroff failures of VR260s >- -------------------------------------------------------------------------------- >> Does anyone have any clear cut experience of the on/off >> switch on the unit causing failures? How about unplugging >> the power cord with it turned on? There are three different >> situations here: We leave our VR260s on continuously. We've also seen both types of failures. My VR260 failed to come back after a site-wide power failure. At least two other VR260s I know about failed when turned off (with the power switch) in order to move the systems to another location. ================================================================================ Note 336.11 VSII monitor (VR260) power failure problems 11 of 16 EISNER::PROVOST 31 lines 28-MAR-1989 11:08 -< More data >- -------------------------------------------------------------------------------- re: -< Power Failure vs Powering Off >- > Does anyone have any clear cut experience of the on/off > switch on the unit causing failures? How about unplugging > the power cord with it turned on? Our monitors are connected to the uVAXen via patch panels which allow us to connect any monitor to any reasonably close uVAX. We power off monitors only to move them or for scheduled power failures. When there is a scheduled power outage, we power down all computer equipment in advance, and power up only after voltage is stable. When there is an unscheduled power failure, we power down all computer equipment, and power up only after voltage is stable. We have never lost a monitor through power cycling for moving it. We have heard that the order of shut down of power for the monitor and the uVAX matters. In the past we have shut the uVAX down first and turned it on last. We are now shutting the monitor down first and turning it on last. There have been no failures since, but we have only one data point. We have lost a monitor when we allowed the system to go down and up without attention. ================================================================================ Note 336.12 VSII monitor (VR260) power failure problems 12 of 16 EISNER::PROVOST 10 lines 15-MAY-1989 15:04 -< Latest from Atlanta >- -------------------------------------------------------------------------------- re: < Note 336.11 by EISNER::PROVOST > DEC is aware of the problem... A component on the video deflection board was under spec. Digital chose not to issue ECO-FCO's. Component will be swapped as boards come in for repair. This is rough on my 3rd party maintenance company. If anyone has more details, I'd be interested. ================================================================================ Note 336.13 VSII monitor (VR260) power failure problems 13 of 16 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 10 lines 22-JUN-1989 19:36 -< 5 volt non-polarised cap needed >- -------------------------------------------------------------------------------- A F/S type confessed that he had on several occasions taken to fixing these himself, but asked us to NEVER mention it to his boss. These must be dying fast enough so there is a supply problem through normal DEC channels. He said it is always a 5 mfd non-polarised cap, but I didn't get the voltage. Obviously get higher voltage than what is there. For Boston area folks: HE buys his caps at You-DO-IT electronics near channel 5. ================================================================================ Note 336.14 VSII monitor (VR260) power failure problems 14 of 16 EISNER::KOZAM 9 lines 22-JUN-1989 19:58 -< Great Note for Self-Maintainers >- -------------------------------------------------------------------------------- Terrific news. It's far easiser to replace a cap than it is to stomach sending the board out at substantial cost and waiting for repair. Too bad DEC has become so stubborn about hardware docs. > He buys his caps at You-DO-IT electronics... We used to call it You-Blew-IT, for obvious reasons. Their selection is the best I've seen anywhere (except, of course, the large distributors who are resistant to sell anything quantity 1). ================================================================================ Note 336.15 VSII monitor (VR260) power failure problems 15 of 16 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 37 lines 27-SEP-1989 13:33 -< My eyballs are getting tired of bouncing up and down >- -------------------------------------------------------------------------------- < Note 336.4 by EISNER::MAYHEW "Bill Mayhew" > > I have one VR260 that has been here for almost two years, and > gets power-cycled daily. Still works fine, though the raster > is not stable any longer, a problem that began about two weeks > ago and which I have been too lazy to get fixed. (Every now > and again the screen image starts jittering, shrinking and > growing, as a garden-variety TV would if its power source > became very unstable.) Perhaps my problem is the forerunner of > the more serious failure. Methinks I will get off my butt and > get it fixed... Did you ever find out what caused the jitter? Mine has gotten rather annoying lately and won't go away. Of course it may be related to the latest "incident": I leave my monitor on almost all the time. Several evenings ago, the raster started collapsing and expanding at a rapid rate, accompanied by an arcing noise from inside the monitor. Examination showed that the connector for the deflection yoke (on the board at the left side of the monitor) had overheated and failed. Both the male and female connectors were melted and charred around one pin, and the PC board was also charred around the same pin. Fortunately I was able to find an exact replacement Molex connector and remove the charred area from the PC board. One etch has now been replaced with a piece of wire. I thought that both this problem and the jitter were due to dirty connectors or other bad connections. However, wiggling all of the connectors seems to have no effect on the jitter. Could this be caused by the bad capacitor mentioned in an earlier note? The other obvious possibility was dirty trimpots, but that proved false. By the way, the jitter is only vertical, not horizontal. ================================================================================ Note 336.16 VSII monitor (VR260) power failure problems 16 of 16 EISNER::MAYHEW "Bill Mayhew" 25 lines 27-SEP-1989 16:56 -< Jitter gone, linearity problems won't go away >- -------------------------------------------------------------------------------- > Did you ever find out what caused the jitter? Well, yeah, it was the monitor. ;-} DEC replaced it, inducing various and sundry other problems. I think they cycled 4 replacement monitors through here over the past few months, the latest couple being VR262s. The VR262 I have now is marginally acceptable, where "acceptable" really means it's too much trouble fighting with DEC to get it _right_ (if indeed they have any that are right). The problems with the replacements have mostly revolved around nonlinearities in the display, of various types. (For instance this one has a _slight_ inward bow on both the left and right sides (about 1 pixel worth), a slight blip upwards at the top left corner area (maybe 2 pixels), and all four corners are somewhat "pushed in" in both directions for about an inch along each side -- "pushed in" by about 3 pixels. Evidently I was lucky with my original VR260's display being virtually perfectly rectangular. Unfortunately by the time I figured out that I was lucky, it was way too late to get the original VR260 back and demand that they fix the jittering problem without swapping everything else out... The one I have now is the least objectionable of any of the replacements, but if I were to see it at a demo or trade show, I'd make a mental note not to buy a VR262.  ================================================================================ Note 337.0 DILOG DQ153 - Tape Drive, Tape Formatter - Pertec 1 reply EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 11 lines 25-MAR-1989 01:44 -------------------------------------------------------------------------------- I just unpacked my first Dilog DQ153 and have two questions not covered in the manual..... 1) The manual shows *FOUR* tape drives with *FOUR* formatters connected to the DQ153. Well I knew the Pertec interface supported two formatters but four? Anyway the manual does not explain how to identify which drive goes with what formatter. Does the interface have this capability - and how do I use it? 2) Does anyone know what jumpers J1 through J10 are used for? ================================================================================ Note 337.1 DILOG DQ153 - Tape Drive, Tape Formatter - Pertec 1 of 1 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 14 lines 31-MAR-1989 13:20 -< TO ANSWER MY OWN QUESTION >- -------------------------------------------------------------------------------- > 1) The manual shows *FOUR* tape drives with *FOUR* formatters > connected to the DQ153. Well I knew the Pertec interface supported two > formatters but four? Anyway the manual does not explain how to identify > which drive goes with what formatter. Does the interface have this > capability - and how do I use it? The DQ153 does *NOT* support the formatter address line (FAD). If you want to use multiple formatters you must be able to set the formatters so they will only respond to unit numbers that are attached to them or switch the TAD0 or TAD1 line to FAD. > 2) Does anyone know what jumpers J1 through J10 are used for? They are used to convert the DQ153 to a DQ152  ================================================================================ Note 338.0 How's the LNO1 ? 12 replies EISNER::SIMONS "Paul, Not that CONVEX!" 3 lines 27-MAR-1989 17:05 -------------------------------------------------------------------------------- We have the opportunity to obtain a LN01S-LH cheap. Does anyone use these and care to comment on their reliability/performance ? Is there a way that I could convert it to an LP11 interface ? ================================================================================ Note 338.1 How's the LNO1 ? 1 of 12 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 4 lines 27-MAR-1989 17:49 -< Close... >- -------------------------------------------------------------------------------- We have a printer based on the same Xerox print engine as the LN01. Its big, prone to break, is touchy about supplies, has mediocre print quality, and costs a mint to maintain. Other than that, its great. ================================================================================ Note 338.2 How's the LNO1 ? 2 of 12 EISNER::COY "Dale E. Coy" 7 lines 27-MAR-1989 21:28 -< A good production printer >- -------------------------------------------------------------------------------- You'll either love it, or hate it. It's an EXCELLENT high-volume production printer, but needs to be kept fairly busy. If you're hunting a workgroup printer (less than about 10 users), it's expensive to maintain. The serial interface was an add-on. The original parallel interface should still be there. ================================================================================ Note 338.3 How's the LNO1 ? 3 of 12 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 13 lines 28-MAR-1989 07:14 -< LN uh-oh 1 >- -------------------------------------------------------------------------------- > We have a printer based on the same Xerox print engine as the < > LN01. Its big, prone to break, is touchy about supplies, has < > mediocre print quality, and costs a mint to maintain. Other < > than that, its great. < Essentially ditto. We have an LN01. The print quality is OK, but we've had Field Service out here every three weeks this year fixing the fool thing. We bought 2 LN03 printers (with which we've had setup problems; see earlier topics on LN03's); we figured the maintenance costs on two LN03's will be less than on one LN01. When it works, it's a good printer. When it breaks (which is too often), it's an expensive boat anchor. ================================================================================ Note 338.4 How's the LNO1 ? 4 of 12 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 3 lines 28-MAR-1989 13:07 -< Xerox 2700 engine: works great, lasts a long time >- -------------------------------------------------------------------------------- The LN01 uses the Xerox 2700 engine, which we have found to be a very reliable engine, when maintained by Xerox (which they will do for about $2,800 per year no matter what the volume of paper you push. ================================================================================ Note 338.5 How's the LNO1 ? 5 of 12 EISNER::CERMAK "Denny Cermak -3M- (612)733-3242" 7 lines 28-MAR-1989 15:01 -< Xerox Engine? >- -------------------------------------------------------------------------------- Has the LN01 always used the Xerox engine? We bought one back in 85 for our Mtn View site, and during one of the MANY service calls the FE told me it had a Canon engine. If you've got to buy DEC I'd recommend the LN03...cheaper upfront and for service (last I looked). ...Denny ================================================================================ Note 338.6 How's the LNO1 ? 6 of 12 EISNER::BURNET "John Burnet" 15 lines 28-MAR-1989 19:33 -< Old Xerox engine: Is it worth the maintenance expense? >- -------------------------------------------------------------------------------- > Has the LN01 always used the Xerox engine? Yes. Although this engine may be a workhorse when it's in a good mood, it's pretty obsolete compared to current laser-engine technology. Usually, DEC's monthly maintenance charges are a fairly good indicator of how reliable and easy to maintain a particular technology is... for example, compare the monthly charges for VT100s and VT220s. We just took all of our VT1xx terminals off of DEC service because of this. We'll use them until they die, and replace them with 220s or 320s. That's not feasible for an LN01, of course; so you might want to consider before buying it just what you'll be paying for service, and whether it would be better to take that money, add it to the cost of the LN01, and get an LPS40 or a few Talaris 1590s or something. ================================================================================ Note 338.7 How's the LNO1 ? 7 of 12 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 15 lines 29-MAR-1989 08:58 -< Xerox Owner >- -------------------------------------------------------------------------------- I have had a Xerox 2700 for over 5 years now and what everybody else has said is basically true. When they work, they work good. When they don't you had better get Xerox Service to fix them because DEC just doesn't seem to have the touch. I just got a note from Xerox about increase in rates for maintaining 2700's. The new rates are $205/month + $0.012 over 10,000 pages/year Maybe the reason your getting that LN01 cheap is the former owner couldn't afford the maintenance ???!!?! ================================================================================ Note 338.8 How's the LNO1 ? 8 of 12 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 12 lines 29-MAR-1989 10:01 -< Are we going to get stuck? >- -------------------------------------------------------------------------------- RE: < Note 338.7 by EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" > -< Xerox Owner >- | $205/month | + $0.012 over 10,000 pages/year Oh, oh. If we get that notice, we'll be in trouble. While the per month charge is compatible with what we now pay, we run an average of over 750,000 pages per year through each of our 2700 engines (and 1,440,000 per year through our 3700 engine). That's another $9,000. I have a hard time believing that's not 10,000 pages per month, since the 2700 engine certainly can handle more than that. ================================================================================ Note 338.9 How's the LNO1 ? 9 of 12 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 23 lines 29-MAR-1989 10:05 -< 2700 Mechanism has its problems. >- -------------------------------------------------------------------------------- We had a Xerox 2700 on site for about three years. It was used for document-quality printing, and we averaged about 15,000 pages/ month. Print quality was very good for the first year (about 180,000 impressions), then began to decline -- and Xerox was never able to repair it so as to obtain our original (and desired) print quality, despite completely replacing the print mechanism at least three times. The biggest problem we had was with irregularities in the paper feed speed; there would be areas of expanded and/or compressed print on the sheet. Toner coverage became irregular, as well. Toward the end of our usage, it was necessary to send each multiple-page document twice to be sure of being able to assemble one of suitable quality. The Xerox repair person, of whom we saw a lot, reported that most 2700's exhibited these problems after about 200-250 K sheets. Xerox marketing types denied that there was any known wear-out factor to the 2700 (and 3700, which uses a very similar paper path). When the print quality became completely unacceptable and it became obvious that Xerox could not improve matters except with a new printer, we purchased another unit (Dataproducts 2665) which has shown none of the problems we had with the 2700. I would personally find it very difficult to recommend acquisition of a used printer that includes a Xerox 2700 mechanism, knowing the problems that we had with Xerox-supplied reconditioned mechanisms. ================================================================================ Note 338.10 How's the LNO1 ? 10 of 12 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 4 lines 29-MAR-1989 10:09 -< That's 10000 pages/MONTH >- -------------------------------------------------------------------------------- >> + $0.012 over 10,000 pages/year ^^^^ At least for us, that was per MONTH. ================================================================================ Note 338.11 How's the LNO1 ? 11 of 12 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 1 line 29-MAR-1989 15:07 -< ooppss >- -------------------------------------------------------------------------------- Yes, that's 10,000/month sorry ================================================================================ Note 338.12 How's the LNO1 ? 12 of 12 EISNER::SIMONS "Paul, Not that CONVEX!" 4 lines 30-MAR-1989 19:59 -< Thanks, but no thanks. >- -------------------------------------------------------------------------------- Thank you all for responding. When we told our in house hardware types we where thinking of buying LN01's, they did some fast checking, said, "That's where those went!", and told us no way were they going to maintain them.  ================================================================================ Note 339.0 System Industries shadowing No replies EISNER::MERRIMAN "George Merriman - CCA/NY" 4 lines 27-MAR-1989 18:38 -------------------------------------------------------------------------------- A client is looking into some kind of shadowing product supplied by System Industries based on a controller board and a couple of disk drives. That's all I know about it. Does anyone out there know any more? Does anyone care to comment?  ================================================================================ Note 340.0 TA90 VERSUS TK70 DUMB QUESTION 4 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 11 lines 29-MAR-1989 08:27 -------------------------------------------------------------------------------- !>> As I recall, one TA90 cartridge holds 400MB, while a 10.5 inch magtape !>> holds about 140 MB, so we're talking about 3:1 in capacity. ! !I thought that the IBM 3480 (or is it 3840? :-), aka TA90 in DECland, !had a capacity of either 270 or 290 MBs (my memory, not dual-capacities) !at this time. This makes the TA90 to 10.5 reel only 2:1 in per unit !storage. If this is the case - other than speed - why would someone buy a TA90 over a TK70? ================================================================================ Note 340.1 TA90 VERSUS TK70 DUMB QUESTION 1 of 4 EISNER::KENNEDY "Terry Kennedy" 17 lines 29-MAR-1989 08:50 -< 4.5 reasons (pardon the double #4) >- -------------------------------------------------------------------------------- > If this is the case - other than speed - why would someone buy a > TA90 over a TK70? 1) Real IBM 3480's have auto-loaders. You can fill up at least a dozen tapes and let it go. Actually, some variant is also available which feeds tapes to available drives as the become idle... 2) Media interchange with IBM shops 3) Reliability. Enough people have fond memories of the TK50 that they won't go near a TK-anything. Any low-end drive that has to have a set of traffic light LEDs to indicate when pulling the handle will not destroy the drive is a bit silly... 4) Not being sole-source trapped for media 4) Speed (still important) ================================================================================ Note 340.2 TA90 VERSUS TK70 DUMB QUESTION 2 of 4 EISNER::SHANNON "Terry C. Shannon, IDC" 2 lines 29-MAR-1989 19:51 -< But wait, there's more! >- -------------------------------------------------------------------------------- The TA90 cartridges are less expensive than TK50/TK70 cartridges, too! ================================================================================ Note 340.3 TA90 VERSUS TK70 DUMB QUESTION 3 of 4 EISNER::COY "Dale E. Coy" 3 lines 29-MAR-1989 21:13 -< TA91 ? >- -------------------------------------------------------------------------------- And, IBM has announced higher-density and higher-transfer-rate improvements for this device (but I've no idea what DEC's support plans are). ================================================================================ Note 340.4 TA90 VERSUS TK70 DUMB QUESTION 4 of 4 EISNER::KILGALLEN "Larry Kilgallen" 3 lines 30-MAR-1989 00:10 -< TA90.5 ? >- -------------------------------------------------------------------------------- Some of our LUG members told the DEC speaker they thought DEC would have to get a new controller board for the HSC in order to support the higher transfer rate. He did not disagree (as I recall).  ================================================================================ Note 341.0 TA81 cached? 2 replies EISNER::NORTON "Bill Norton" 11 lines 31-MAR-1989 09:49 -------------------------------------------------------------------------------- When I do a $SHOW DEVICE/FULL $1$MUA3: (a TA81 on an HSC50) it responds with everything normal except any mention of cache support. I would *swear* that before I installed V5.1, I helped debug a backup procedure based on what $SHO DEV reported about the tape cache being enabled or disabled on this same hardware. Can someone verify for me whether a TA81 actually has a cache? ^ ================================================================================ Note 341.1 TA81 cached? 1 of 2 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 2 lines 31-MAR-1989 17:16 -------------------------------------------------------------------------------- Nope, the TA81 does not have a cache. ================================================================================ Note 341.2 TA81 cached? 2 of 2 EISNER::RENES "Stuart Renes, DFWLug Chair" 6 lines 6-APR-1989 09:44 -< + Cache >- -------------------------------------------------------------------------------- It's the TU81+ that has a cache. Note, however that as of VMS 5.1 there is still a problem of the cache not showing as enabled, when in fact, is is. Stuart.  ================================================================================ Note 342.0 SIGMA DHV11'S 4 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 7 lines 31-MAR-1989 23:22 -------------------------------------------------------------------------------- We just upgraded a MVII system to a PDP-11/83. In the process of doing so we found the SIGMA DHV11/16's hung the system. Be advised that the early versions of the SIGMA DHV11/16's will not work with floating point 11/83's. The board can be identified by the fact it has only one square metal clock crystal instead of two on the new ones. The board must be returned to SIGMA for a re-work. ================================================================================ Note 342.1 SIGMA DHV11'S 1 of 4 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 1 line 3-APR-1989 10:36 -< DEC wouldn't like your definition of "upgrade" >- -------------------------------------------------------------------------------- > We just upgraded a MVII system to a PDP-11/83. ================================================================================ Note 342.2 SIGMA DHV11'S 2 of 4 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 8 lines 3-APR-1989 16:11 -------------------------------------------------------------------------------- ! -< DEC wouldn't like your definition of "upgrade" >- ! !> We just upgraded a MVII system to a PDP-11/83. AH! but the customer does. If you replace a poorly written VMS application with a well written RSTS application most end users will consider it an upgrade. ;-) ================================================================================ Note 342.3 SIGMA DHV11'S 3 of 4 EISNER::SMITHJR "Charles T. Smith, Jr." 5 lines 3-APR-1989 22:09 -< If you'd really rather have a PDP... >- -------------------------------------------------------------------------------- > We just upgraded a MVII system to a PDP-11/83. Would you care to trade the MVII cpuset for, say a PDP-11/60? ================================================================================ Note 342.4 SIGMA DHV11'S 4 of 4 EISNER::SCOPELLITI "Whatsa behind is uva no importanc" 1 line 3-APR-1989 23:43 -< Depends what you're after >- -------------------------------------------------------------------------------- If your goal was floating point performance, in a microsecond!  ================================================================================ Note 343.0 Digi-Data 1740 series tape drive 14 replies EISNER::LEDERMAN "Bart Z. Lederman" 15 lines 4-APR-1989 11:07 -------------------------------------------------------------------------------- We have an opportunity to connect a Digi-Data tape drive to our MicroVAX-II. (The drive is currently on a CCC system.) The model number is 1749-86-4-120-FD-UL which, according to our manual, means it's 45 IPS, 800/1600 BPI, embedded formatter. What it doesn't say is if the interface is a standard (such as Pertec). It does use two 50 pin ribbon connectors, which look fairly standard. Has anyone used this drive, or another of the same series (1140, 1640, 1740, 1840), and if so with which coupler or controller? Also, the manual doesn't give a current address or telephone number for Digi-Data and I no longer have access to Gold Book so an address for Digi-Data and/or a local contact for a coupler (Dilog?) supplier (I'm now in Cambridge/Boston) would be appreciated. ================================================================================ Note 343.1 Digi-Data 1740 series tape drive 1 of 14 EISNER::FRISBIE "Veni, Vidi, $cmkrnli, rebooti" 6 lines 4-APR-1989 12:49 -< Formatter connections >- -------------------------------------------------------------------------------- >> What it doesn't say is if the interface is a standard (such as >> Pertec). It does use two 50 pin ribbon connectors, which look >> fairly standard. If you have the pin-outs for the formatter, I can tell you if it matches the documentation for anything I have. ================================================================================ Note 343.2 Digi-Data 1740 series tape drive 2 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 78 lines 4-APR-1989 14:03 -< Here's what the manual says. >- -------------------------------------------------------------------------------- Here is what I can get out of the formatter operation manual. Transmitters are 7406/7407 or equiv. Open collector Low true. Receivers are 8837 or equiv. There are two sets of 50 pin connectors: Digi-Data interface pair JA / JB JA JB __________________________________ __________________________________ | 1 | (+5V) | GND | 2 | | 1 | REW | GND | 2 | | 3 | WP.C | GND | 4 | | 3 | TAD1 | GND | 4 | | 5 | W0 | GND | 6 | | 5 | TAD0 | GND | 6 | | 7 | W1 | GND | 8 | | 7 | LWD | GND | 8 | | 9 | W2.B | GND | 10 | | 9 | FPT | GND | 10 | | | | | | | | | | | | 11 | W3.A | GND | 12 | | 11 | LPT | GND | 12 | | 13 | W4.8 | GND | 14 | | 13 | EOT | GND | 14 | | 15 | W5.4 | GND | 16 | | 15 | NRZ | LOL | 16 | | 17 | W6.2 | GND | 18 | | 17 | SGL | GND | 18 | | 19 | W7.1 | GATEWP | 20 | | 19 | SPD SEL | GND | 20 | | | | | | | | | | | | 21 | FBY | GND | 22 | | 21 | WSTR | GND | 22 | | 23 | DBY | IDENT/CCG | 24 | | 23 | RSTR | GND | 24 | | 25 | DEN STAT | GND | 26 | | 25 | R0 | GND | 26 | | 27 | HER | CER | 28 | | 27 | R1 | GND | 28 | | 29 | FMK | GND | 30 | | 29 | R2.B | GND | 30 | | | | | | | | | | | | 31 | CCG | 7 TK | 32 | | 31 | R3.A | GND | 32 | | 33 | RDY | GND | 34 | | 33 | R4.8 | GND | 34 | | 35 | ONL | GND | 36 | | 35 | R5.4 | GND | 36 | | 37 | RWD | GND | 38 | | 37 | R6.2 | GND | 38 | | 39 | THR2 | THR1 | 40 | | 39 | R7.1 | GND | 40 | | | | | | | | | | | | 41 | ERASE | GND | 42 | | 41 | RP.C | GND | 42 | | 43 | EDIT | WFM | 44 | | 43 | FEN | GND | 44 | | 45 | WRITE | GND | 46 | | 45 | FAD | GND | 46 | | 47 | REV | GND | 48 | | 47 | GO | GND | 48 | | 49 | OFL | GND | 50 | | 49 | PAR | DEN | 50 | ---------------------------------- ---------------------------------- Alternate interface connectors JC / JD. JC JD __________________________________ __________________________________ | 1 | GND | FBY | 2 | | 1 | RP.C | R0 | 2 | | 3 | GND | LWD | 4 | | 3 | R1 | LPT | 4 | | 5 | GND | W4.8 | 6 | | 5 | GND | R4.8 | 6 | | 7 | GND | GO | 8 | | 7 | GND | R7.1 | 8 | | 9 | GND | W0 | 10 | | 9 | GND | R6.2 | 10 | | | | | | | | | | | | 11 | GND | W1 | 12 | | 11 | GND | HER | 12 | | 13 | GND | SGL | 14 | | 13 | GND | FMK | 14 | | 15 | GND | LOL | 16 | | 15 | GND | CCG/IDENT | 16 | | 17 | GND | REV | 18 | | 17 | GND | FEN | 18 | | 19 | GND | REW | 20 | | 19 | GND | R5.4 | 20 | | | | | | | | | | | | 21 | GND | WP.C | 22 | | 21 | GND | EOT | 22 | | 23 | GND | W7.1 | 24 | | 23 | GND | OFL | 24 | | 25 | GND | W3.A | 26 | | 25 | GND | NRZ | 26 | | 27 | GND | W6.2 | 28 | | 27 | GND | RDY | 28 | | 29 | GND | W2.B | 30 | | 29 | GND | RWD | 30 | | | | | | | | | | | | 31 | GND | W5.4 | 32 | | 31 | GND | FPT | 32 | | 33 | GND | WRITE | 34 | | 33 | GND | RSTR | 34 | | 35 | GND | THR2 | 36 | | 35 | GND | WSTR | 36 | | 37 | GND | EDIT | 38 | | 37 | GND | DBY | 38 | | 39 | GND | ERASE | 40 | | 39 | GND | SPEED | 40 | | | | | | | | | | | | 41 | GND | WFM | 42 | | 41 | GND | CER | 42 | | 43 | GND | THR1 | 44 | | 43 | GND | ONL | 44 | | 45 | GND | TAD0 | 46 | | 45 | GND | TAD1 | 46 | | 47 | GND | R2.B | 48 | | 47 | GND | FAD | 48 | | 49 | GND | R3.A | 50 | | 49 | GND | DEN | 50 | ---------------------------------- ---------------------------------- ================================================================================ Note 343.3 Digi-Data 1740 series tape drive 3 of 14 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 13 lines 4-APR-1989 14:46 -< PERTEC INTERFACE >- -------------------------------------------------------------------------------- >Alternate interface connectors JC / JD. > > JC JD >__________________________________ __________________________________ >| 1 | GND | FBY | 2 | | 1 | RP.C | R0 | 2 | > >| 45 | GND | TAD0 | 46 | | 45 | GND | TAD1 | 46 | >| 47 | GND | R2.B | 48 | | 47 | GND | FAD | 48 | >| 49 | GND | R3.A | 50 | | 49 | GND | DEN | 50 | >---------------------------------- ---------------------------------- Looks like a standard PERTEC interface to me. ================================================================================ Note 343.4 Digi-Data 1740 series tape drive 4 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 29 lines 18-MAY-1989 22:05 -< Drive works when vector problems are solved >- -------------------------------------------------------------------------------- The project got the money to put in a Dilog coupler (153, I think, I'm at home). It works, and the manual even mentions the Digi-Data drive series a little. JC / JD is the right connection. The only problem is one of configuration. This coupler is one of the new devices where the vector is set by the operating system at run time. Very nice: except that all of the other devices (especially the DPV11 and the two DHQ11s) are the old style devices where the vector is set on the board. VMS sees the controller, decides to put it at a vector which is already in use by one of the other devices. SYSGEN SHOW /CONF tells you that the DHQ11s have "moved" to new vector assignments, but of course they haven't. So the tape controller works, but all of the terminal lines go dead. I suppose I could have manually set the device vectors in SYSGEN during system boot, but that would mean setting it up again for every major system upgrade, and possibly S/A Backup wouldn't work or would require a lot of extra manual effort on booting which untrained personnel wouldn't know how to do (besides, I didn't think of that approach in time). So we had to pull boards and reset vector assignments. (Did you know that the DPV11 sets address and vectors with wire-wrap terminals instead of switches? I suppose I should be happy it didn't use soldered jumpers.) I hope we never have to pull the tape coupler out for awhile, as all the terminals will go out of operation again. I suppose if all of the devices had software controlled vectors, this feature would be really neat. Otherwise, I'd just as soon have all vectors fixed in the device. ================================================================================ Note 343.5 Digi-Data 1740 series tape drive 5 of 14 EISNER::HARVEY "Jack Harvey" 21 lines 18-MAY-1989 22:46 -< SYSGEN CONFIG for Safe Vector/CSR Addresses >- -------------------------------------------------------------------------------- > VMS sees the controller, > decides to put it at a vector which is already in use by one of > the other devices. SYSGEN SHOW /CONF tells you that the DHQ11s > have "moved" to new vector assignments, but of course they haven't. Your explanation is perfectly correct, Bart. I'd just like to emphasize a point: This behavior is *not* caused by the third party controller. It's a perfectly normal change due to the floating vector and CSR address protocol observed by VMS. (One might argue for an improved protocol, but then we get into version compatibility problems, etc.) So long as you know to expect this, and know how to use the SYSGEN facility to advise you of the required changes, it's no big deal. The alternative, where you get to pick any old vector/CSR addresses you please, leads to RSX11M type sysgens. Remember them? Want to go back? I don't. I like having SYSGEN tell *me* where to put them, knowing that Field Service will also be able to find them with no hassle. ================================================================================ Note 343.6 Digi-Data 1740 series tape drive 6 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 5 lines 19-MAY-1989 07:49 -< System too smart for it's (our) own good >- -------------------------------------------------------------------------------- I agree that it's nice to have the system tell you where it expects to find devices (and by the way, there's a way to get RSX to tell you where you should place the devices as well: there's a command file in [200,200]). What I don't like is the system trying to assign floating vectors to devices that don't support floating vectors. ================================================================================ Note 343.7 Digi-Data 1740 series tape drive 7 of 14 EISNER::HARVEY "Jack Harvey" 31 lines 21-MAY-1989 19:09 -< Is an Alternate Site a Consideration? >- -------------------------------------------------------------------------------- > What I don't like is the system trying to assign > floating vectors to devices that don't support floating vectors. I presume you mean devices, such as the DZQ11, which require switch register setting of vectors. What's the alternative for such devices? Would you have the system boot process ask the operator what the vector is? Would you have a file on the system disk (a variation on the RSX scheme) tell the boot process what the vector is? And then what happens when you try to boot your disk on another hardware configuration down the hall? Remember, under RSX11M, it wouldn't work. Under M+, you were able to use CONFIG to correct things sometimes. But do you want operators doing that in the middle of the night on an emergency switch-over to an alternate site? Yes, the VMS scheme requires rigid adherence to the protocol, and that means when you add a device, you may have to alter one(s) already installed. And when you remove a device, you have to alter others back again. But when followed, *any* VMS system seems to be able to run on the hardware. (I'm sure there are exceptions that will blow it up, but I like the intent.) Although I've never tried it, I expect that someone having good and sufficient reason to break the protocol and put devices at addresses of their choice can do so with edits to SYSTARTUP_Vn.COM. ================================================================================ Note 343.8 Digi-Data 1740 series tape drive 8 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 18 lines 22-MAY-1989 08:01 -< VMS could do better >- -------------------------------------------------------------------------------- I'm not really all that concerned about what RSX does anymore (sorry). Since the disks other than RA60's are fixed, I'm not all that concerned about what happens when disks are moved. However, that shouldn't really affect the situation any, since I'm not proposing a configuration file on disk either. What I think VMS should do is have a list of what devices have dynamically allocatable vectors (like TMCSP controllers) and which devices do NOT have dynamically allocatable vectors (like DPV11 and DHQ11 controllers). VMS should look for the non-dynamic devices first, and let them use whatever vectors they are already set for. Then it should look for dynamic controllers and, since they can have any vector assigned to them, assign a vector to them which isn't already in use by something else. That would be "real" intelligence (as opposed to the current "artificial intellegence" approach). ================================================================================ Note 343.9 Digi-Data 1740 series tape drive 9 of 14 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 21 lines 22-MAY-1989 13:55 -< most cards can't tell youtheir vector. >- -------------------------------------------------------------------------------- There seems to be a terminology conflict here -- I think some people are saying "floating vector" when what they really mean is "programmable vector", ie the system can tell the device what its vector is. (I'm not sure that "programmable vector" is the official term, either. In the note quoted below, Bart refers to these as "dynamic vectors", which likewise may not be the official term.) Also... >. VMS should look for the non-dynamic devices > first, and let them use whatever vectors they are already set for. A nice idea, but not possible. There is no way to read the vector from the card. With some devices the host could force the device to interrupt and then see which vector was asserted by the device, but this isn't possible with all devices, and in any case would be very difficult to do robustly, esp. at other than boot time (worse than doing a show/unibus on a running system!). ================================================================================ Note 343.10 Digi-Data 1740 series tape drive 10 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 33 lines 2-JUN-1989 10:25 -< Seeing odd, irregular problems. >- -------------------------------------------------------------------------------- We are still seeing some odd problems. We can usually mount and read all of the contents of old tapes. When we try to write, the tape may go about 1/2 way through, then start re-trying very often. Trying to re-initialize that tape ends in a parity error. If we let things sit for a while and reboot, the problem goes away for a while. Dilog suggested setting the controller parameters to say "no remote density select". This means that 800/1600 is set by the button on the front of the drive (no problem, as it's almost always going to be 1600). But when we do a SHOW DEVICE/FULL MUB0: it tells us that the tape is at 6250 BPI instead of 1600. Does anyone know if this is a problem? Should we always put /DENSITY=1600 when we init and mount tapes, or doesn't VMS care? The other potential cause of the problem may be dirt. We will run a tape through, and then get errors. Cleaning the head will make the write errors go away (though even before we clean we can read old tapes fine). I just got that again this morning, using a set of tapes that had just gone through the data centers' tape certifier and cleaner. At the risk of starting another tape brand war: we are using Wabash DataTech QUADPlus tapes; does anyone know anything good or bad about them? This problem has been so intermittent (drive will work find for several reels of tape, then go nuts) that it's really driving the system manager and myself crazy. I should also mention that we've tried three different (physical) tape drives with the same Dilog coupler, and I think it's highly unlikely that all three could all be defective in the same way, especially as they were all working when disconnected from the other systems. ================================================================================ Note 343.11 Digi-Data 1740 series tape drive 11 of 14 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 8 lines 2-JUN-1989 12:02 -< Ask about ECOd micro-code for error handling >- -------------------------------------------------------------------------------- The Dilog 153 was going to get some microcode upgrades to improve error handling. I don't know when such changes were/willbe available, but check with Dilog about how current your microcode is, even if you got the controller fairly recently. One of the best guys they have for such questions is Vic Herring. He may have moved on to be PLM for their disc controllers, but will still KNOW what you need to find out. ================================================================================ Note 343.12 Digi-Data 1740 series tape drive 12 of 14 EISNER::NAGY "Frank J. Nagy, VAX Guru" 9 lines 3-JUN-1989 10:39 -< DQ153 needs new firmware for VMS V5.1 use >- -------------------------------------------------------------------------------- >The Dilog 153 was going to get some microcode upgrades to improve >error handling. I don't know when such changes were/willbe available, >but check with Dilog about how current your microcode is, even if you >got the controller fairly recently. After upgrading to VMS V5.1, our DQ153 caused VMS to crash when writing tapes (machine check) and programming the density stopped working. The spare DQ153 has been sent back to Dilog for a firmware upgrade and should be back soon. ================================================================================ Note 343.13 Digi-Data 1740 series tape drive 13 of 14 EISNER::LEDERMAN "Bart Z. Lederman" 3 lines 4-JUN-1989 12:21 -< It ought to be up to date! >- -------------------------------------------------------------------------------- Our DQ153 is only a few weeks old, and was purchased new from a Dilog rep. Is there any number I can check (like in the configuraton menu) that will tell me if I have the right firmware? ================================================================================ Note 343.14 Digi-Data 1740 series tape drive 14 of 14 EISNER::NAGY "Frank J. Nagy, VAX Guru" 3 lines 5-JUN-1989 15:11 -< DQ153 Rev. E current? >- -------------------------------------------------------------------------------- The revision label is supposedly stamped very visibly on the board. To the best of our memories here, our DQ153s are being upgraded to Rev. E, so we assume Rev. E is current.  ================================================================================ Note 344.0 Used VAXmate source? 1 reply EISNER::MAYHEW "Bill Mayhew" 2 lines 8-APR-1989 00:47 -------------------------------------------------------------------------------- Does anybody know of a source of used (or new, but cheap) VAXmates? We need one and have been unable to locate any. ================================================================================ Note 344.1 Used VAXmate source? 1 of 1 EISNER::CONROY "Alan Conroy" 3 lines 8-APR-1989 04:05 -< You probably already thought of these, but... >- -------------------------------------------------------------------------------- Besides the obvious (Midwest, Newman, etc), you might try the IRUG newsletter (magazine?) "BUY/SELL/TRADE" column. Seems to me I've seen something in past issues... Also, don't forget the SWAP conference!  ================================================================================ Note 345.0 MicroVAX II to VAXstation Conversion 2 replies EISNER::HARVEY "Jack Harvey" 16 lines 10-APR-1989 19:12 -------------------------------------------------------------------------------- With DECwindows now bundled, I'm interested in a cheap conversion of a MicroVAX II to a VAXstation II/GPX. I know DEC sells a kit for this, but since my interest has a limited budget, I would like to go with used parts. I know about the monitor(s), keyboard, mouse and video cable. What I'm looking for is the nomenclature of the circuit board(s) that go in the Q-Bus backplane, and the cabinet kit. I'm getting conflicting stories and am looking for other opinions. Also, I am aware that converting the MVII to a VSII will cause the LMF to convert it to a single user system. In my case it's not an issue. I mention it to warn readers who aren't aware of the problem, decide to do this also, and get burned. ================================================================================ Note 345.1 MicroVAX II to VAXstation Conversion 1 of 2 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 5 lines 17-APR-1989 11:35 -< I wonder if the price is going to jump because of DECwindows? >- -------------------------------------------------------------------------------- >> I'm interested in a cheap conversion of a MicroVAX II to a >> VAXstation II/GPX. Why not just call Midwest Systems (to name only one of many) and buy the complete kit? ================================================================================ Note 345.2 MicroVAX II to VAXstation Conversion 2 of 2 EISNER::HARVEY "Jack Harvey" 3 lines 18-APR-1989 01:03 -< I'm holding my breath >- -------------------------------------------------------------------------------- < -< I wonder if the price is going to jump because of DECwindows? >- Nah, DEC marketing is far too sharp to let that happen, wern't it?  ================================================================================ Note 346.0 8mm Tape/HSC/Reliable? 3 replies EISNER::MCCREADY "Gary, former DEC, DCL is Fun!" 30 lines 12-APR-1989 16:50 -------------------------------------------------------------------------------- I was thinking to add this to 335, but that was getting a bit long... So far I've only seen 8mm tape devices on an HSC from Micro Technology, and they are actually installing their first one this month, but already have a 45-day order backlog (so says salesperson). Not suprising, since the marriage of the 2.3 Gig backup device and HSC's seem to be perfect. Price on this one is about 20K for a dual (two 2.3 gig drives) unit. Not much cheaper for the single unit. so does anyone else out there make an 8mm/HSC? If so, how has it worked, throughput, etc..? And it general, is there any PROOF (aside from the 10**13 stat) as to how good these things really are for long term storage of data versus 9-track? Like published reviews from reliable sources, or BIG users doing year or multi-year comparisons? When I told the salesperson that I needed more than a letter from Exabyte saying they were good, she told me to do the research myself (!!!). So I am. Also any info on Micro Technology would be good as well. Anyway FYI, you can allededly stack up to 10 (I think) off of one tape "channel", and can allegedly access each one independtly. There are also throughput stats, but until I hear it from a "user"... --Gary McCready ================================================================================ Note 346.1 8mm Tape/HSC/Reliable? 1 of 3 EISNER::MCCREADY "Gary, former DEC, DCL is Fun!" 15 lines 16-APR-1989 15:51 -< Some Stats for above >- -------------------------------------------------------------------------------- Just a few stats from Micro Technology to add to above -Model Number MA-24, TA/TMSCP drive emulation -can stack up to 16 2Gig units to backup up to 32 Gig unattended -Connect 4 drives per HSC port, 16 per tape (HSC5X-CA)card -Has a 512KB buffer so it has a streaming transfer rate of 246KB/Second, and it has a peak transfer rate of 1.5 MB/Second (they say 32 Gig in 3 hours !!?!) - Size (single or dual drive unit) H5.25,W17.25,D25,25 -MTBF - 20,000 Hours Well, it sounds real good.... --Gary McCready ================================================================================ Note 346.2 8mm Tape/HSC/Reliable? 2 of 3 EISNER::RICE "Gary Rice" 16 lines 17-APR-1989 08:09 -------------------------------------------------------------------------------- > Just a few stats from Micro Technology to add to above > -Model Number MA-24, TA/TMSCP drive emulation > -can stack up to 16 2Gig units to backup up to 32 Gig > unattended ^^^^^^^^^^ > -Connect 4 drives per HSC port, 16 per tape (HSC5X-CA)card > -Has a 512KB buffer so it has a streaming transfer rate > of 246KB/Second, and it has a peak transfer rate of > 1.5 MB/Second (they say 32 Gig in 3 hours !!?!) > - Size (single or dual drive unit) H5.25,W17.25,D25,25 > -MTBF - 20,000 Hours > Well, it sounds real good.... Just one slight problem with MT's configuration. The cassette tape is too small to back up 1 of their 4-spindle "disks" if it is above 90% full. ================================================================================ Note 346.3 8mm Tape/HSC/Reliable? 3 of 3 EISNER::KENNEDY "Terry Kennedy" 15 lines 17-APR-1989 08:28 -< I don't think it's a good idea... >- -------------------------------------------------------------------------------- > -MTBF - 20,000 Hours Who are they trying to fool? Even if that's power-on hours, the heads are only rated for 1,500 hours - and that's for consumer-grade video. Just because you can watch the movie through the snow doesn't mean your system can read backups through it. 8mm technology may be better than previous video technologies, but it is still a consumer-grade system. Fast-forwarding or rewinding introduces an *incredible* amount of stress on both heads and tape. The heads are designed to 'win' by permanently deforming the tape wherever the tape is started/stopped. And guess what's there when you're doing backups- you gott it - the backup header! This is but one of the problems of the system, which will generally show up after about 500 hours of tape motion. How often has *your* transport been back to the factory? Think about it.  ================================================================================ Note 347.0 LQP02 Form length? 1 reply EISNER::WOOD_J "Allan Wood, STS Consultants, Ltd." 16 lines 12-APR-1989 19:01 -------------------------------------------------------------------------------- I need to adjust the form-feed length of an LQP02 printer. We are using it to print checks from an AP/financial package (Collier-Jackson) and want to be able to feed up the last check printed without losing the alignment of the checks. The software adjusts the form feed lenght when a check is printed, but not for the case of pressing the form feed switch. The manual we have doesn't address the problem; only reference is to form feed switch which "when pressed ... the paper moves forward a predetermined number of lines." It doesn't say how to predetermine them! If anyone can tell me how to do this and/or point to me some source that would do the same thing, I would appreciate it very much. ================================================================================ Note 347.1 LQP02 Form length? 1 of 1 EISNER::KOZAM 5 lines 12-APR-1989 19:18 -< Setting lines per page >- -------------------------------------------------------------------------------- From the 1985 Terminals and Printers Handbook: The appropriate escape sequence is: ESC [ Pn t Pn is the number of lines per page (must be between 1 and 169).  ================================================================================ Note 348.0 Unibus controllers for Fujitsu M2361A & Cipher M990? 2 replies EISNER::BURNET "John Burnet" 38 lines 13-APR-1989 22:20 -------------------------------------------------------------------------------- Here are a few questions that I hope one or more of you hardware wizards who have experience with third-party controllers and peripherals can answer. We just bought (used, at auction) a MicroVAX II that happens to have a Fujitsu M2361A disk and a Cipher M990 tape drive, controlled by an Emulex QD32 and a TC03, respectively. (Of course we got almost no documentation on the peripherals or controllers, the hardware is all long out of warranty, and we have no hardware or software support yet.) The MicroVAX will be deployed elsewhere, but we want to keep the mass-storage goodies around to use on an 11/750 and/or an 8350. I have read the comments on Emulex, Fujitsu, and Cipher stuff in this conference. The questions that I have are: 1. What's the colloquial name for the Fujitsu disk? The M2351 is the Eagle; is the 2361 the next generation? With the Emulex controller, this disk looks like two RA81s to the VAX. 2. Assuming that we'll be trashing the Emulex Q-bus controllers, and assuming that we'll steer well clear of Emulex in the future (after reading all of the war stories here): What Unibus controllers can we choose from in order to hang the Cipher and the Fujitsu on the 750? The 750 currently has a UDA50 and a TS11 (TU80) controller, as well as an MBA (from Harvard -- oops, scratch that). Either DEC or third-party controllers would be okay. 3. The tape drive will have to go on the 750, but the disk can alternatively (and preferably) go on the 8350... is there any VAXBI controller that we can use? (Not likely, right?) 4. The TC03/M990 combination uses an Emulex driver. We have the driver (binary) for V4 but not for V5. If we use some (unspecified) Unibus controller, what driver (non)availability problems will we have to contend with? Is there any way to make the M990/any-controller combination look like some random DEC 6250bpi tape drive so that we can use DEC's TUDRIVER in the future? I guess that uses up my question quota for today. Thanks in advance for any and all info. ================================================================================ Note 348.1 Unibus controllers for Fujitsu M2361A & Cipher M990? 1 of 2 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 46 lines 14-APR-1989 00:24 -< its still an Eagle >- -------------------------------------------------------------------------------- The disk is the Eagle+ or Super Eagle, or whatever. We, too, had one on a MVII with the QD32, but the QD32 couldn't hack faster drives (Sabre 3.0Mbytes/sec), and we wanted Q/U compatible formatting (for 9766s dual ported on the same controllers) so we put in Dilog DQ256 on Q and DU256 on U (each does 4 physical SMD drives). The Eagle+ had been set up as 1 drive on the QD32, and could be 1-4 logical drives on the Dilogs. It is now DU0: + DU1: on a U bus 11. If we ever add the dual port card, we will be able to backup the 11s disk using BAC on the VAX. You need the Emulex Diag tape to do much with that QD32. The Dilogs setup, etc is built in to the controller. If you were using the Tape on Q, the Dilog DQ153 is a super choice, and it is less $ that a TS11 emulating Dq142. It is TMSCP (TU:). On U, they don't have TMSCP emulation yet as far as I know. The TU80 controller you have is a single TS11 emulation on a DEC produced Dilog DU132. If you got the Dilog proms, 4 drives daisy chain off those 2 Pertec cables (each with a seperate CSR + VEC - TS11s were single drives). If the TU80 controller is there WITHOUT the TU80, simply add your drive. If you are getting a NEW tape controller, see how high a # Dilog has gotten to. There may be a DU142. You can ask if there is a DU153 yet, but I doubt it. 1-800-DILOG-32 (I have no connection). As for the BI machine - If every one would STOP buying machines they can't plug drives of their choice into, DEC might get the message. For you NEXT $10,000,000. CPU order with DEC, SPECIFY as a prerequisite the ability to use controllers and drives of any source, and DON'T give DEC the order if they fail. After a few hundred of such lost orders, even DEC would catch on. The folks that make the adapters to be able to use any SMD drive in a UDA50, KDA50, KDB50, HSCx all want to sell you the drive too, and after looking at their prices, DEC isn't the only bandit in town. We probably have manuals on the Eagle+, and QD32. If only you were local, you could drop by. If there is something specific I could look up, ask. ================================================================================ Note 348.2 Unibus controllers for Fujitsu M2361A & Cipher M990? 2 of 2 EISNER::BURNET "John Burnet" 18 lines 15-APR-1989 18:52 -< Looks like Eagle+ --> U and M990 --> Q for us >- -------------------------------------------------------------------------------- Thanks. It looks like we'll be putting the Eagle+ on the 750. The M990 will go on another MicroVAX (one that has an RA81 and can take advantage of the high-performance tape drive for backups), since apparently we can't use it on the Unibus or BI systems. Your point about closed vs. open bus architectures is well taken, and I myself couldn't agree more. Unfortunately, it's not always easy to get the people who are purchasing the systems to see things that way! Also, in our case, it would be a quixotic gesture, hurting only ourselves, since we spend much less than the $500,000 per year that DEC seems to consider the minimum acceptable hardware purchase before they'll admit that a customer exists. (Actually, what we need to do is to drop DEC and move to Sun, but that's another story...) Another factor is performance. Considering *only* DEC's stand on VAXBI, a boycott makes perfect sense; but thinking technically, what are the alternatives to the 13.3 MB/s BI bus? Unibus? Slow and almost dead. Q-bus? Almost as slow. If DEC would only adopt VME, we'd be in great shape.  ================================================================================ Note 349.0 MV2 power fail trashes disk 7 replies EISNER::WICHMAN "Jon Wichman" 85 lines 14-APR-1989 12:10 -------------------------------------------------------------------------------- A MicroVax system that I manage on an as needed basis for another division has had disk drive problems ever since I started managing it about 9 months ago (just coincidence, I'm sure - they must've had them before and didn't know about it). These are the dreaded "forced error flags". The drives involved are an RD52 and a RD53. The system configuration is as follows: MicroVax II ! Upgraded MV1 in BA23 cabinet Internal RX50 Internal RD52 External TK50 External RD53 RQDX2 RQDXE DEQNA DZQ11 MS630-?? ! 2 meg. board MS630-BB ! 4 meg. board They had been getting seemly random forced errors at a rate of about one every three weeks (at least that's how often it hits a valuable file that has to be restored). During one particulary bad spate about 6 months ago (when I was installing V5), DUA1: (the RD53) became unrecognizable to the system and after a long, drawn-out service call, DEC ended up replacing the RD53, the RQDX2, and the RD52. Things were fairly stable for a while although they were still getting occasional forced errors until a week ago Tuesday (April 4). After the three-day weekend (ain't NATO day wonderful!), the system wouldn't reboot after a power failure. This was found to be caused by a bad/missing DUDRIVER.EXE. I restored the system disk from a image backup. This past Monday, I got another call because the system only came up partially. I arrived and found out that "SYS$LIBRARY:SCRSHR.EXE" was missing. I restored the drive again and was talking to the customer when she remarked that the system always gets screwed up after a power failure. I'd heard that they had power problems (mainly, the power goes out during practically any thunderstorm), but had never connected it to the disk errors before. With their permission, I killed the power, waited a few seconds, and turned it back on. Bingo! Two consecutive files in SYS$SYSTEM were trashed - "SHOW.EXE" and "SHUTDOWN.COM" (a nice combination). I copied those files from another V5.02 system on the network, deleted the old files, and repeated the experiment (BTW, anyone know the story behind the mysterious BADBLOCK_SCAN process that runs after you delete a file with forced errors?). This time, two different files were trashed - "ERFTAPE.EXE" and "ERF7XXX.EXE", again two files adjacent to each other on the disk (since it'd just been restored). DEC FS was called. The rep said that he thought he'd heard of this problem, would do some research on it, and then come out. He came out the next AM and checked the rev. levels on all of the modules. He said that there was a known problem with certain rev. levels of RQDXE's, but that ours was above the required rev. level. He said that he'd replace the board anyway the next day (they have BASIC service). Unfortunately, at that time, the customer was getting hot and got her DEC sales rep on the phone. He said that he'd call FS and find out what their "Action Plan" was. So, since that time, we haven't heard back from DEC. I assume they're busy writing up a action plan instead of out there working on the system (or at least - board-swapping). In the meanwhile, I've had a batch job running $ANAL/DISK/READ continuously on both disks. In almost 2 days, DUA0: (RD52) has logged 11 errors and DUA1: (RD53) has logged 39 errors. $ANAL/ERROR/SUMM shows 11 (DUA0) and 41 (DUA1) UCB's. The /FULL log indicates these are all DATA ERRORS, uncorrectable ECC errors, and has done a block replacement on them (that's what all these errors always show up as). Any ideas? Anyone ever heard of this? Short of a UPS (suggested, but she doesn't have the budget) - I don't have any ideas except for DEC FS finding something sooner or later. If you need more specific information, error logs, etc. let me know. The machine's about two blocks away and is on DECnet, so I can get back with the info fairly rapidly. P.S. I just located notes 11.9 and 11.19 which sound like relevant ECO's. Perhaps I'll double check on DEC now... ================================================================================ Note 349.1 MV2 power fail trashes disk 1 of 7 EISNER::KENNEDY "Terry Kennedy" 18 lines 14-APR-1989 19:26 -< Here is what to do to fix it >- -------------------------------------------------------------------------------- 1) Don't believe FS (in this instance) 2) You need: a) RQDXE replaced with the newer 1E variety - it's a different module number b) The ribbon cable between the RQDX2 and the RQDXE replaced with the shielded kind - there is an FCO for this c) The ribbon cable between the RQDXE and the BA23 disk signal distribution board replaced with the shielded kind. *THIS IS ALSO NEEDED BUT THE FCO FOR B) DOES NOT INDICATE IT* d) The RD53 would be better in the BA box and the RD52 in the tabletop, but this is not required. 3) If you do 2A thru 2C, your problem *will* go away... ================================================================================ Note 349.2 MV2 power fail trashes disk 2 of 7 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 3 lines 16-APR-1989 21:03 -------------------------------------------------------------------------------- > 3) If you do 2A thru 2C, your problem *will* go away... And, if your problem doesn't go away, call Terry - he said it would! ================================================================================ Note 349.3 MV2 power fail trashes disk 3 of 7 EISNER::KENNEDY "Terry Kennedy" 2 lines 17-APR-1989 02:05 -< That's right - see my WHO_AM_I entry for # >- -------------------------------------------------------------------------------- > And, if your problem doesn't go away, call Terry - he said it would! ================================================================================ Note 349.4 MV2 power fail trashes disk 4 of 7 EISNER::WICHMAN "Jon Wichman" 11 lines 17-APR-1989 11:25 -< Ready to do battle with DEC >- -------------------------------------------------------------------------------- ******* 1) Don't believe FS (in this instance) ******* I seldom do! Last time I believed them was in 1982 - right before FS replaced a memory board BACKWARDS in my 11/70. ("Hey, this a PDP, and this guy's worked on 11/23's before - let's send him out!") Nice craters in several of the chips and a very bad burning smell were among the more obvious results of this... Thanks for the info - I feel better going into battle with a little armament with me! ================================================================================ Note 349.5 MV2 power fail trashes disk 5 of 7 EISNER::GRANZOW 9 lines 17-APR-1989 12:21 -------------------------------------------------------------------------------- < Note 349.4 by EISNER::WICHMAN "Jon Wichman" > -< Ready to do battle with DEC >- ******* 1) Don't believe FS (in this instance) ******* DON'T FORGET THAT THERE IS ALSO AN FCO FOR THE POWER CABLES ON A BA23 BOX. THIS CAN ALSO CAUSE THE TYPE OF PROBLEMS YOU ARE SEEING. (CABLES BETWEEN THE SUPPLY AND BACKPLANE.) ================================================================================ Note 349.6 MV2 power fail trashes disk 6 of 7 EISNER::KENNEDY "Terry Kennedy" 12 lines 17-APR-1989 18:11 -< Good to have; probably not the culprit >- -------------------------------------------------------------------------------- > DON'T FORGET THAT THERE IS ALSO AN FCO FOR THE POWER CABLES ON A BA23 > BOX. THIS CAN ALSO CAUSE THE TYPE OF PROBLEMS YOU ARE SEEING. > (CABLES BETWEEN THE SUPPLY AND BACKPLANE.) While it is true that the FCO is a 'good thing' to have (I lost a system to it), I don't think it's the problem in this case. Bad power cables would cause random problems, or problems at every power-up, not just after power failures. Note that the original *replacement* part also will burn up. I just gave up and used Teflon-jacked nickel/silver wire brazed onto the power supply lugs. See my posting in HARDWARE as well as "Burned up over BA23 fire" in SOAPBOX. ================================================================================ Note 349.7 MV2 power fail trashes disk 7 of 7 EISNER::WICHMAN "Jon Wichman" 26 lines 19-APR-1989 09:56 -< And the correct answer is "1" >- -------------------------------------------------------------------------------- !!!! 1) Don't believe FS (in this instance) Boy, you sure called this one Terry! When FS rep. #2 came out yesterday to replace the RDXQE which was up to rev. level - just in case it was bad, he found that it WASN'T up to rev.!! To refresh everyone's memory, FS #1 supposedly checked the board and verified the rev. level with Atlanta last week. I guess he was looking at his watch or something instead of the board and got the numbers confused?:-) Anyway, he installed the new (correct rev.) board and replaced the power cable mentioned in (0.+2?). He also re-formatted the disk (20 errors). The customer whipped out her copy of the replies to .0 and showed it to FS #2 upon arrival. He was rather quiet and subdued during the rest of the call. Because the customer needed the system back up ASAP due to a deadline, we decided to go try it with the above two changes and see what happened. If the problem continues, I'll call them back and push for the other changes and, of course, let y'all know about it too. If things go true to form, it shouldn't be more than a couple of weeks before we know one way or the other...  ================================================================================ Note 350.0 Weird DF242 modem problem 3 replies EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 31 lines 19-APR-1989 12:19 -------------------------------------------------------------------------------- I ran across an interesting problem with a DF242 modem last week and am wondering if anyone else has seen it. The modem is on a remote system in Texas, so I can't get to it to examine the switches, etc. The problem is a cycle of three repeating steps: 1) I dial into the remote system, get a good connection, do my work, log out and hangup from my end. The remote DF242 is on a DL11-type line (system console) and therefore has no modem control lines in use. So far, so good. By the way, this is all at 1200 bps. 2) I dial in a second time, get a connection (MCR or console ODT prompts), but the remote modem seems to be sending the remote system nothing but a string of upper case "A", followed by a Line-Feed. I get all the echoes from the remote system OK. Needless to say, the remote system doesn't know what to make of these strange command lines! Nothing I do at my end seems to have any effect. I then hang up my end. 3) I dial in a third time, but the phone is never answered. If someone cycles power on the remote modem, we go back to step 1. Any ideas? The modem at this end is a Vadic 2400-VP. I do not have a DF242 to play with here, nor do I have a manual for it. The people at the remote system are somewhat less than helpful, as they know little about the DF242 and are not very interested in learning. ================================================================================ Note 350.1 Weird DF242 modem problem 1 of 3 EISNER::MAYHEW "Bill Mayhew" 14 lines 19-APR-1989 14:22 -< Sounds like a huh-what? loop to me... >- -------------------------------------------------------------------------------- I have not touched a DF242, but I have seen similar behavior on a DF224. The "a" is the modem's "attached" message, much like the "CONNECT 1200" that a Hayes-type modem gives. What I think is happening is that the modem is uttering that message to the computer, which is echoing it back to the modem, which is echoing it back to the computer, ... I would try turning off the "command character echo", or whatever it's called, on the DF242, and also put it in "quiet" mode if possible (so it doesn't send the first "a" to begin with). I don't have DF242 docs here, so can't provide more specific info. ================================================================================ Note 350.2 Weird DF242 modem problem 2 of 3 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 84 lines 19-APR-1989 14:38 -< DEC'S SETTINGS FOR THE DF242 >- -------------------------------------------------------------------------------- Operational : 1 Modem protocol : (1200) Bell 212 =0 (300 ) Bell 103 =1 (2400) V.22bis =2 (1200) V.22 =3 [2] 2 Character echo : Off=0, On=1 [0] 3 Answer mode : Auto=0, Man=1 [0] 4 Attached msg on incoming call : Off=0, On=1 [0] 5 Delete key effect : CRT=0, Hardcopy=1 [0] 6 Enable Rem Dig Loopback : Off=0, On=1 [0] 7 Blind Dialing : Off=0, On=1 [0] 8 Reserved For Digital Use 9 Reserved For Digital Use 10 Reserved For Digital Use 11 Automatic Retrain : Off=0, On=1 [0] 12 Speaker enable : Off=0, On=1 , On if no CD=2 [0] 13 Speaker volume : Low=1, Med=2, Hi=3 [1] 14 Ring to Answer on Ring = 1 to 10 [2] 15 Callback Timer Time = 1 to 6 [1] Communication : 1 Data Bits : 7 bits=3 8 bits=4 [4] 2 Parity : None=1, Even=2, Odd=3 [1] 3 Stop bits : One=1, Two=2 [1] 4 Off-line XON/XOFF recognition Off=0, On=1 [0] 5 Communication Protocol : Async=0, Sync=1 [0] 6 Sync mode timing : Int=0, Ext=1, Slave=2 [0] 7 DSR forced : On=1, Normal=2 [2] 8 EIA CD/CTS Signals Separate = 0 Common = 1 [0] Disconnect : 1 Send Space Disconnect Off=0, On=1 [0] 2 Receive Space Disconnect Off=0, On=1 [0] 3 Carrier Loss Disconnect Off=0, On=1 [1] 4 Current Loss Disconnect Off=0, On=1 [1] Error Correction : 1 Error Correction : Off=0, MNP=1, X.PC=2 [1] 2 Auto Fallback : Off=0, On=1 [1] 3 Transmit Break Reaction : Disregard=0 Destructive=1 Non-Destructive Immediate=2 Non-Destructive Sequenced=3 [3] 4 Receive Break Reaction : Disregard=0 Destructive=1 Non-Destructive Immediate=2 Non-Destructive Sequenced=3 [3] 5 Reserved For Digital Use 6 MNP Disconnect Type : Non-disruptive=0 Disruptive=1 [0] Speed Buffering : 1 Speed Buffering State : Off=0, On=1 [1] 2 DTE Flow Control : None=0 X-ON/X-OFF=1 RTS/CTS=2 Both=3 [1] 3 Autobaud state : Off=0, On=1 [0] 4 Local baud rate : 1200 bps=0 2400 bps=1 4800 bps=2 9600 bps=3 [3] Access Security : 1 Callback Memory Access : Off=0, On=1 [0] 2 Dial Memory Access : Off=0, On=1 [0] 3 Parameter Access : Off=0, On=1 [0] Statistical : 1 Bit Error Rate : 10E6 2 Line Level : 45 ================================================================================ Note 350.3 Weird DF242 modem problem 3 of 3 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 3 lines 20-APR-1989 13:27 -< I suspect you're right, but not the entire answer >- -------------------------------------------------------------------------------- >> -< Sounds like a huh-what? loop to me... >- But why the 3-step cycle of working, sorta-working, no-answer?  ================================================================================ Note 351.0 EXABYTE /BACKUP 1 reply EISNER::GRANZOW 5 lines 19-APR-1989 19:52 -------------------------------------------------------------------------------- I HAVE BEEN SEARCHING THE BASE AND CAN NOT FIND ANSWERS TO TWO QUESTIONS I HAVE. 1. DOES ANYONE HAVE RELIABILITY EXPERIENCE OR FIGURES ON EXABYTE, AND 2. WHERE DOES ONE GET THE UNIT REPAIRED AFTER WARRANTY RUNS OUT? THANKS IN ADAVANCE ================================================================================ Note 351.1 EXABYTE /BACKUP 1 of 1 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 11 lines 19-APR-1989 20:51 -< It's there >- -------------------------------------------------------------------------------- RE: < Note 351.0 by EISNER::GRANZOW > -< EXABYTE /BACKUP >- $ search decuserve_all_titles exabyte shows notes in topic 13 in this conference $search decuserve_all_titles 8mm shows notes in 13, 320, 335, and 346 (although the latter has no replies).  ================================================================================ Note 352.0 MSCP CONTROLLER PRIORITY 2 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 12 lines 21-APR-1989 19:40 -------------------------------------------------------------------------------- The UDA-50 manual (EK-UDA50-UG-002), page 2-13, says: "All UDA50 modules are shipped with a level 5 priority plug. This is the recommended priority level for UDA50 Disk Subsystems and thus, the prio- ority plug need not be changed for the majority of installations." Both Sigma and TD systems ship their MSCP controllers at priority 4 and claim that is the norm for VMS. In TD's case they can't change it. RSTS requires they run at priority 5. What does VMS really require? What does VMS really recommend? What priority are you running your third party MSCP controllers at? ================================================================================ Note 352.1 MSCP CONTROLLER PRIORITY 1 of 2 EISNER::KENNEDY "Terry Kennedy" 14 lines 23-APR-1989 20:47 -< The Q stuff is level 4, U is 5, clear as mud? >- -------------------------------------------------------------------------------- > Both Sigma and TD systems ship their MSCP controllers at priority 4 and > claim that is the norm for VMS. In TD's case they can't change it. RSTS > requires they run at priority 5. What does VMS really require? What > does VMS really recommend? What priority are you running your third > party MSCP controllers at? Bothe the RQDX2 and RQDX3 are fixed at 4 and are not changable (EK-RQDX2- UG-001, P. 3-4; EK-RQDX3-UG-001, P. 2-9) So, it seems that is depends on the controller; although it may be a Q/U issue. The VMS V5 manual set is unresponsive; the only reference to 'Bus', 'BR level', or 'BIRQ' is in the Device Support manual, SYSGEN is mum. The RSTS manaul claims to want BR5 (apparently, only if it can get it, which it can't on RQDXen.) ================================================================================ Note 352.2 MSCP CONTROLLER PRIORITY 2 of 2 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 7 lines 24-APR-1989 14:49 -< According to one of the designers... >- -------------------------------------------------------------------------------- >> What priority are you running your third party MSCP controllers at? Andromeda Systems ships all of their controllers at BR4, but you can jumper them to either 5 or 6. It is certainly better to run the controller at 4 when the software is expecting 5 than the other way around.  ================================================================================ Note 353.0 Large Capacity Paper Tray for Talaris 802? 2 replies EISNER::WOOD "Jim Wood (609)423-1040" 2 lines 25-APR-1989 09:21 -------------------------------------------------------------------------------- Does anyone know of a vendor who sells a large capacity paper input tray for the Talaris 802? (Talaris doesn't) ================================================================================ Note 353.1 Large Capacity Paper Tray for Talaris 802? 1 of 2 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 16 lines 25-APR-1989 21:33 -< try QMS's Laser Connection >- -------------------------------------------------------------------------------- >Does anyone know of a vendor who sells a large capacity paper input >tray for the Talaris 802? (Talaris doesn't) I don't know which engine is in there, but if that is a typical desktop rig with a Ricoh or Cannon engine, there should be a feeder to fit. The Laser Connection folk's Source Book (they are the retail end of QMS for low end 8 ppm or less lasers) has 2 pages of feeders, but I don't see your model mentioned. They are at 1-800-523-2696 for VERY GOOD product and application help, and nearest dealer info. If you ask nicely, they may be able to suggest which dealers are known for BIG discounts. It DOES make a big difference... (I have NO connection...) ================================================================================ Note 353.2 Large Capacity Paper Tray for Talaris 802? 2 of 2 EISNER::WOOD "Jim Wood (609)423-1040" 8 lines 26-APR-1989 16:18 -< No Luck >- -------------------------------------------------------------------------------- < try QMS's Laser Connection > I did, thanks to your suggestion. Unfortunately, they don't have a large capacity feeder for the Talaris 802. As you said, though, they were very helpful and a pleasure to talk to. Thanks for the info.  ================================================================================ Note 354.0 "VAX 8900"? 2 replies EISNER::HASSINGER "Bob Hassinger" 2 lines 25-APR-1989 11:47 -------------------------------------------------------------------------------- Note reference in INDUSTRY_NEWS 301.2 and .3 to "VAX 8900". Have I missed something? ================================================================================ Note 354.1 "VAX 8900"? 1 of 2 EISNER::MAYHEW "Bill Mayhew" 5 lines 25-APR-1989 16:13 -< 8974 or 8978? >- -------------------------------------------------------------------------------- I assume this refers to these machines, which were announced at about the same time as the 8700/8800 and are essentially marketing-created "systems" consisting of 4 or 8 clustered 8800s (or is it 8700s?) ================================================================================ Note 354.2 "VAX 8900"? 2 of 2 EISNER::SHANNON "Terry C. Shannon, IDC" 3 lines 25-APR-1989 20:04 -< A Marketing Ploy From The Past. >- -------------------------------------------------------------------------------- VAX 8900 does in fact refer to the VAX 8974 and 8978, which are preconfigured, clustered VAX 8800 systems. They were introduced slightly more than two years ago.  ================================================================================ Note 355.0 MTI/Maxtor disk drives 5 replies EISNER::BRUCKER "Cyberspace Jockey" 10 lines 28-APR-1989 14:17 -------------------------------------------------------------------------------- Has anyone had any experience with MTI disks connected to an HSC? We are looking at purchasing the MTI equiv to the RA90. The disk is the MDI-476 (actually 4 Maxtor 8760s seen as 2 logical units) and has a formatted capacity of 2.2GB. Pros? Cons? Thanks! -Ken ================================================================================ Note 355.1 MTI/Maxtor disk drives 1 of 5 EISNER::RICE "Gary Rice" 22 lines 29-APR-1989 09:53 -------------------------------------------------------------------------------- > Has anyone had any experience with MTI disks connected to an HSC? PROs: Inexpensive when compared to comparably sized drives CONs: Slow transfer rate Rumor (not confirmed by experience) that under certain conditions associated with losing a spindle of the 4 drive set, the controller writes garbage to the disk but doesn't inform VMS that anything is wrong. Company just emerging from Chapter 11 Also on a personal level, I have come to dislike the Sales Rep that we dealt with during a recent competitive bid evaluation between MTI, Emulex, SI and Dilog. By the way, Dilog was my vendor of choice between these RA90 clone vendors. Gary ================================================================================ Note 355.2 MTI/Maxtor disk drives 2 of 5 EISNER::BRUCKER "Cyberspace Jockey" 16 lines 3-MAY-1989 12:36 -< more MTI questions >- -------------------------------------------------------------------------------- > < Note 355.1 by EISNER::RICE "Gary Rice" > >CONs: Slow transfer rate Do you have any numbers to back this up. From indications I have the RA90s have a transfer rate of 2.8 MB/sec. The Maxtors are 1.9MB/Sec/Spindle and since 2 spindles are in use, and the data is interleaved between the two spindles according to MTI, you get an aggregate throughput of 3.8MB/sec. When you refer to the 4 drive set, is that when all 4 spindles are connected to the HSC via a single SDI port, or when they are paired as two RA90 equiv using two ports? - Ken ================================================================================ Note 355.3 MTI/Maxtor disk drives 3 of 5 EISNER::RICE "Gary Rice" 16 lines 3-MAY-1989 14:53 -------------------------------------------------------------------------------- > Do you have any numbers to back this up. From indications I have > the RA90s have a transfer rate of 2.8 MB/sec. The Maxtors are > 1.9MB/Sec/Spindle and since 2 spindles are in use, and the data > is interleaved between the two spindles according to MTI, you get > an aggregate throughput of 3.8MB/sec. Interesting. The sales "glossies" and discussions with the engineers at the plant yielded a transfer rate of 2.5 mb for Maxtor configuration. Furthermore, they told us that the drives are not really intreleaved, rather, they are daisy-chained. When you refer to the 4 drive set, is that when all 4 spindles are connected to the HSC via a single SDI port, or when they are paired as two RA90 equiv using two ports? All 4 into 1 port. ================================================================================ Note 355.5 MTI/Maxtor disk drives 5 of 5 EISNER::BRUCKER "Cyberspace Jockey" 24 lines 3-MAY-1989 18:33 -< MTI info from the source >- -------------------------------------------------------------------------------- < Note 355.4 by EISNER::BRUCKER "Cyberspace Jockey" > -< MTI misinformation? >- >Interesting. The sales "glossies" and discussions with the engineers at the >plant yielded a transfer rate of 2.5 mb for Maxtor configuration. Furthermore, >they told us that the drives are not really intreleaved, rather, they are >daisy-chained. Just got off the phone with my local MTI rep and got a little better picture of what is going on. First off, some of the MDI476 units use an Hitachi drive and others use the Maxtor drive. The Maxtor drive has a transfer rate of 1.9mb/sec. The data would seem to be laid down on disk in lbn order. lbn 0 is block 0 of drive 0, lbn 1 block 1 of drive 0, etc until drive 0 is filled. Then LBN ~1,100,000 starts on block 0 of drive 1. The sales rep is checking with the factory to verify this information. My rep said the faster transfer rate (2.5mb) is a customer requested option. One implication of the data layout is that the indexf.sys file will end up at the end of drive 0 and beginning of drive 1, assuming a 2 disk configuration. This is also being verified by the rep. -Ken  ================================================================================ Note 356.0 RA90 microcode upgrade for HSC V3.9 No replies EISNER::MOZDZEN "Don Mozdzen" 73 lines 1-MAY-1989 13:10 -------------------------------------------------------------------------------- We upgraded our RA90s a couple of months ago, and since I haven't seen this posted anywhere else, I thought it might be of interest. When we received the latest HSC software upgrade, V3.9, the following note was found in the Release Notes at the beginning of the software installation section. NOTE If your HSC cluster has RA90 disk drives connected to it, verify that the RA90 drives report a minimum software revision level of MC = 7. Issue the SHOW DISKS command to perform the verification. If the drives are not at that minimum software revision, ask your DIGITAL Field Service Representative to install FCO RA90X-O001 prior to installing HSC Version 3.90 software. Since we received an early shipment of RA90s and a SHOW DISKS reported MC - 5, we contacted Field Service to upgrade our drives. After a lot of phone calls, our engineer came up with the following description for the FCO. To: Field Service Engineers SUBJECT: SA600/RA90 MICROCODE CHANGES PROBLEM: CSSE has observed in some customer applications, SA600/RA90 drives exhibiting the following symptoms: o Large numbers of Uncorrectable ECC errors, and 6-8 symbols in error. o Occurs predominately on the first 5 heads. o Occurs predominately from cylinders 1900 thru 2649. o BBR was being invoked and the blocks were being replaced about 80% of the time. o Cause of data errors was not related to head/media/data path problems, but "fine position" servo control, which is controlled by the drive microcode. o Data loss due to writing data off-track and unable to recover data thru error recovery. This is significant risk if subsystems is invoking drive error recover levels 7(.) thru 12(.) in order to recover the customer data. The customer applications/operations that are prone to show this syndrome include: o Image Backups, in particular when many "short" length files are involved. o Use of 'third party' defragmenters where variable length files are involved. o Those applications that are able to sustain high level I/O requests involving short block transfers on each I/O. After reading this, we had our Field Service engineer order the microcode update cartridge and scheduled the disks to be upgraded. This was done over a weekend, since the drives had to be formatted and verified, and our disk now are at MC - 8.  ================================================================================ Note 357.0 UNIBUS loading 11 replies EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 17 lines 3-MAY-1989 09:07 -------------------------------------------------------------------------------- Okay all you old timer (and I mean that with the utmost respect), how do you know when you have too many devices on a UNIBUS (other than the fact that things don't seem to work) ? Case in point. We have an 8650 with 1 UNIBUS (two boxes hooked together) that has the following : 1 DMF32 6 DHU11 5 DMR11 1 DEUNA 1 Third party tape controller This seems like a bit much to me, but how do you convince people otherwise ? ================================================================================ Note 357.1 UNIBUS loading 1 of 11 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 62 lines 3-MAY-1989 13:27 -< Watch your POWER >- -------------------------------------------------------------------------------- > We have an 8650 with 1 UNIBUS (two boxes hooked together) that has > the following : > 1 DMF32 > 6 DHU11 > 5 DMR11 > 1 DEUNA > 1 Third party tape controller > This seems like a bit much to me, but how do you convince people > otherwise ? Which cards are where in which backplanes in which kinds of boxes with which regulators, and what # Unibus jumper is between adjacent backplanes? Without looking anything up, a quick answer is that I would GUESS you are 'ok' on bus loading, but you have some power PIGS there. I will also guess that you have 2 BA11KE or BA11KU boxes. Not only do you have to watch the power in each box, but you have to watch the power on a single DD11DK backplane, and on each regulator. Also DMRs may have problems if 2 are stuffed into adjacent slots, and I seem to recall that there is a prefered sequence for the DMR's CPU and LINE card to be stuffed into the backplane. In any case, 2 DMRs (thats 4 cards) is all you want in a DD11DK (and separate the 2 pairs by a slot or more, and preferably have each pair fully on the other side of center from the other), and at that point you are out of steam if your regulator is an H744. BA11KEs normally had 24 amp H744s, and BA11KUs had the 32 amp H7441s, but look, as folks swap them around. I have seen F/S split a DMR between backplanes, and with each card on a different H7441. That is a poor thing to do, especially since used BA11KUs (with the desirable 32 amp H7441s) are VERY inexpensive. You can probably get 2 for well less than $500. The DEUNA is also a POWER PIG. I have no idea about DHUs and DMF32s or whatever 3rd party controller you have. I would guess the average 3rd party tape controller is about 4.5 amps. That 5th System Unit on a BA11KE/U that you could stuff a DD11CK into, is best left open, as it shares the 2nd regulator that is also needed for the DD11DK (normally) in SUs 3+4. If you add a 3rd box, the U bus cable to it should NOT be the same length as that between the first 2 boxes. There are probably a few other rules/ideas that someone else will remember, but I bet this is the general area to watch. If, for some reason, you are going to make more of these, and terminal servers are not an option, consider an Emulex CS11 (with DMF32 emulation) that will do 64 ASYNC (full modem control) ports from a single Unibus card drawing 3.4 amps, or the CS32 controller that will do 128 ports. Both will need a CP34 panel for each 16 ports (and that is where the $s are), but those panels will also work on a Q bus CS04 controller that does 64 ports, so when you dump your Unibus machines, your $s aren't all lost. The Emulex panels can sometimes be found used cheap. I once found several of the older CP12 panels (will work up to 64 (ONLY) ports worth on CS11s and CS32s, BUT WON'T work on the Q bus CS04 - damn) for $67 each. Also use the DELUA rather than the DEUNA. ================================================================================ Note 357.2 UNIBUS loading 2 of 11 EISNER::HIBBITS 56 lines 3-MAY-1989 14:25 -< Some details >- -------------------------------------------------------------------------------- < Note 357.0 by EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" > -< UNIBUS loading >- > We have an 8650 with 1 UNIBUS (two boxes hooked together) > > This seems like a bit much to me, but how do you convince people > otherwise ? (Prepared while Barton responded, so somewhat redundant) (Figures from 1989 April-June VAX Systems and Options Catalog) dc Amps Drawn Bus Loads 5V 15V -15V ac dc Totals 5 15 -15 ac dc 1 DMF32 8.0 0.5 0.5 6.0 1.0 8.0 0.5 0.5 6.0 1.0 6 DHU11 6.0 0.4 0.4 2.2 1.0 36.0 2.4 2.4 13.2 6.0 5 DMR11 12.0 0.08 0.19 1.0 60.0 0.4 0.96 5.0 1 DEUNA 8.0 1.0 4.0 1.0 8.0 1.0 4.0 1.0 1 Third party tape controller ????????????????? 112 3.3 4.86 23.2 13.0 (Add your own nuts and toppings for the tape controller, the information should be with the board documentation.) Now for some suppositions. If I were to guess, you have 2 BA11 boxes, with six DD11 units in each box. The rules go something like this: Per box, there are 96 amps @5 volts, 5 amps @15 volts, and 6 amps @ 15 volts, provided that you don't exceed 32 amps @ 5 volts per adjacent pair of DD11 units. This is a board placement issue, I'd think. Furthermore, there are limits to the bus loads in each box, the numbers escape me, but if my rememory serves as to scale, if you break up the cards according to the rules above, you shouldn't bump your noggin on the bus load limit. If the suppositions are not correct, you could still easily be playing within the rules above, if the board placement is closely monitored. Of course, the other issue is, how much traffic is on the UNIBUS itself, and is performance a problem!!! Separating the devices onto multiple UNIBUSes (:-)) may help get over throughput/bandwidth problems. Also, placement of boards with respect to one another on the bus can give interesting interactions (knee bone connected to the neck bone?) Any way, happy hardwaring!!!!!! Jim ================================================================================ Note 357.3 UNIBUS loading 3 of 11 EISNER::HIBBITS 8 lines 3-MAY-1989 14:36 -< Caveat Ethernet >- -------------------------------------------------------------------------------- Forgot to let you know: my figures for the DEUNA are based on the published values for a DELUA. Barton is right about the DEUNA being a power hog, but even if you multiply by 2, you should be in good shape. Jim ================================================================================ Note 357.4 UNIBUS loading 4 of 11 EISNER::PROVOST 93 lines 3-MAY-1989 15:13 -< Unibus loading rules >- -------------------------------------------------------------------------------- Below are some Unibus configuration rules taken from the PDP-11 Bus Handbook, which is no longer available. What it comes down to is that if you have too many devices you will experience bus hangs, lost interrupts, etc. If you are close, you can sometimes improve the system by adding repeaters, adding cable, eliminating cable symetry, etc. Looks like your AC loading is a bit high. Separate the heavy loads into separate boxes, and make sure the cable connecting the boxes is a different length from the one connecting the first box to the Unibus adapter. Unibus Segment - A Unibus segment is a portion of a Unibus sys- tem between and including two terminators. A bus segment con- sists of a terminator, a 120-ohm transmission path (cable) with options having drivers and receivers, and another terminator (in that order). A single bus system has one bus segment. A multiple bus system contains more than one bus segment, usually separ- ated by bus repeaters or bus switches. For example, a DT03 bus switch contains bus repeaters. Lumped Load - A lumped load is a group of Unibus elements (other than cables or jumpers) that are interconnected via the Unibus (backplane wire or PC etch). The group is not a lumped load if it uses a cable to interconnect the elements, or if the ele- ments are separated by a bus repeater. Semi-Lumped Load - A semi-lumped load is a group of lumped loads interconnected by 91.44 cm (3 ft) or less of cable and not separated by a bus repeater. This cable may be one of the fol- lowing: Cable Length BC11-A-OF 15.24cm (0.5 ft) BC11-02 60.96cm (2 ft) BC11A-03 91.44cm (3 ft) M9202 60.96cm (2 ft) AC Bus Load - AC loading is the amount of capacitance a module presents to a bus signal line. AC loading is expressed in terms of ac loads. A unibus segment can support up to 20 ac loads. For more than 20 ac loads, an M9202 (jumper with cable) is used to separate a large load into two Unibus segments. A number is assigned to each module indicating its ac load. The number is derived using reflection techniques in the time domain. The ac load causes a reflection to occur on the bus line shortly after a signal is asserted. This reflection shows up on an oscillo- scope as a spike shortly after the asserting edge of the signal. The ac load rating is usually based on worst case reflections on BBSY, SSYN, and MSYN signal lines. DC Bus Load - DC bus loading is the amount of leakage current a module presents to a bus signal line. DC loading is expressed in terms of dc loads; one dc load is approximately 105uA. This rep- resents 89uA for the receiver plus 25uA for the transmitter (driver). One dc bus load can also be thought of as one bus driver and one bus receiver (or one transceiver). Most modules present 1 dc bus load to the Unibus. A Unibus seg- ment can support a maximum of 20 dc bus loads. This limit is set to maintain a sufficient noise margin. For more than 20 dc bus loads, a Unibus repeater option (DB11-A) or equivalent is used. (For example, a DT03 bus switch also contains a bus repeater.) The DB11-A imposes 1 dc bus load on the first Unibus section, but then it can drive an additional 19 dc bus loads and 50 feet of cable. DC loading is more important on SSYN than on data lines. Unibus Configuration Rules 1. Maximum cable length of a Unibus segment should not exceed 15.24m (50 ft). 2. Maximum dc loading should not exceed 20. 3. Maximum ac loading should not exceed 20 for any lumped load (unless the entire segment consists of one lumped load). 4. Different cable lengths should be selected when: a. a lumped load requires cables longer than 2.59m (8.5 ft) for its bus-in and bus-out connections. b. the sum of the ac loads in the two lumped loads connected by the cables exceeds 18. c. the sum of the ac loads in the two semi-lumped loads con- nected by the cables exceeds 36. The length of cables selected should differ by 1.52m (5 ft) or more with the longer cable connected to the semi-lumped load with the greater ac load if possible without increasing the total cable length more than 1.52m (5 ft). 5. Voltage margin tests should be completed for system accep- tance: a. when the system is originally configured, and b. when any Unibus element is added, deleted, or swapped (including a defective module or backplane). ================================================================================ Note 357.5 UNIBUS loading 5 of 11 EISNER::KENNEDY "Terry Kennedy" 9 lines 3-MAY-1989 18:03 -< Performance is in utilization; not just loading specs >- -------------------------------------------------------------------------------- > how do you know when you have too many devices on a UNIBUS (other > than the fact that things don't seem to work) ? One other thing to consider here is whether your CPU / UBA can handle all the traffic this configuration can generate. You can still have a legal con- figuration and get miserable performance. What is the line speed on the DMR's, and are they half or full duplex? Also, what is the average character throughput for inbound and (seperately) outbound data on the terminal inter- faces? ================================================================================ Note 357.6 UNIBUS loading 6 of 11 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 32 lines 3-MAY-1989 19:01 -< Most BA11Kx boxes NOT 96 amps >- -------------------------------------------------------------------------------- >Now for some suppositions. If I were to guess, you have 2 BA11 >boxes, with six DD11 units in each box. The rules go something >like this: >Per box, there are 96 amps @5 volts, 5 amps @15 volts, >and 6 amps @ 15 volts, provided that you don't exceed 32 amps >@ 5 volts per adjacent pair of DD11 units. Those #s must be for the NEW BA11 boxes that have 3 regulators, and space for 6 System Units (each DD11DK uses 2). My comments in .1 all assumed the slightly older BA11KE or BA11KU boxes that have 2 regulators, and 5 SU (2 DD11DK + 1 DD11CK - MAX). SUs 1+2 are fed by the first regulator, and SUs 3-5 by the second. These are very common/popular. In U bus land, slightly old == VERY inexpensive... If you must add a box, try for a used BA11KU @ less than $200. Maybe they will include DD11DKs for that price. Make sure you get 32 amp h7441 regulators. If you want an OLD OLD box that has 3 h744 regulators, there was the BA11F class box for 11/45s and 11/70s. I think it holds 9 SUs (without walking 50' to look). You then need DD11xF backplanes that have the longer power wire harnesses. There also was a nifty box the 10/20 folks had that had a cable tunnel stamped into the center of the top cover. It was otherwise BA11Kx shaped, but I don't know how many SUs or regulators it took. ** WHICH BOXES DO YOU HAVE? ** ================================================================================ Note 357.7 UNIBUS loading 7 of 11 EISNER::PROVOST 37 lines 4-MAY-1989 09:37 -< Replies to misplaced NOTE >- -------------------------------------------------------------------------------- I mistakenly entered my UNIBUS LOADING reply in stream 257 instead of 357. I then entered it here. However, there were some relevant replies to the other entry, so I copied them here. <<< EISNER::DUA0:[NOTES$LIBRARY]HARDWARE_HELP.NOTE;1 >>> -< HARDWARE_HELP >- ================================================================================ Note 257.6 DMZ-32 DMA problems and FCO information 6 of 8 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 8 lines 3-MAY-1989 17:04 -< OK for the theory, how about in practice? >- -------------------------------------------------------------------------------- >> 5. Voltage margin tests should be completed for system accep- >> tance: >> a. when the system is originally configured, and >> b. when any Unibus element is added, deleted, or swapped >> (including a defective module or backplane). How does one do this operation? Is is also necessary/desirable for the Q-Bus? <<< EISNER::DUA0:[NOTES$LIBRARY]HARDWARE_HELP.NOTE;1 >>> -< HARDWARE_HELP >- ================================================================================ Note 257.7 DMZ-32 DMA problems and FCO information 7 of 8 EISNER::HARVEY "Jack Harvey" 8 lines 4-MAY-1989 07:52 -< Margin Testing was Marginal >- -------------------------------------------------------------------------------- Tom's information is really old. The laws of physics haven't changed, so all the loading rules still apply, but voltage margin testing hardware disappeared long ago. The reason it disappeared is that it rarely ever did much good and cost quite a bit. And we never did have it for the Q-Bus equipment. ================================================================================ Note 357.8 UNIBUS loading 8 of 11 EISNER::PROVOST 24 lines 4-MAY-1989 09:38 -< Old information >- -------------------------------------------------------------------------------- Unibus margin tester long gone. Basically, it enabled the FS tech to vary the terminator voltage radically. This would blow all the weak tranceivers. Any system that would run while margined would probably not incur intermittents. I copied the information on loading verbatim from the "blue book", so the margin box was included. There are too many variables for theory to predict what will and will not work. I have found that understanding the theory enables me to intelligently select configurations which have a higher probability of functioning. Adding Unibus is contra-intuitive, for example. Qbus systems usually lack significant cable. All devices appear as a single lumped load. Systems are usually not as complex as Unibus configurations, which can have cable interconnecting boxes and running to external peripheral controllers. Information is old, but theory is valid. The only more recent publication I know of is less complete and also out of print. ================================================================================ Note 357.9 UNIBUS loading 9 of 11 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 14 lines 4-MAY-1989 11:17 -< Traffic on the UNIBUS >- -------------------------------------------------------------------------------- Thanks for all the good info, folks ! Actually, I think we are in good shape with AC and DC bus loads as our FE is an old hand at these things (right Stu?) !!! Actually what I was referring to is the "traffic" on the UNIBUS. All the async ports are running 9600 and some of them carry lots of traffic (graphics terminals). 1 of the DMR's is not connected. 3 of the DMR's are running 9600 half dup. The fifth one will be running 56/64 K full dup. There isn't much traffic on the Ethernet and right now but we can't seem to get the tape drive to work ;-). ================================================================================ Note 357.10 UNIBUS loading 10 of 11 EISNER::LEDERMAN "Bart Z. Lederman" 16 lines 4-MAY-1989 13:43 -< Actual throughput lower than theoretical >- -------------------------------------------------------------------------------- Just adding up the baud rates of the various lines isn't enough to tell you how much Unibus traffic you will have (though it might give a worst case). Even when terminals are hooked up and 9600 baud and there is someone using each one, the transfer rate is rarely the full 960 characters per second for more than a short burst. And if you have older (i.e. VT100) terminals your real rate will be less than that. We hooked up an 11/70 once with two DH11 equivalents, a TM11 tape controller equivalent, a UDA50 disk controller, and THREE DMR-11 network devices, all set for the full 1 MB local cable transfer rate. This obviously goes 'way beyond the theoretical Unibus limit. But total DECnet traffic was not too great, the tapes don't use up all that much bandwidth, and the system disk was an RP06 on the Massbus with the RA81s on the UDA50 for data storage. So we never ran into any Unibus problems. ================================================================================ Note 357.11 UNIBUS loading 11 of 11 EISNER::FULLER_S "There's only 1 'C' in SERVICE" 19 lines 4-MAY-1989 15:47 -------------------------------------------------------------------------------- > how do you know when you have too many devices on a UNIBUS (other > than the fact that things don't seem to work) ? What doesn't work? The only thing that I've heard about not working is the 8mm tape drive interface board. What does MTI (your vendor) have to say about it? For the others watching, the two boxes are of the BA11-A variety. The first is the normal BA11-A found in an 86xx system (also contains the 4 slot Q-bus backplane for the RLV controller), and the second box is another BA11-A. Each box is rated at +5v at 100A. Power loads are spread out on each backplane, so that no one backplane is handling all the high power loads. The Unibus typically will drive 20 dc bus loads, and each device typically presents 1 dc bus load, therefore (including the DW780), we're well within the dc bus load specs. Each backplane is separated by a 2ft Unibus jumper cable to minimize the lumped loading problems. Stu (Jack's Field _SERVICE_ person)  ================================================================================ Note 358.0 VAXstation 3100, Scholar Modem, 'Nuff Said 3 replies EISNER::MCMICHAEL 53 lines 3-MAY-1989 14:00 -------------------------------------------------------------------------------- I'm attempting to run dynamic asynch DECnet from a VAXstation to a VAX 6220. I've SYSGENed NOA0 & VTA0 as the manual says. The physical setup is this: VAXstation 2000 (VMS v5.0-2) or VAXstation 3100 (VMS v5.1-B) | TTA2: (/DIALUP/HANGUP/MODEM/EIGHTBIT/TYPE_AHEAD/AUTOBAUD) | | BC16E-50 DECconnect office cable (6 pin) | H8571-C Male 25-pin adapter DF242 (Rev C ECO 2) Scholar Plus 2400 baud modem (factory defaults) | | Dialup line | DF242 (Rev C ECO 2) Scholar Plus 2400 baud modem (factory defaults) | | EC25-0010-MM Black Box 25-pin straight connection | TXB0: (/DIALUP/HANGUP/MODEM/EIGHTBIT/TYPE_AHEAD/AUTOBAUD) DMB32 | VAX 6220 (VMS v5.0-2) When I issue a SET HOST/DTE from the 3100, I can log into the 6220, but 30 seconds later the connection breaks and the messages %REM-S-END, control returned to node VS3100 %SYSTEM-F-HANGUP, data set hang-up appear on my screen. When I SET HOST/DTE again, my old process is waiting for me to reattach, but 30 seconds later I'm thrown out again. The word from Colorado Springs is that 6 pins aren't enough for modem control, so that is will never work. However, I find that when I move the cable from TTA2: on the 3100 to TTA2: on the 2000, things work perfectly! Colorado Springs vaguely recalls that the port on the 2000 was "doctored" by Engineering to get this to happen. Does anyone know of differecnces between the 2000 & 3100 ports? Are there switch settings on the Scholar to force the VAXen to "make nice" with one another? NOTES: 1) Issuing a SET HOST/DTE from the 6220 to the 3100 gives me a "VAX/VMS host system modem not wired correctly - contact your system manager" message instead of a username prompt. 2) I called DECdirect to confirm the cable between the VAXstation & the Scholar. The consultant had to call me back since his reference gave conflicting answers, but he did confirm my part numbers. 3) In today's mail I got a flyer for the VAXstation 3100, announcing that "...all VAXstation 2000 applications are immediately transportable to your new workstations." It's Day 6 of modem madness here at Muppet Labs and I'm begining to get a tad sceptical. ================================================================================ Note 358.1 VAXstation 3100, Scholar Modem, 'Nuff Said 1 of 3 EISNER::PRIGOT "Jonathan M. Prigot" 9 lines 3-MAY-1989 15:12 -< 'Taint the same >- -------------------------------------------------------------------------------- It should be interesting to hear their answer, as the MMJ on the VS3100 supports only the DTR and DSR modem control lines while a port set /MODEM is looking for CD and CTS as well. (ref: I/O User Manual P8-14). I suspect that the H8571-A adapter that you are using on the VS2000 is putting out a signal on the CD line (pin 8), probably tied to DSR (pin 6), while the VS3100 will only see DSR and never CD or CTS. After all, we are talking about a 25 pin connection on the 2000 vs. 6 on the 3100. ================================================================================ Note 358.2 VAXstation 3100, Scholar Modem, 'Nuff Said 2 of 3 EISNER::MCCORMICK_B "Bob McCormick" 14 lines 4-MAY-1989 19:03 -< DEC423 usually Data Leads Only! >- -------------------------------------------------------------------------------- Quite a few of DEC's implementation of the EIA 423 standard on the MMJ connectors provides: 1 RDY Out o --+ Ready and Busy (pins 1 & 6, one of the 2 TX+ o | three pairs) are jumpered together. 3 TX- o | For example, this is standard with the 4 RX- o | H3104 cable concentrator with all 5 RX+ o | DS200/DL & CXA16 interfaces. 6 RDY In o --+ I'm not sure what the occasional console (like occasional chair...) has for wiring, or the VS3100 ... but I would guess they only have the transmit & receive signals, and either a constant voltage source, or the jumpered connection, as noted above. ================================================================================ Note 358.3 VAXstation 3100, Scholar Modem, 'Nuff Said 3 of 3 EISNER::MCMICHAEL 8 lines 15-MAY-1989 15:38 -< VAXstation 2000 vs. VAXstation 3100 >- -------------------------------------------------------------------------------- The latest word I have from Field Service is that the VAXstation 2000 data line has a control register with bits for DSR, DTR, & CD. When DSR is turned on, DTR & CD are forced. The VAXstation 3100 was engineered with the data line control register containing a bit for DSR, but not DTR nor CD. Thus the problem. Other customers have complained, and DEC Engineering is reportedly arguing internally whether to modify the hardware, or come up with a software fix.  ================================================================================ Note 359.0 FCO policy changes - 3/89 3 replies EISNER::KENNEDY "Terry Kennedy" 64 lines 3-MAY-1989 21:49 -------------------------------------------------------------------------------- From the March 1989 "At Your Service" Newsletter (Vol. VI, Issue 4): PRODUCT UPDATE FIELD CHANGE ORDER POLICY CHANGES Effective immediately, DIgital has revised the worldwide Field Change Order Policy. This new policy enhances and simplifies FCO Services Order Policy. For example, all FCO Services for Self-Maintenance customers will now be available from a single location: the Customer Returns Center 1-800-225-5385 This new policy also redefines the current FCO categories to more accurately reflect the purpose of the Field Change Order. There have also been some changes made to the material charging process; as outlined in this article. The major changes to the new FCO codes are: Category and Description S PRODUCT SAFETY - AN FCO issued to reduce or eliminate a potential risk of a personal injury or substantial property damage. O PRODUCT OPERATION - An FCO issued to reduce or eliminate an imminent risk of significant product degradation and/or abnormal failure(s). F PRODUCT FUNCTIONALITY - An FCO issued to reduce or eliminate the risk of loss or degradation of a specific function within a product, or to restore a product to its original level of functionality or performance. I PRODUCT IMPROVEMENT - An FCO issued to make a product compliant with new Digital and/or industry standards, or to expand product compatibility. MATERIAL CHARGING: The non-contract/non-warranty customer will be charged for all FCO kits with the exception of the "S" coded Product Safety FCO's. The matrix below summarized the material charging for each code. S Digital pays for the FCO Material O Customer pays DECmailer/Factory Repair Cost Recovery Price F,I Customer pays DECmailer/Factory Repair Price MATERIAL ORDERING: All FCO's should be ordered through the Customer Returns Center at 1-800-225-5385. All FCO's that have been released previous to the new policy with the old FCO codes of M and R will continue to be provided by the Local Field Service Delivery Units. The field service delivery units will continue to provide material on a best effor basis for a period of two (2) years beyond the FCO release date. ================================================================================ Note 359.1 FCO policy changes - 3/89 1 of 3 EISNER::KENNEDY "Terry Kennedy" 23 lines 3-MAY-1989 21:58 -< Yet another dud policy! >- -------------------------------------------------------------------------------- > O Customer pays DECmailer/Factory Repair Cost Recovery > Price Of course, this pricing is not in any DEC price book... > F,I Customer pays DECmailer/Factory Repair Price This would appear to have four *serious* implications: 1) You need a DECmailer account to get FCO kits 2) You have to return your whole board to DEC for FCO application (that's what the DECmailer bit means), so you lose the use of the board for an unpredictable amount of time (contrary to the literature, which states a 5-day turnaround, the manager of the Customer Returns Center told me that there is *NO* guaranteed turn-around time). 3) You will have to pay the (much higher) board repair price to get the FCO installed. For example, the $35 FCO to support RD53's on an RQDX2 would now cost $650 (the DECmailer price for an RQDX2). 4) As loose FCO kits will no longer be available, your 3rd-party service costs may go up, as they can no longer get the FCO kits either. ================================================================================ Note 359.2 FCO policy changes - 3/89 2 of 3 EISNER::KENNEDY "Terry Kennedy" 58 lines 7-OCT-1989 05:13 -< Many months; no apparent action >- -------------------------------------------------------------------------------- It has been quite some time since this was announced. I spoke to a person within Digital (a Mr. Allen) who said it would be "looked into.". I have seen no evidence of any action on this, and it is a major issue for us. In order to test the policy, I arranged for an RQDX3 to be run through the new system from a customer who did not have a DECmailer agreement. Here is what I found: > 1) You need a DECmailer account to get FCO kits Not quite true. You have the "option" of calling the CRC with PO in hand. If you have open account status they will takje your PO and send you a "coupon" for the board exchange. It took 11 working days to get the coupon. > 2) You have to return your whole board to DEC for FCO application (that's > what the DECmailer bit means), so you lose the use of the board for an > unpredictable amount of time (contrary to the literature, which states > a 5-day turnaround, the manager of the Customer Returns Center told me > that there is *NO* guaranteed turn-around time). The board was returned by FedEx overnight and returned 8 working days later. > 3) You will have to pay the (much higher) board repair price to get the FCO > installed. For example, the $35 FCO to support RD53's on an RQDX2 would > now cost $650 (the DECmailer price for an RQDX2). The price was $310, the same as returning a defective board for repair. The total labor involved was to remove the two socketed firmware chips and replace them with a new version. Two chips with a market value of $15.00 for $310.00? Get real! What will be next, a need to buy a "right to use new version" for controller firmware? This FCO was not for new functionality, it was to FIX BUGS with the board. With a policy like this, it's cheaper to go with Emulex (who I have derided here in past times for their firmware upgrade pricing). At least the most I've ever had to pay them was $250, and you *can* bully them into doing it free if you push hard enough. What ever happened to the "old" DEC? You know, the one that went out of its way to help you? Now it's let's-make-a-buck and who cares about the customer. This is the same attitude that brings you the DELQUA upgrade "plan" (buy a new one at list price). What's next - will the early 9000-series CPU's require a return-to-factory FCO for $200,000.00 in a year or so? I wouldn't be surprised given the current attitudes up there. I don't expect new functionality for free, but bug fixes are another matter. Of course, DEC has carefully covered themselves with the new pseudo-warranty program. However, I think that a case could be made that the new warranty terms (return-to-factory unless *DEC* service is purchased, along with the new FCO policy) could be construed as an illegal tying of DEC Field Service to DEC hardware products. This example is especially ridiculous since it's a two-chip change. Possibly some DEC service people may not be up to it, but anyone involved in a large (or even a small) self-maintenance operation should be able to change chips. Even clerical people do it in PC's - it's not hard,.. DEC should recognize this. ================================================================================ Note 359.3 FCO policy changes - 3/89 3 of 3 EISNER::HANRAHAN "Jamie Hanrahan, jeh@simpact.com" 17 lines 7-OCT-1989 14:42 -< Another company that did it right. >- -------------------------------------------------------------------------------- > With a > policy like this, it's cheaper to go with Emulex (who I have derided here in > past times for their firmware upgrade pricing). At least the most I've ever had > to pay them was $250, and you *can* bully them into doing it free if you push > hard enough. Or Dilog. Our DQ153 (9-track tape controller) and Dsomethingorother (controller for Exabyte tape drive) both needed firmware upgrades for VMS 5.1. Called Dilog customer support. Boards are still under warranty (one year, not 90 days!), so new ROMs were sent for free. Plugged 'em in and sent the old ones back. No huhu. Total downtime, maybe an hour (most of that because the cabinet that holds the backplane that holds the board is VERY crowded, had to remove some distribution panels to get to the Qbus). Don't know what the upgrade would have cost out of warranty...  ================================================================================ Note 360.0 8820 vs 6220 floating-point results 15 replies EISNER::PROVOST 10 lines 4-MAY-1989 09:45 -------------------------------------------------------------------------------- A physicist here ran identical binary programs with identical input on an 8820 and a 6620. Results varied slightly. This is quite disturbing. Has anyone else noticed differences in calculations between the two machine classes? We are currently testing the program on a 780 to see who might be broken. ================================================================================ Note 360.1 8820 vs 6220 floating-point results 1 of 15 EISNER::CHADWICK "Keith Chadwick, Fermilab" 19 lines 4-MAY-1989 11:12 -< More information please >- -------------------------------------------------------------------------------- > A physicist here ran identical binary programs with identical > input on an 8820 and a 6620. Results varied slightly. This is > quite disturbing. What were the programs doing (large matrix inversions perhaps, or using lots of SIN/COS/TAN functions)? What versions of VMS are running on both of these processors (there is a documented problem with the MTH$DCOSD, MTH$DSIND, and MTH$DTAND functions in VMS V5.0 which is fixed in V5.0-1) Oh, and by the way, what is a 6620 (or when is it going to be announced 8-)? > Has anyone else noticed differences in calculations between > the two machine classes? We have previously seen differences between 8600's and 8650's which turned out to result from different versions of the processor microcode. -Keith. ================================================================================ Note 360.2 8820 vs 6220 floating-point results 2 of 15 EISNER::KALETKA 22 lines 4-MAY-1989 12:20 -< Further info on the 8820/6220 anomaly. >- -------------------------------------------------------------------------------- -< 8820 vs 6620 results >- The 8820 and 6220 are clustered, running a single copy of VMS 5.0-2. The physics code involved is a large (many 1000's of lines) FORTRAN Monte-Carlo code (LUND). The results from the 8820 and 6220 runs agree for the first seven loops of the Monte-Carlo, but a slight divergence after that (7th significant digit) compounds to cause wildly different final results. I've been able to reproduce a difference using REAL*4 and the MTH$SINCOSD routine. Out of about 3600 calculated values, there are 19 sines calculated differently on the 8820 than on the 6220. The 8820 and 6220 agree on all the cosine calculations. Two 6220's and a VAXStation-II all agree exactly. Where the 8820 and the 6220 disagree, the behavior seems to be that the 8820 is "rounding up" the least-significant-bit of the fraction part of the REAL*4 result, where the 6220 is truncating it The same test, only using REAL*8 and the MTH$DSINCOSD routine, produces results which are identical for the 8820 and the 6220. Have we got a software, hardware, or architecture problem here? ================================================================================ Note 360.3 8820 vs 6220 floating-point results 3 of 15 EISNER::MCMICHAEL 14 lines 4-MAY-1989 14:17 -< Emulation? >- -------------------------------------------------------------------------------- >>Have we got a software, hardware, or architecture problem here? You can issue the following on your systems: $ X = F$GETSYI("F_FLOAT_EMULATED") $ Y = F$GETSYI("D_FLOAT_EMULATED") $ Z = F$GETSYI("G_FLOAT_EMULATED") $ W = F$GETSYI("H_FLOAT_EMULATED") $ SHOW SYMBOL X $ SHOW SYMBOL Y $ SHOW SYMBOL Z $ SHOW SYMBOL W and see if they differ. ================================================================================ Note 360.4 8820 vs 6220 floating-point results 4 of 15 EISNER::FULLER_S "Funky cold madina" 6 lines 4-MAY-1989 15:58 -------------------------------------------------------------------------------- According to the VAX Architecture Handbook (the older "pocketsize" version), "The precision of an F_floating datum is approximately one part in 2**23 [approximately 7 decimal digits]." Stu ================================================================================ Note 360.5 8820 vs 6220 floating-point results 5 of 15 EISNER::NAGY "Frank J. Nagy, VAX Guru" 6 lines 4-MAY-1989 16:48 -< Sounds like trying to exceed 32-bit floating accuracy >- -------------------------------------------------------------------------------- In reply .2 of this stream, it seems to be stated that the application is using 32-bit floating point (F_Floating or REAL*4) and the user is upset by a difference in the 7th decimal digit. As pointed out in reply .4 (just previous), F_Floating is accurate to *approximately* 7 decimal digits. Sounds like the user is expecting too much from the particular architecture he is using; have you tried REAL*8 (D_Floating)? ================================================================================ Note 360.6 8820 vs 6220 floating-point results 6 of 15 EISNER::HARVEY "Jack Harvey" 9 lines 4-MAY-1989 20:20 -< A VAX is a VAX is a VAX???? >- -------------------------------------------------------------------------------- Frank, I think the concern here is that there is a difference between two machines, not the absolute accuracy. A single machine should certainly be completely repeatable even though the seventh decimal digit may be in error. So why should another VAX model get a different seventh digit? It would appear that some microcode for floating point is indeed behaving differently. Which of course shakes up the physicist. ================================================================================ Note 360.7 8820 vs 6220 floating-point results 7 of 15 EISNER::TERRAZAS "Mike Terrazas" 3 lines 5-MAY-1989 09:37 -< CPU classes at question >- -------------------------------------------------------------------------------- It is at least consistent with the uVAX micorcode, since that is what is on the 6220 & the consistency with uVII was noted. So what we have is a difference between uVAX & real VAX. ================================================================================ Note 360.8 8820 vs 6220 floating-point results 8 of 15 EISNER::HASSINGER "Bob Hassinger" 12 lines 5-MAY-1989 11:10 -< Best to do a double precision test at this point >- -------------------------------------------------------------------------------- I believe that differences in FP roundoff have been common across different VAX models, maybe even between the FPP option and non-FFP option versions of certain models. I think you noted that for the first few iterations things are OK then they diverge. Sounds very typical of insufficient precision for the algorithm. In other words, a test with double precision might show that using single precision is insufficient in any case. As I understand it a DP test is considered good practice in work like this. It would be very interesting to know if the results from one CPU match at SP and DP while the results from the other do not. ================================================================================ Note 360.9 8820 vs 6220 floating-point results 9 of 15 EISNER::ROECKEL 12 lines 5-MAY-1989 22:30 -< Hardware vs. Software? >- -------------------------------------------------------------------------------- > < Note 360.7 by EISNER::TERRAZAS "Mike Terrazas" > > -< CPU classes at question >- > >It is at least consistent with the uVAX micorcode, since that is what >is on the 6220 & the consistency with uVII was noted. So what we have >is a difference between uVAX & real VAX. Am I way off base here, or is it due to the fact that some of the instructions for the uVAX are implemented in 'software', were all the 'real' VAXen implement them in hardware? ================================================================================ Note 360.10 8820 vs 6220 floating-point results 10 of 15 EISNER::TERRAZAS "Mike Terrazas" 12 lines 15-MAY-1989 09:47 -< Hardware vs. firmware, or where's the beef? >- -------------------------------------------------------------------------------- > < Note 360.9 by EISNER::ROECKEL > > -< Hardware vs. Software? >- > Am I way off base here, or is it due to the fact that some of the > instructions for the uVAX are implemented in 'software', were all > the 'real' VAXen implement them in hardware? Software is probably too strong a term. It's usually referred to as firmware, since it really is a program, but it's burned into a ROM or a PLA, depending upon the hardware. But yes, I figure the behavior is due to a difference between real VAX microcode for handling the instructions and uVAX firmware for emulating it. ================================================================================ Note 360.11 8820 vs 6220 floating-point results 11 of 15 EISNER::COY "Dale E. Coy" 5 lines 15-MAY-1989 23:02 -< Where from here? >- -------------------------------------------------------------------------------- Has an SPR been submitted? A DSIN Query? A quick, very-incomplete check of the L&T campground at Atlanta did NOT turn up anybody who knew of this problem. (Yes, IMHO it's a L&T problem area) ================================================================================ Note 360.12 8820 vs 6220 floating-point results 12 of 15 EISNER::PROVOST 23 lines 16-MAY-1989 09:58 -< Resolution coming... >- -------------------------------------------------------------------------------- RE: < Note 360.11 by EISNER::COY "Dale E. Coy" > > -< Where from here? >- > Has an SPR been submitted? A DSIN Query? NO, better. > A quick, very-incomplete check of the L&T campground at Atlanta > did NOT turn up anybody who knew of this problem. I have been pursuing this problem with my hardware contacts in Digital. At this point the ucode engineers are looking at it, and will get back to me for more information. I will keep you posted. > (Yes, IMHO it's a L&T problem area) I believe the problem is related to the hardware implementations of the VAX architecture. Thus, the quickest route to resolution is through the HMS SIG. This is not to say that the problem is not of concern to everyone else. Tom ================================================================================ Note 360.13 8820 vs 6220 floating-point results 13 of 15 EISNER::PROVOST 10 lines 4-AUG-1989 11:18 -< THE SOLUTION! >- -------------------------------------------------------------------------------- 8820 utilizes a hardware POLY instruction with certain standards for roundoff before or after normalization. 6220 uses software emulation for POLY. Implementation doesn't quite follow above standards. Software will be fixed in a possible future release of the VMS RUNTIME system. Tom ================================================================================ Note 360.14 8820 vs 6220 floating-point results 14 of 15 EISNER::CHADWICK "Keith Chadwick" 18 lines 7-AUG-1989 12:26 -< Does this affect all VAXen with emulated instructions? >- -------------------------------------------------------------------------------- > < Note 360.13 by EISNER::PROVOST > > > 6220 uses software emulation for POLY. As do all microVAX II and CVAX based processors ... which leads to the next question: > Implementation doesn't quite follow above standards. > Software will be fixed in a possible future release of > the VMS RUNTIME system. Is the problem specific to the 6200 series, or is it in the entire line of VAXen with emulated instuctions? Is there any hope of getting a patch quickly rather than waiting for a future (minor?) release of VMS? -Keith. ================================================================================ Note 360.15 8820 vs 6220 floating-point results 15 of 15 EISNER::PROVOST 17 lines 9-AUG-1989 10:10 -< All emulations fail >- -------------------------------------------------------------------------------- re: < Note 360.14 by EISNER::CHADWICK "Keith Chadwick" > > -< Does this affect all VAXen with emulated instructions? >- This applies to all CVAX based processors. I assume it applies to all systems which software-emulate the POLY. > Is the problem specific to the 6200 series, or is it in the entire > line of VAXen with emulated instuctions? Again, entire line. > Is there any hope of getting a patch quickly rather than waiting for > a future (minor?) release of VMS? I am only an egg. Tom  ================================================================================ Note 361.0 VT340 and SET PASSWORD 2 replies EISNER::HARVEY "Jack Harvey" 13 lines 5-MAY-1989 08:31 -------------------------------------------------------------------------------- Why does SET PASSWORD hate my VT340? It claims I'm entering an invalid character. It doesn't matter what password is entered from the VT340. They are all rejected. 12345678 is invalid. Using a VT320 plugged into the same port and process works fine. VMS V5.0 and VMS V5.1 exhibit this. (Didn't try the 'tweens.) LOGIN works fine with the "invalid" password via the VT340 after I enter it with a VT320. The VT340 is in VT300-7 bit mode with character format 8 bit, no parity. The obvious procedure is to hang a data analyzer on it. Obviously, I haven't got one. Any ideas? ================================================================================ Note 361.1 VT340 and SET PASSWORD 1 of 2 EISNER::WICHMAN "Jon Wichman" 4 lines 5-MAY-1989 09:57 -< "NEW LINE" parameter perhaps? >- -------------------------------------------------------------------------------- Is the error message perhaps "%SET-E-Pwdsyntax", Illegal password syntax" ? Check to see if somehow "NEW LINE" got set in the VT340 setup. ================================================================================ Note 361.2 VT340 and SET PASSWORD 2 of 2 EISNER::HARVEY "Jack Harvey" 5 lines 5-MAY-1989 17:23 -< VT340 does a >- -------------------------------------------------------------------------------- Yes, I expect that's it. I discovered independently that the VT340 is emitting a when I hit the return key. That's probably what SET PASS is upset about. Thanks.  ================================================================================ Note 362.0 Help with MicroVAX II/RQDX3/External Drives 12 replies EISNER::DAVIS 5 lines 5-MAY-1989 13:26 -------------------------------------------------------------------------------- I have a MicroVAX II system which currently has an internal RD53 and 1 external RD53. I want to add an additional external RD53 and need some input on how to configure the RQDX3/RQDXE combination. Also, the appropriate drive select setting for the second external RD53. ================================================================================ Note 362.1 Help with MicroVAX II/RQDX3/External Drives 1 of 12 EISNER::MCCORMICK_B "Bob McCormick" 9 lines 5-MAY-1989 15:45 -< Just cables between boxes >- -------------------------------------------------------------------------------- Boy its been a while since I did this, and I don't have the (old) documentation here (maybe somebody can confirm or elaborate)... I believe all you need to do is get the external disk in the desk top box (leprechan [sp?]) and use a cable out of the J2 port and into the J3 port on the second external disk. If nobody posts a better suggestion/answer send me a mail message and I'll attempt to find the doc and fax you the info. ================================================================================ Note 362.2 Help with MicroVAX II/RQDX3/External Drives 2 of 12 EISNER::ROECKEL 14 lines 5-MAY-1989 22:34 -------------------------------------------------------------------------------- > < Note 362.1 by EISNER::MCCORMICK_B "Bob McCormick" > > -< Just cables between boxes >- > > Boy its been a while since I did this, and I don't have the (old) > documentation here (maybe somebody can confirm or elaborate)... > > I believe all you need to do is get the external disk in the desk > top box (leprechan [sp?]) and use a cable out of the J2 port and > into the J3 port on the second external disk. We researched this about a year ago and I beleive you are correct. Simply daisy-chain the external boxes. ================================================================================ Note 362.3 Help with MicroVAX II/RQDX3/External Drives 3 of 12 EISNER::WICHMAN "Jon Wichman" 3 lines 8-MAY-1989 10:42 -< Requires BC17Y-1J cable >- -------------------------------------------------------------------------------- > Simply daisy-chain the external boxes. The part number for the cable is "BC17Y-1J". ================================================================================ Note 362.4 Help with MicroVAX II/RQDX3/External Drives 4 of 12 EISNER::DAVIS 50 lines 8-MAY-1989 11:41 -< More detail on the specific problem. >- -------------------------------------------------------------------------------- I have all the appropriate equipment, and according to the documentation (skimpy as it is), the cable is run from the J3 connector of the first box to either the J1 or J2 connector on the second external drive. The RQDXE is configured for 1 internal drive (an RD53 in the BA23 cabinet) as RD0, and for 2 external drives RD1 and RD2 (both RD53s in an expansion cabinet, same power supply as the tabletop but just a rackmount cabinet). In the documentation this is designated as configuration F. The first external drive, RD1 (or DUA1), is set with drive select 4 (DS4 jumper pin). There is no mention in my documentation on how to set the drive select jumper pins for RD2 (or DUA2). We have tried various combinations: 1) DS3 set and the cable in J1 2) DS4 set and the cable in J1 3) DS3 set and the cable in J2 4) DS4 set and the cable in J2 5) DS3 and DS4 set and the cable in J1 6) DS3 and DS4 set and the cable in J2 When we attempt to boot, i.e. >>> B DUA2, the activity light on RD2 flashes, but the error DEVOFFLINE is consistently returned. The activity light in fact flashes upon any boot command, B DUA0, B DUA1, etc. It just doesn't seem to recognize RD2 as DUA2. Since I don't have the documentation on how to configure the drive select (DS?) on RD2 that is an unknown. I assume someone out there has this configuration... To make the questions more specific: 1) How should the RQDXE be jumpered for this configuration? 2) How should the cables be routed? 3) How should the drive select pins be jumpered? Current setup is: +-----------+ +------------+ +------------+ | BA23 (RD0)|--------J1| RD53 (RD1) |J3------J1| RD53 (RD2) | +-----------+ +------------+ +------------+ DS4 sel DS? sel By the way, DUA0 and DUA1 work fine. Steve. ================================================================================ Note 362.5 Help with MicroVAX II/RQDX3/External Drives 5 of 12 EISNER::DAVIS 5 lines 8-MAY-1989 11:49 -< Yes, I thought of this also. >- -------------------------------------------------------------------------------- As an additional note, the RD53 that will be DUA2 is a valid VMS V5.1 system disk from another processor. It boots properly if inserted in place of the current DUA1 with no problem. If the current DUA0 is inserted in place of DUA2, it returns the same DEVOFFLINE error. ================================================================================ Note 362.6 Help with MicroVAX II/RQDX3/External Drives 6 of 12 EISNER::MAYHEW "Bill Mayhew" 17 lines 8-MAY-1989 11:57 -< Here's what I think... >- -------------------------------------------------------------------------------- By the DS setting, I assume you are referring to the jumper or switch on the drive itself. This should be set the same as your first external disk. The distinction between units is governed by which rear-panel connector you plug the cable into. The cables want to run (a) from the RQDXE to J1 on the first disk, and (b) from J3 on the first disk to J2 on the second. I assume there is no RX50 involved here anywhere, since the RQDX controllers can only talk to 4 units, and the RX50 counts as 2. According to my documentation, you should have the following jumpers set on the RQDXE: A1-A3, B1-B3, C1-C3, D1-D3, E1-E2, F1-F3, H1-H2, H3-H4, K1-K3, K2-K4, L3-M1, L4-M2. ================================================================================ Note 362.7 Help with MicroVAX II/RQDX3/External Drives 7 of 12 EISNER::MAYHEW "Bill Mayhew" 21 lines 8-MAY-1989 12:18 -< Finishing those thoughts... >- -------------------------------------------------------------------------------- Almost forgot: You also need N1-N2 and N4-P2 to enable the *internal* drive. Aside from that, from what I can tell and remember, you should have no other jumpers on the RQDXE. I just realized that you have these things in a rack-mount box of some type. Is it a BA23-C? If so, then you are building an "unsupported configuration" since you can't have two full-height RD drives in a BA23... and this is why the combo isn't documented in the book. If, OTOH, you are using an H-whatsis (H9302?), which DEC made explicitly to rack-mount two RD5x-R rack-mountable boxes, then the settings I gave you here and in .-1 should be correct. Also, be sure you have an RQDXE (M7513), and not an RQDX1-E; these are two different animals. And, I assume you have an RQDX3, not an RQDX1 or RQDX2, since only the 'DX3 will support more than two winchesters; and, finally, I assume there's no RX50 in the configuration, since an RQDXn can only handle 4 units and the RX50 counts as two. ================================================================================ Note 362.8 Help with MicroVAX II/RQDX3/External Drives 8 of 12 EISNER::DAVIS "Steven P. Davis" 9 lines 8-MAY-1989 18:32 -< More detail. >- -------------------------------------------------------------------------------- There are no RX50s in the configuration and the controller is an RQDX3 and RQDXE. The boards are also up to rev level. The RQDXE is configured as noted in the previous two notes. The drive select on RD2 is jumpered to DS4. It has been tried with the cable running from RD1 J3 to RD2 J2 also. I'll try it again and see what happens, thanks for the input. ================================================================================ Note 362.9 Help with MicroVAX II/RQDX3/External Drives 9 of 12 EISNER::DAVIS "Steven P. Davis" 18 lines 9-MAY-1989 10:44 -< The latest in the continuing saga... >- -------------------------------------------------------------------------------- I have tried connecting up the drives again, and there are some new results. Both external drives are set DS4 (drive select pin 4 jumpered). I ran the disk drive cable from the BA23 box to the J1 connector of the external drive I want to be DUA2 (in other words I switched the order of the external drive cabling); this would now make DUA2 become DUA1. I then ran the next drive cable from J3 of that previous box to the J2 connector of what was the DUA1 box, thus making it DUA2. The command >>> B DUA1 now boots the drive which I want to be DUA2; however, the >>> B DUA2 command now produces the message CTRLERR. I then reconfigured the whole thing so that the cabling is from the back of the BA23 into J1 of the DUA1 box, from there out of J3 to J2 of the DUA2 box and everthing works as previously mentioned in this topic. The command >>> B DUA1 boots fine but >>> B DUA2 produces the message DEVOFFLINE. ================================================================================ Note 362.10 Help with MicroVAX II/RQDX3/External Drives 10 of 12 EISNER::MAYHEW "Bill Mayhew" 22 lines 9-MAY-1989 17:13 -< How many straws left to grasp at? >- -------------------------------------------------------------------------------- Arrrgh... Don't you just love these subsystems which attempt to be "smart"?? How I wish Terry Compton was still with us... he was my guru on all these matters. Anyway... Try connecting ONLY one of your expansion boxes at a time. For each box, try plugging into J1 and then try J2. Note any differences in behavior ;-} ... oughta be the same. Next, check the DS setting on DUA0 inside your BA23; I found an obscure note that says it needs to be DS3. I also find notes indicating that it doesn't matter what the DS settings of the external-box drives are, as long as they're set to *something*. Take two aspirin and call me in the morning. {sigh} This is sounding an awful lot like your controller doesn't want to be an RQDX3... it wants to be an RQDX2. If all else fails, trash one of the RD53s and buy an RD54! ================================================================================ Note 362.11 Help with MicroVAX II/RQDX3/External Drives 11 of 12 EISNER::ERSKINE "Chris Erskine" 1 line 10-MAY-1989 07:17 -< Is there any bus terminators which may need to be pulled? >- -------------------------------------------------------------------------------- ================================================================================ Note 362.12 Help with MicroVAX II/RQDX3/External Drives 12 of 12 EISNER::CROWELL "Shefth of the Fourth Order" 3 lines 14-MAY-1989 20:44 -< DUA2 sez you! >- -------------------------------------------------------------------------------- The RQDX3 is probably deciding for itself which drive goes with which unit number. (This drove our field service guy nuts for a long time.)  ================================================================================ Note 363.0 Relative speed of VAXes 6 replies EISNER::CONROY "Alan Conroy" 9 lines 7-MAY-1989 03:18 -------------------------------------------------------------------------------- Over the years I have been collecting information on the relative performance of various DEC processors (CPU speed, bus bandwidth, maximum memory, etc). This is basically for my own use in comparing cost/performance of systems available on the used market. Anyway, I think I have information on every VAX ever made, except for the CPU speeds of three processors (MicroVAX I, VAX-11/725, and VAX-11/730). So far as I know, DEC never published any relative performance information on these. Does anyone have any benchmarks/documentation which could help me in completing my list? All I'm missing is raw CPU speed (in VUPs, of course). ================================================================================ Note 363.1 Relative speed of VAXes 1 of 6 EISNER::RICE "Gary Rice" 9 lines 7-MAY-1989 09:44 -------------------------------------------------------------------------------- > ... Anyway, I think I have information on every VAX ever made, > except for the CPU speeds of three processors (MicroVAX I, VAX-11/725, and > VAX-11/730). All I'm missing is raw CPU speed (in VUPs, of course). I seem to recall a DECdirect catalog that listed the 11/730 at 0.3 VUPs. Gary ================================================================================ Note 363.2 Relative speed of VAXes 2 of 6 EISNER::MAYHEW "Bill Mayhew" 2 lines 7-MAY-1989 13:58 -< And the answer is... >- -------------------------------------------------------------------------------- The "Systems Growth Planning Guide for Digital CSOs" lists the MicroVAX I at 0.36 VUPs, and the 725 and 730 at 0.30. ================================================================================ Note 363.3 Relative speed of VAXes 3 of 6 EISNER::WICHMAN "Jon Wichman" 9 lines 8-MAY-1989 10:46 -< Sounds great! - Can we have it? >- -------------------------------------------------------------------------------- < Note 363.0 by EISNER::CONROY "Alan Conroy" > -< Relative speed of VAXes >- > Over the years I have been collecting information on the relative performance >of various DEC processors (CPU speed, bus bandwidth, maximum memory, etc). This >is basically for my own use in comparing cost/performance of systems available >on the used market. Anyway, I think I have information on every VAX ever made, Would you considered posting it when you're done? ================================================================================ Note 363.4 Relative speed of VAXes 4 of 6 EISNER::MAYHEW "Bill Mayhew" 11 lines 8-MAY-1989 12:23 -< Similar chart as published by DEC >- -------------------------------------------------------------------------------- You might also be interested in a similar chart that was published by DEC's Northeast sales folks in "DECreport", a new-ish publication that they put out about every other month. It basically covers the same subjects mentioned in .0, plus number/type of bus(ses), max. local disk, "active office users" and "total number of subscribers" (which I believe are measured in terms of ALL-IN-1). It only covers the current machines in the 8000, 6000, and 3000 families, plus the 2000, MVII, and (for reference) the 11/780. The editor for this publication seems to be at: DEC, GBD Editor, 25 Constitution Drive, Bedford, NH 03102-9938. ================================================================================ Note 363.5 Relative speed of VAXes 5 of 6 EISNER::CONROY "Alan Conroy" 105 lines 14-MAY-1989 02:24 -< Here's what I got... >- -------------------------------------------------------------------------------- > The "Systems Growth Planning Guide for Digital CSOs" lists the MicroVAX > I at 0.36 VUPs, and the 725 and 730 at 0.30. Does this list ALL of the VAXes? How do I get a copy? Does it give other specs too? ---- Well, I was afraid someone would ask me for a copy of my list. Fortunately, I'm using my Rainbow and POLYTERM instead of trying to go from the VAX via the server and modem (etc...). Just a couple disclaimers and other notes. The memory capacity column is what DEC says the maximum for the system is. Numbers in parentheses indicate max memory using third party products. It has been an interesting exercise, putting this together. DEC is doing better now, but for a while, it seemed they published about three conflicting specs per CPU. I simply took the most recent specs and ignored the others. This was not intended for public consumption, but since you asked for it... And about the PDPs... Well, I couldn't resist the urge to compare apples and oranges. For machines that DEC currently sells, you will find more detailed information in the VAX Systems and Options Catalog. The blank spots indicate things I don't know.  DEC CPU COMPARISONS Aggregate CPU CPU speed (VU) Max Memory (Mb) Bus(es) I/O speed --- -------------- --------------- ------- ---------- PDP-8M 32Kb PDP-11/03 28Kb LSI PDP-11/04 28Kb PDP-11/05 28Kb U PDP-11/10 PDP-11/40 U PDP-11/45 256Kb U PDP-11/60 256Kb U PDP-11/23 .12 4 PDP-11/34 .13 PDP-11/24 .15 4 PDP-11/53 .3 4 Q 3.3 VAX-11/725 .3 3 U VAX-11/730 .3 5 U 1.5 PDP-11/53+ .36 4 Q 3.3 MicroVAX I .36 4 Q 3.3 PDP-11/44 .39 4 U PDP-11/73 .42 4 Q 3.3 PDP-11/70 .6 4 U 5.8 VAX-11/750 .6 8 (15) U,CMI PDP-11/83 .72 4 Q 3.3 PDP-11/84 .72 4 U MicroVAX II .9 16 Q 3.3 VS II .9 16 Q 3.3 MicroVAX 2000 .9 14 (16) n/a n/a VS2000 .9 14 n/a n/a VAX-11/780 1 64 U,SBI 8 VAX-8200 1 128 BI 13.3 VS8000 1.2 32 BI 13.3 VAX-11/785 1.2 64 U,SBI 8 VAX-8250 1.2 128 BI,U 13.3 VAX-11/782 1.6 80 U,SBI 13.3 (ASMP) VAX-8300 1.9 128 BI, U 13.3 VAX-8350 2.3 128 BI,U 13.3 (SMP-2) MicroVAX 3300 2.4 52 Q 3.3 MicroVAX 3400 2.4 52 Q 3.3 VS3100 2.7 32 SCSI 1.2 MicroVAX 3200 2.7 16 Q 3.3 MicroVAX 3500 2.7 64 Q 3.3 VS3200 2.7 16 Q 3.3 VS3500 2.7 32 Q 3.3 MicroVAX 3600 2.7 32 Q 3.3 VS3600 2.7 64 MicroVAX 6210 2.8 256 BI, U 60 VAX-8500 3.1 MicroVAX 3800 3.8 64 Q 3.3 MicroVAX 3900 3.8 64 Q 3.3 VAX-6310 3.8 256 BI,U 60 VAX-8530 4 VAX-3520 5.4 64 Q 30 (SMP-2) VS3520 5.4 64 Q 30 (SMP-2) VAX-6220 5.5 256 BI,U 60 VAX-8600 6 260 U, SBI 8 VAX-8810 6 512 BI,U 30 VAX-8550 6.1 320 BI,U 16 VAX-8650 6.1 260 BI, SBI 8 VAX-8710 6.1 256 BI 13.3 VAX-6312 7 512 BI,U 60 (2x6310) VAX-6320 7.5 256 BI,U 60 (SMP-2) VAX-6230 8.3 256 BI 60 VAX-3540 10.8 64 Q 30 (SMP-4) VS3540 10.8 64 Q, SCSI 30 (SMP-4) VAX-6240 11 256 BI,U 60 (SMP-4) VAX-6330 11.3 256 BI 60 (SMP-3) VAX-8820 11.4 512 BI,U 60 (SMP-2) VAX-8800 12 256 BI 13.3 VAX-6340 15 256 BI 60 (SMP-4) VAX-8830 17 512 BI,U 90 (SMP-3) VAX-6350 18.6 192 BI 40 (SMP-5) VAX-6360 22 192 BI 40 (SMP-6) VAX-8840 22 512 BI,U 120 (SMP-4) VAX-8842 22 1024 BI,U 30 (2x8820) VAX-8974 24 2048 BI,U 30 (4x8810) VAX-6333 30 320 BI 60 (3x6330) VAX-8978 48 4096 BI,U 60 (8x8810) ================================================================================ Note 363.6 Relative speed of VAXes 6 of 6 EISNER::MAYHEW "Bill Mayhew" 33 lines 14-MAY-1989 14:36 -< Alternate sources >- -------------------------------------------------------------------------------- > Does this [Systems Growth Planning Guide for Digital CSOs] list > ALL of the VAXes? No. It doesn't list the clustered-pseudo-VAXen, for instance (6312, 6333, 8974, et al.), nor does it list some older machines (8200, 8300, 8500). > How do I get a copy? Don't know if you *can*. It is published for DEC Complementary Solutions Organizations (resellers, cooperative marketing partners, ...). I'm sure there is something similar for DEC direct sales people, don't know if there's something similar for end-user customers. > Does it give other specs too? Not really. It lists "latest memory chip technology" (i.e. if the DEC memory for a given computer uses 64K or 256K or 1Mb chips; not all that interesting). It does NOT list bus bandwidth, etc. That's about it. There *is*, OTOH, a chart in the Systems volume (Vol.5) of the Digital Reference Service that covers this stuff and a lot more (max. communications lines, CPU cache size/access time, etc.) but only on the relatively current machines. There's one in the same volume for PDP-11s as well, but it references a PDP-11/70 as "1" for its CPU comparisons (a PDP Unit of Processing, or PUP, no doubt ;-} ). In the Apr-Jul 1989 edition of this book, look at pages 2.12 ff. and 3.18 ff., respectively.  ================================================================================ Note 364.0 Formatting disks on a VAXstation 3100 4 replies EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 12 lines 19-MAY-1989 21:44 -------------------------------------------------------------------------------- <<>> I am playing with a VAXstation 3100 at a customer's site and am having a bit of a problem: how do you format a floppy disk on the thing? There appears to be a ROM program for formatting the SCSI hard disks, but not the floppy. Is there a VMS-based disk formatter? It is always possible that I am using the wrong type of diskettes. The ones I have are Sony MFD-2DD Double-sided, Double-density. They are brand new and have never been in a machine. ================================================================================ Note 364.1 Formatting disks on a VAXstation 3100 1 of 4 EISNER::BURNET "John Burnet" 0 lines 21-MAY-1989 03:37 -< Wouldn't INIT do it? >- -------------------------------------------------------------------------------- ================================================================================ Note 364.2 Formatting disks on a VAXstation 3100 2 of 4 EISNER::PRIGOT "Jonathan M. Prigot" 4 lines 21-MAY-1989 13:25 -< INIT/DENSITY=DOUBLE >- -------------------------------------------------------------------------------- I have always used INIT/DENSITY=DOUBLE DUA2: mumble. I seem to remeber that if I tried just plain old INIT, the format bombed. I do remember that the ROM based disk diag will work with the floppy as well. TEST 70, I believe, although my VS3100 Release Notes are at work. ================================================================================ Note 364.3 Formatting disks on a VAXstation 3100 3 of 4 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 13 lines 30-MAY-1989 12:50 -< Problem solved (I think) >- -------------------------------------------------------------------------------- >> I have always used INIT/DENSITY=DOUBLE DUA2: mumble. I seem to remeber that if >> I tried just plain old INIT, the format bombed. I do remember that >> the ROM based disk diag will work with the floppy as well. >> TEST 70, I believe, although my VS3100 Release Notes are at work. The problem turned out to be that I was using DSDD floppies instead of DSHD (High Density) floppies. I haven't actually tried it, but feel confident that that was the problem. I don't think that TEST 70 will work for formatting, however. My brief look at it seems to indicate that it is only for the SCSI drives, not for floppies. INIT/DENSITY=DOUBLE does seem to be the correct command. ================================================================================ Note 364.4 Formatting disks on a VAXstation 3100 4 of 4 EISNER::PRIGOT "Jonathan M. Prigot" 6 lines 30-MAY-1989 17:03 -< 0 memory errors logged >- -------------------------------------------------------------------------------- > I don't think that TEST 70 will work for formatting, however. > My brief look at it seems to indicate that it is only for > the SCSI drives, not for floppies. My guess (from your offline MAIL), is that the information was published ONLY in the VMS V5.1-B Cover Letter (AV-MM58A-TE).  ================================================================================ Note 365.0 Adding RD5x's to a BA23 MicroVAX 18 replies EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 35 lines 22-MAY-1989 14:57 -------------------------------------------------------------------------------- I have a MicroVAX II with the following: o BA23 cabinet o RD54, RQDX3 o TK50, TQK50 o DHV11 o 7 Mb (a 4Mb board, a 2 Mb board, and 1 Mb on the CPU) The '54 was a recent upgrade; the original drive was a '53, which is currently doing paperweight duty. I also have a '52 hanging around. The question - What hardware will I need in order to get the '52 and the '53 hooked up? I know that a "vanilla" DEC solution would invole an RQDXE and cab kit, a couple external cables, and two leprechaun boxes. However, leprechaun boxes are not available (unless you buy an RD5x-DA). I looked into Digital Basic's disk cabinets, which seem somewhat reasonable (I say "somewhat", since I have a hard time figuring out why a dinky 100W power supply and some sheet metal should cost what everybody charges for these things!), but the "gotcha" is their "cable" kit, which is a couple C notes! I have a *real* hard time paying $200 for a few hunks of ribbon cable! To make this more interesting, my BA23 has the front-panel used for RX50 configuration, so it can only write-protect and online/offline the internal drive. I would like to be able to write product my external drives, as well as take them on and off-line, via a front-panel on the drive cabinet. This is my home system, so I'm trying to go as cheaply as possible. As such, I'm also not bound by the usual "Gee, you can't put this non-standard configuration under a Field Service contract" stuff. Anybody have any suggestions? Ed ================================================================================ Note 365.1 Adding RD5x's to a BA23 MicroVAX 1 of 18 EISNER::KENNEDY "Terry Kennedy" 14 lines 22-MAY-1989 19:02 -< Econo-lodge? >- -------------------------------------------------------------------------------- > This is my home system, so I'm trying to go as cheaply as possible. > As such, I'm also not bound by the usual "Gee, you can't put this > non-standard configuration under a Field Service contract" stuff. Try this: 1) Order kit BA23-UC, which gives you the 6-button front panel. 2) Order a PC hard disk box. Snake the cables from the BA23 'Fixed disk 1' connectors into the PC box. (The restriction on one hard disk in a BA23 is due to power supply loading. By using an external supply you can use the BA23's internal distribution panel to get the data lines you need. P.S. - The second drive must be jumpered as drive 4 of 4 (your existing drive is 3 of 4) ================================================================================ Note 365.2 Adding RD5x's to a BA23 MicroVAX 2 of 18 EISNER::MAYHEW "Bill Mayhew" 9 lines 22-MAY-1989 19:20 -< Holiday Inn -- no surprises ;-} >- -------------------------------------------------------------------------------- If you do want something a little neater, order an RD51-D, which is still sporadically available from DEC (check the E-store, it comes and goes) and others for a few hundred $$. Yank the RD51 and sell it or take it apart to see how it works ;-} ... and insert the RD53. The RD51-D comes with the outside-the-box cable (which is a long, shielded, round cable, which meets FCC specs and so forth). You still need the RQDXE, of course. This, plus the BA23-UC Terry mentioned, is the cheapest way to build a "legal, official" configuration that I know of. ================================================================================ Note 365.3 Adding RD5x's to a BA23 MicroVAX 3 of 18 EISNER::KOZAM 8 lines 22-MAY-1989 19:52 -< Bad Boy Configuration >- -------------------------------------------------------------------------------- To hell with legal - let's get REALLY radical. Take an old CPU box - BA11-MA seem to be best. Yank out the card cage. Now you've got a housing with power supply. All you need is some Radio Shack ribbon cable and connectors. On a more practical side, the CPU boxes can handle the power requirements quite nicely. ================================================================================ Note 365.4 Adding RD5x's to a BA23 MicroVAX 4 of 18 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 22 lines 25-MAY-1989 09:53 -< Good info so far - how about daisy-chaining? >- -------------------------------------------------------------------------------- Thanks for all the suggestions - I particularly like the PC hard disk box with the ribbon cables coming from the BA23 dist. board - Cheap and totally in keeping with the home-brew nature of this project! The only problem is that I have *two* drives I want to add, not one. Is anybody out there savvy with the details of the ST412 interface? In talking with the people that offered me a $200 cable kit, they said that DEC engineered their disk controllers/wiring the hard way, and that there is no need to have things like RQDXE's, as the ST412 interface allows disks to be connected without all additional hardware DEC requires. My assumption is that these folks really used the drive select jumpers on the drives as drive selects, unlike DEC, which states that there must be a drive select jumper in place on the external drives (or was it the drives in a BA123? I forget), but the actual jumper setting is not important. This, and the fact that I see terminator resistor packs on the drives leads me to believe that it should be possible to simply daisy-chain the little buggers, just like Massbus drives. Does anybody have a clue on how to set this sort of thing up? Ed ================================================================================ Note 365.5 Adding RD5x's to a BA23 MicroVAX 5 of 18 EISNER::KENNEDY "Terry Kennedy" 24 lines 25-MAY-1989 16:15 -< Yet more info on the distribution panel >- -------------------------------------------------------------------------------- Well, first some background. ST-506 interface hard disks essentially use a floppy-style interface, they just put the data and some other stuff on the 20-pin cable. Knowing that, it is easy to see how a hybrid controller for both floppies and hard disks can be built. The distribution board in the BA23 takes drive selects 0 and 1 and routes them to the RX50. Drive select 3 comes out on the fixed disk 0 connector, and drive select 4 comes out on the fixed disk 1 connector. Thus, all 4 possible unit selects are used. All the RQDXE does with its zillions of jumpers is re-route some of these off to the external box (diverting them from the BA23 dist. panel, by the way). Several things conspire to make it difficult for you to easily do your own thing with cables - first, the BA23 control panel is also wired to the distribution board to force write protect and ready. Secondly, the native signal connector on a RQDXn is a 50-pin - you'd need to split it out into the component parts. That isn't really hard (it's totally pas- sive), but it is a bear of a wiring job. Last, the RQDXn conspires by fiddling the interface lines to see if the device at each unit number is a floppy (and which type, for an RQDX3), or a hard disk (and which type). Attempting to fool the controller will only hang it up and require a bus reset to get it un-wedged. ================================================================================ Note 365.6 Adding RD5x's to a BA23 MicroVAX 6 of 18 EISNER::HARVEY "Jack Harvey" 20 lines 25-MAY-1989 23:26 -< Ribbon Cables and the BA123 Box >- -------------------------------------------------------------------------------- The ribbon cables between the board edge connectors on the RD5n drives and the Berg connectors on the BA23 backplane assembly that some people want $200 for appear to be identical to the cables one gets at the local PC shop to add hard disks to an IBM PC or clone. Length may be a problem, so just get some raw ribbon and make whatever length and width you need using the PC cable connectors. There is a marvelous board in the BA123 box that breaks out the cables from an RQDX3 to connect to the several RD5n disks that can be mounted in this box. The board does *not* need to be plugged into a backplane and only needs the DC power. I've used it to pile (literally) two additional hard disks on a BA23 by just soldering power wires to the board edge connector and connecting them to the chassis power. It looks like hell, but VMS doesn't mind. Three RD54's on a BA23 is a goodly machine. What's the module number, Terry? I haven't got a BA123 handy. ================================================================================ Note 365.7 Adding RD5x's to a BA23 MicroVAX 7 of 18 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 9 lines 26-MAY-1989 00:38 -< M9058 >- -------------------------------------------------------------------------------- > There is a marvelous board in the BA123 box that breaks out > > What's the module number, Terry? I haven't got a BA123 > handy. The BA123 signal distribution board is a M9058. The read/write switch that plugs into it is a 54-16244-02 ================================================================================ Note 365.8 Adding RD5x's to a BA23 MicroVAX 8 of 18 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 12 lines 26-MAY-1989 11:41 -------------------------------------------------------------------------------- Hmmmm, the M9058 sounds pretty interesting - I just looked at a BA123-based MicroVAX, and noticed that it plugs into a "nano-backplane" of some sort (I assume just for power). Will this module work and play well plugged into the backplane of my BA23, or do I have to cobble up something to get power to it? Thanks to all for your suggestions! Now to figure out which path to pursue (based, as usual, on funding). Thanks again, Ed ================================================================================ Note 365.9 Adding RD5x's to a BA23 MicroVAX 9 of 18 EISNER::HARVEY "Jack Harvey" 23 lines 26-MAY-1989 12:40 -< Mounting and Powering the M9058 >- -------------------------------------------------------------------------------- >Will this module work and play well >plugged into the backplane of my BA23, or do I have to cobble up something >to get power to it? Yes, it should work fine plugged into the BA23 backplane, either the AB or CD side, since it uses the standard Q-Bus power fingers on its connector. However, the animal, with its Berg connectors standing on end and cables coming out at right angles will take up lots of space in your box. I cobbled. Three wires, +5, +10 and ground, soldered to the board, connected to the power supply lines on the BA23. You can then put the animal where your ribbon cables fit best. Like in a shoebox under an outboard RD53. If you want a socket to plug the board into, look for an old BA11-M box. That was the LSI-11, or PDP-11/03 cabinet. People don't know what to do with them now. But they have +5 and +10 and will probably power a disk drive or two. Hey, we want a report on how you make out! Some of us will be right behind you. ================================================================================ Note 365.10 Adding RD5x's to a BA23 MicroVAX 10 of 18 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 11 lines 26-MAY-1989 16:28 -< Alms for the poor.... >- -------------------------------------------------------------------------------- < Note 365.9 by EISNER::HARVEY "Jack Harvey" > > Hey, we want a report on how you make out! Some of us > will be right behind you. No problem - I'll let all of you know how it works out. Actually, you may see a note from me in SWAP_MEET, as my employer has little in the way of random parts that would prove useful in such an endeavour. Anybody have a BA123 they don't know what to do with? ;-) Ed ================================================================================ Note 365.11 Adding RD5x's to a BA23 MicroVAX 11 of 18 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 13 lines 12-JUN-1989 14:47 -< Status report >- -------------------------------------------------------------------------------- Well, here's the scoop so far. I've found a dual drive enclosure with power supply for $60.00, and two sets of drive cables for $10.00. This was the simple part. Now while I'm waiting for these goodies to show up, I have to figure out a way to get the right signals out of the BA23 for the two drives. I've checked local sources for an M9058 or an RQDXE, but no dice. I'm going to give SWAP_MEET a try next... As an interim step, I'll probably pull the "fixed disk 1" signals off of the BA23's distribution panel, so at least I'll have one external drive going. Thanks again for everyone's help - I'll let you know whether my system passes the smoke test as soon as the parts arrive... Ed ================================================================================ Note 365.12 Adding RD5x's to a BA23 MicroVAX 12 of 18 EISNER::HARVEY "Jack Harvey" 1 line 12-JUN-1989 20:14 -< Keep Up the Good Work! >- -------------------------------------------------------------------------------- ================================================================================ Note 365.13 Adding RD5x's to a BA23 MicroVAX 13 of 18 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 26 lines 23-JUN-1989 16:08 -------------------------------------------------------------------------------- Well, here's the scoop so far. The enclosure arrived and is quite nice! It will hold two drives, has sufficient power, and even has some room for an interface board. After taking the DEC standard skid plate off, the drive bolts right in (using those side mounting holes that you thought you'd never use). All in all, a very good buy for $60.00. I also ordered some drive cables; 22 inches of ribbon cable with all the proper connectors. These went for about $6.00 a set. If anybody is interested in getting the addresses of the companies I purchased this gear from, send me mail. I have no stake in either company, other than being a satisfied customer. Anyway, SWAP_MEET led me to an RQDX1-E that was looking for a new home, so I know have a way to get the signals out of my BA23. Yesterday, some of my DEC buddies were kind enough to cough up some technical specs on the RQDX3 and related items, so I'm learning about how disks are connected to RQDX3's, and how to implement write-protect and online/offline buttons. I've also learned some interesting things about the capabilities of ST-506/ST-412 interface drives, and how DEC utilizes them. I'm still a little ways from designing and building the little black box that will take input from the DB50 provided by the RQDX1-E and a set of buttons, and put it out onto the ribbon cables my drive expects, but I'm close. The interface is conceptually simple, the electronics are all good ol' TTL, and there's a Radio Shack right down the street. (What the hell, if DEC uses R. S., so can I) ;-) Ed ================================================================================ Note 365.14 Adding RD5x's to a BA23 MicroVAX 14 of 18 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 31 lines 16-OCT-1989 13:54 -< Another installment in the never-ending saga... >- -------------------------------------------------------------------------------- Here's an update on my situation. I was able to upgrade my RD52 to a '53, so now I have two '53s sitting in my industrial surplus expansion box. I've purchased a used RQDX1-E, and obtained a schematic for same. In the course of studying the '1-E schematic, I found that there would be a bit of a problem using it, as it doesn't really support the highest-order head select line. This isn't really an issue, unless one tries to hang an RD54 off of it, but the fact that that line (and one or two others) didn't have the "anti-glitch" circuitry on it makes me a bit uneasy using it (especially given the dependability of our local electric company. So, I looked around, and was able to snap up a brandy-new RQDXE for about half price. I then got (Thanks, Alan!) a schematic for the BA23's signal distribution board. Looking this jewel of a document over, it should be a piece of cake. My plan is to set the RQDXE up to speak to a second BA23, use the RQDXE cab kit to get the signals out onto the big round 50 conductor cable, use a second cab kit (left over from the RQDX1-E) to convert from "big round" cable to ribbon cable, and to get the signals into my external box. Next, using some surplus drive cables, four switches, four LEDs, a Radio Shack proto board, a couple transistors and resistors, and some hookup wire, I'll convert the 50 conductor ribbon cable from the cab kit into the two 34 conductor, and two 20 conductor drive cables, including online/offline and write protect switches. I have all the parts I need (with the exception of the transistors, which I need to look-up in a cross-reference). Once I get a lazy afternoon/evening away from my nine month-old (who just *loves* to grab wires, knawing/pulling on same), I'll put it together. If there's sufficient interest, I'll provide/post more detailed wiring info for other home-brew crazies out there.... Ed ================================================================================ Note 365.15 Adding RD5x's to a BA23 MicroVAX 15 of 18 EISNER::HARVEY "Jack Harvey" 1 line 16-OCT-1989 19:45 -< Thanks for the progress report >- -------------------------------------------------------------------------------- ================================================================================ Note 365.16 Adding RD5x's to a BA23 MicroVAX 16 of 18 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 42 lines 24-OCT-1989 13:50 -< A rose by any other name would still have 50 pins... >- -------------------------------------------------------------------------------- AAARRRGGGHHH!!! Last night I sat down with a copy of the BA23 signal distribution board schematic, RQDX3 user's guide, RQDXE user's guide, and RD53 technical manual, and started to put together my own little schematic for a board that will tie all the goodies together. As I started to jump from document to document, cross-referencing connections, etc., I found a disconcerting problem: the numbering schemes used by each document are different! The big gotcha is the BA23 signal distribution board schematic, which shows its input (from an RQDX-whatever) as being the opposite of what's shown coming out of the RQDX3! (ie, the signal on pin 1 of the RQDX3 is shown as coming in on pin 50 of the sig. dist. board). So, I thought I'd go to the source, and check the hardware. I pulled out the cab kit for my RQDXE, and the 50-pin Berg connector's pin 1 is connected to pin 50 of the external DB50 connector! To make things even worse, this is the exact oppposite of the cab kit I got with an RQDX1-E! I'm not so sure *what* to think now. Can anybody tell me the location (and signal name) of pin 1 on an RQDX3? If I have this, I can trace out the rest. I need the signal name, so I can figure out what pin this corresponds to on my schematic. I also have the same problem with the connectors on my RD5x's. The RD53 tells me where pin 1 is for each connector, but it doesn't tell me where any other pin is, so I can't be sure about the actual location of each signal. For example, if we were talking about an 8 pin connector, it could be numbered: 1357 2468 1234 8765 1234 5678 Any light that could be shed on this would be most appreciated! Ed-who-has-found-you-cant-trust-the-little-numbers-molded-on-the-connectors! ================================================================================ Note 365.17 Adding RD5x's to a BA23 MicroVAX 17 of 18 EISNER::KENNEDY "Terry Kennedy" 22 lines 24-OCT-1989 16:58 -< More info >- -------------------------------------------------------------------------------- > So, I thought I'd go to the source, and check the hardware. I > pulled out the cab kit for my RQDXE, and the 50-pin Berg connector's pin 1 > is connected to pin 50 of the external DB50 connector! To make things even > worse, this is the exact oppposite of the cab kit I got with an RQDX1-E! Well, DEC has this little trick - sometimes they use the stripe on the cable to mean Pin 1, sometimes it means Pin N (where N is the highest possible pin). You will frequently find DEC manuals saying "align with the stripe", which often means "...and ignore the Pin 1 markings on the con- nector". > I'm not so sure *what* to think now. Can anyone tell me the > location (and signal name) of pin 1 on an RQDX3? ... It's MFMWRTDT1(H). For reference, 25 is Ground, 26 is Index(L) and 50 is Ready(L). [EK-RQDX3-UG-001, P. B-1) > 1357 > 2468 Although on the drives it's actually 1357.. 2468.. ================================================================================ Note 365.18 Adding RD5x's to a BA23 MicroVAX 18 of 18 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 0 lines 25-OCT-1989 11:25 -< Thanks! I'll give it another try... >- --------------------------------------------------------------------------------  ================================================================================ Note 366.0 VAX Address Alignment for Speed 31 replies EISNER::HARVEY "Jack Harvey" 27 lines 24-MAY-1989 09:38 -------------------------------------------------------------------------------- When the VAX-11/780 first appeared, one of its remarkable features (to PDP-11 programmers) was the fact that words could start at odd addresses. I recall reading about a special 8 byte memory register which (presumably) made this work efficiently. Then about a year after the debut, I was doing some MACRO benchmark testing to try to establish that the 780 could beat the pants off a PDP-11/70. (I failed.) I was advised by DEC software specialists that to get best performance from the 780, I should use the .ALIGN LONG directive to assure fast access to my binary data. Dull thud. "What became of that special memory register that allows words, long words, etc., to be aligned at odd addresses?" "Never heard of it. Align your words if you want speed." I was never able to prove or disprove this assertion and I haven't heard it repeated for years now. I'm starting a new project developing massive amounts of MACRO code for best speed. Does this folk theorem (alignment of binary data) still prevail, or have the new machines successfully solved the problem if it ever existed? I hesitate even to ask. If this is a non-problem, asking the question is likely to revive the rumor. *Please*, let's have facts rather than superstition. ================================================================================ Note 366.1 VAX Address Alignment for Speed 1 of 31 EISNER::LEDERMAN "Bart Z. Lederman" 6 lines 24-MAY-1989 12:54 -< DEC apparently believes it. >- -------------------------------------------------------------------------------- I have not seen any definitive data which shows that alignment will help or hurt. But I think there are some people within DEC who believe that it helps because if you look at the code generated by the FORTRAN compiler (and maybe some other compilers too) you will find that they do longword alignments on blocks of code, especially after jumps and branches and before labels. ================================================================================ Note 366.2 VAX Address Alignment for Speed 2 of 31 EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 10 lines 24-MAY-1989 12:59 -< Delay in memory, not cpu, I think >- -------------------------------------------------------------------------------- As I remember the problem for odd byte addressing was not in the CPU itself but in the data fetch from memory. Memory reads were done as 4 byte chunks, thus calling for a longword that started on an odd byte (or the wrong even byte even) caused *two* memory fetches rather than a single memory fetch, effectively doubling the memory latency. With the better pipelines and caches on the newer machines maybe this has all changed... (Sorry, I don't remember the source, it may have been an early course from Ed Services on programming...) ================================================================================ Note 366.3 VAX Address Alignment for Speed 3 of 31 EISNER::KILGALLEN "Larry Kilgallen" 13 lines 24-MAY-1989 16:40 -< It's like the dancing dog >- -------------------------------------------------------------------------------- The marvel is not how well it dances, but the fact that it dances at all. For various CPU's there is a varying degree of effort to reduce the performance impact of memory access which crosses the natural boundary of memory cells, but I believe that in all cases the VAX hardware designers have chosen to make the memory-to-CPU bus wider than 8 bits. Given that wider bus, I believe all VAX designs have a "natural" alignment for which transfers are more efficient. The 780 was a marvelous step forward in allowing the free access of memory, but just like virtual memory it grants the programmer a chance to be sloppy, not a guarantee that being sloppy will be as efficient as being neat. ================================================================================ Note 366.4 VAX Address Alignment for Speed 4 of 31 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 16 lines 24-MAY-1989 16:41 -< Alignment recommendations confirmed >- -------------------------------------------------------------------------------- Seton is correct. Aligning data on its natural boundaries allows all data <= a quadword in size to be fetched with a single memory reference. (Note that most memory subsystems support 8-byte chunks, not just 4.) >With the better pipelines and caches on the > newer machines maybe this has all changed... Not really. If it's in the cache, fine, but if you have to go to memory you have to go to memory. By the way, there ARE a few instructions -- some, but not all, of the interlocked instructions, used for interprocesor synchronization under SMP -- that DO have strict alignment requirements: Queue headers for INSQTI, et al, mut be on quadword boundaries; and the word accessed by ADAWI must be on a word boundary. (BBSSI and BBCCI have no alignment requirements.) ================================================================================ Note 366.5 VAX Address Alignment for Speed 5 of 31 EISNER::HARVEY "Jack Harvey" 11 lines 24-MAY-1989 22:14 -< Rats >- -------------------------------------------------------------------------------- I'm almost sorry I asked. (But not really.) I think the bottom line, for my case, is that since I can't control what machine the application may run on, and DEC might just introduce a real minimum VAX with no cache, pipe lining, look-ahead buffers, etc., I had better observe the alignment rules. It's probably politically desirable too, and only hurts when it's going to rain. Oh, by the way, does anyone have any test data to back this up? Pure reason is all well and good but... ================================================================================ Note 366.6 VAX Address Alignment for Speed 6 of 31 EISNER::KILGALLEN "Larry Kilgallen" 13 lines 25-MAY-1989 06:34 -< Politically desirable is in the eye of the beholder >- -------------------------------------------------------------------------------- It might be best to *mis-*align the data if you are lobbying for a bigger CPU. As for test data, nothing should be so convincing as a test you wrote yourself and ran on your own machine. Perhaps accessing all through a large array repeatedly would be good. You should do so backwards and skipping alternate longwords/quadwords to avoid taking advantage of CPUs which might store the "left-overs" from the last memory fetch for possible future use. Of course this is a contrived example, and the actual performance impact of alignment on *real* applications will depend upon whether your business is physics or phinance. ================================================================================ Note 366.7 VAX Address Alignment for Speed 7 of 31 EISNER::HARVEY "Jack Harvey" 29 lines 25-MAY-1989 23:06 -< Data Alignment Policy >- -------------------------------------------------------------------------------- > -< Politically desirable is in the eye of the beholder >- Politically desirable policies: 1. Backups - always do so many backups and keep the tapes so long, your boss complains about the time your are spending and the storage space the tapes are taking up. 2. Working Hours - always get to the office before your boss and be there to say goodnight to him when he leaves. 3. Data Alignment - do it even if there isn't a shred of evidence it helps, until your boss complains about the wasted memory. Larry makes a point about testing. I agree. It's hard as hell to devise a test that will get "normal" performance from a cache. So I'm not surprised no one has come forward with test statistics on the effect on non-alignment of binary data types. Jamie points out that some memories return eight byte chunks. If the chunks are aligned on quadword boundaries, that's the memory system designer's choice. There's no physical reason why semiconductor (as opposed to core) memory reads can't start at any byte address. There are, however, (last time I looked) pretty good reasons for most cache schemes to align on some particular boundary. Sigh. ================================================================================ Note 366.8 VAX Address Alignment for Speed 8 of 31 EISNER::BURNET "John Burnet" 8 lines 26-MAY-1989 19:26 -< Random notes about data-structure alignment and VAX C >- -------------------------------------------------------------------------------- When laying out data-structure fields, you can avoid having to worry about alignment too much if you follow two simple (and obvious) rules: First, quadword-align all data structures. Second, store data elements in each structure in decreasing order of size. Please note that VAX C (as of V3.0) now has an "_align" storage-class modifier, so you can specify alignment restrictions in C as well as Bliss, Macro, and probably other languages. (Pascal?) ================================================================================ Note 366.9 VAX Address Alignment for Speed 9 of 31 EISNER::KILGALLEN "Larry Kilgallen" 9 lines 28-MAY-1989 14:12 -< Alignment control in other languages >- -------------------------------------------------------------------------------- Pascal provides full alignment control. Ada provides alignment attributes only within records. Scan does not provide alignment control. Note that in these and "most" VAX languages, the purpose of alignment control is to match some externally specified non-natural alignment. By default compilers do memory allocation so as to align data structures for rapid memory access (as we have been discussing) unless otherwise directed by one of the alignment controls discussed. ================================================================================ Note 366.10 VAX Address Alignment for Speed 10 of 31 EISNER::HARVEY "Jack Harvey" 54 lines 29-MAY-1989 07:51 -< Datatype Alignment Viewed as Harmful >- -------------------------------------------------------------------------------- >< Note 366.8 by EISNER::BURNET "John Burnet" > > -< Random notes about data-structure alignment and VAX C >- > .....Second, store data elements in each >structure in decreasing order of size. Yes, John, that's a simple fool proof procedure ... to which I object strongly. In my opinion, the grouping of data elements should be to make the structure most understandable. Grouping by element size will probably help make life miserable for maintenance programmers. And for what? That's what this thread has failed to establish. These responses largely have been software responses to what is basically a hardware problem. In the PDP-11 (and many other machines as well) alignment was *required*. Not so in the VAX. There is a hardware capability there we are all paying for, but is almost universally argued should not be used: accessibility of most datatypes at any address. What are we gaining, quantitatively, by not using it? Jamie Hanrahan points out that if the datatype crosses a natural boundary in the memory system, two reads are required to access one item. Quite true, if the memory system has natural boundaries. But even so, what does the double read cost? In PDP-8 days, with no cache, no read-ahead, no pipe line, and no SMP, it would have been easy: 1.5 microseconds. It's not so easy anymore. The unused part of the double read may be what's wanted next. (Jamie also points out that certain queue and lock manipulation instructions *require* alignment. That is, I think, an architectural decision to potentially simplify the design of some memory systems. Although interesting, it doesn't have much bearing on application programming.) About the only argument for alignment I've heard is, "It can't hurt." Well, in my opinion, John's suggestion *can* hurt - not in performance perhaps (our topic) but in other ways that shouldn't be ignored. And I would strongly object to a language compiler that arbitrarily aligned data items without advising me, or giving me a chance to overrule. Modern languages like COBOL and Fortran have "reference modification" which permits easy access to data structures at any variable byte address and for any variable field length, just like MACRO. To have the compiler introduce synchronizing bytes behind my back is a sure way to introduce bugs. So, until I see firm performance data, I claim that aligning data on a VAX is somewhat like throwing spilled salt over your left shoulder. It can't hurt, can it? ================================================================================ Note 366.11 VAX Address Alignment for Speed 11 of 31 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 13 lines 30-MAY-1989 13:05 -< For the real nitty-gritty... >- -------------------------------------------------------------------------------- >> Jamie points out that some memories return eight byte chunks. >> If the chunks are aligned on quadword boundaries, that's the >> memory system designer's choice. There's no physical reason >> why semiconductor (as opposed to core) memory reads can't start >> at any byte address. There are, however, (last time I looked) >> pretty good reasons for most cache schemes to align on some >> particular boundary. For some really good reading on the design of main memory and cache, get issue #7 of the Digital Technical Journal (EY-6742E-DP). It has several excellent articles on the design of the 6200 and 3600 systems, including one on the memory and cache sub-system. ================================================================================ Note 366.12 VAX Address Alignment for Speed 12 of 31 EISNER::BURNET "John Burnet" 29 lines 30-MAY-1989 18:38 -< Premature alignment the root of all evil? >- -------------------------------------------------------------------------------- > < Note 366.10 by EISNER::HARVEY "Jack Harvey" > > -< Datatype Alignment Viewed as Harmful >- > Yes, John, that's a simple fool proof procedure ... to > which I object strongly. > > In my opinion, the grouping of data elements should be > to make the structure most understandable. Well, er, ah (blush)... I confess that I myself wouldn't (and don't) align structure elements based on size, so I guess I'm guilty of giving bogus advice. Suggestion, really. Not that I ever practice what I preach. ;-} I quite agree with you regarding readability, maintainability, and all the other "abilities" that are important in software development. When it comes to questions like "To align or not to align?", my attitude has usually been "who cares -- if it makes the programs easier to understand and modify, I'll take the performance hit, if there is one." > (Jamie also points out that certain queue and lock > manipulation instructions *require* alignment. That is, I > think, an architectural decision to potentially simplify the > design of some memory systems. Although interesting, it > doesn't have much bearing on application programming.) Hmmm... I don't know. Now that we (in the VMS world) are starting to get heavily into multiprocessing, I think more and more application programs are going to be taking account of system-level details like interlocked instructions. Or would any program that gets that down-and-dirty with the hardware automatically be classed as systems software? ================================================================================ Note 366.13 VAX Address Alignment for Speed 13 of 31 EISNER::DELARISCH "Arnold S. De Larisch" 19 lines 30-MAY-1989 23:32 -< New RISC chips FORCE QUAD-WORD ALIGNMENT!!! >- -------------------------------------------------------------------------------- Folks, After 'evaluating' a couple of High Performance Workstation's which use the MIPS chip (Silicon Graphics 120GTXB and Ardent Titan II) with their VAX FORTRAN compatible compilers ... let be just say ... if your data structure aren't QUAD WORD ALIGNED ... you are in a heep of trouble! The new RISC chips REQUIRE quad-word alignment. If you force the issue, they will not complie code! The best of the compilers (from SGI) gives you the option of putting an 'error' handler to trap BUS ALIGNMENT errors and have the code 'fix' up the data structures into something usable. Hence, if you are gonna write code which may have to be later ported to this 'technology' you are FAR better off doing QUAD WORD allignment up front! -Arnold Still pulling out my hair trying to get unaligned VAX code to run on the MIPS boxes! ================================================================================ Note 366.14 VAX Address Alignment for Speed 14 of 31 EISNER::HARVEY "Jack Harvey" 39 lines 30-MAY-1989 23:34 -< Alignment has a Down Side >- -------------------------------------------------------------------------------- > ...so I guess I'm guilty of giving bogus advice. Not guilty, John. 366.8 merely gave a procedure to permit alignment without wasting space. It didn't advocate using it. Important information not to be omitted. My reaction was a bit noisy. I finally got my library unpacked from my recent relocation. The following are quotes from the original VAX bible: "VAX 11/780 Architecture Handbook, Volume 1, 1977-78." This is the one with the gray haired couple on the cover. Page 2-5. "VAX-11/780 instructions and data are variable length. They need not be aligned on longword boundaries in physical memory, but may begin at any byte address (odd or even)." ... "The advantage of byte alignment is that instruction streams and data structures can be stored in much less physical memory." Page 2-7, section 2.7.1, Memory Cache. "The memory cache also provides 32 bits of lookahead. On a cache miss, 64 bits are read from main memory - 32 bits to satisfy the miss and 32 bits of lookahead." Well, now we begin to see a way that datatype alignment by inserting unused bytes *can* reduce performance. Those unused bytes are *never* of any interest. If the memory system reads them into a buffer on a lookahead in the forlorn hope they will be wanted, the software .align directive has defeated the hardware attempt to improve access time... Please note that I'm not mad at the .align directive itself. It can be vital when writing data to a tape for reading on more primitive machines. I'm just wondering out loud if it has a performance advantage within the VAX. ================================================================================ Note 366.15 VAX Address Alignment for Speed 15 of 31 EISNER::HARVEY "Jack Harvey" 27 lines 31-MAY-1989 19:31 -< Should Instructions be Aligned? >- -------------------------------------------------------------------------------- While staring at some machine code under the debugger today, it dawned on me that about half of the instructions started on an odd address. Of course. That's the way the VAX instruction set works. Now considering that something approximating one half of memory references are for instruction fetches, you might think the VAX architecture designers would have insisted on instructions with an even number if bytes, if it would improve performance. Bell and Co. chose not to do that. Did they permit instructions on odd boundaries because they figured it didn't matter, or, as some people might say, they didn't mind the VAX being slow? Arnold raises an interesting point with the whizzer MIPS chips that insist on data on quad word boundaries. Arnold, are the instructions on quad word boundaries also? Does RISC mean quad word alignment? Are we about to start a new myth that sticking NOPs between instructions to make them start on quad word boundaries is the secret to super fast VAXen? Yes, let's do that! Who will be the first to give us a set of macros that will insure all non-NOP instructions start on quad word boundaries. (Arnold, I'm not making fun of your note, which is valuable information. I'm just taking the lead and running with it.) ================================================================================ Note 366.16 VAX Address Alignment for Speed 16 of 31 EISNER::KILGALLEN "Larry Kilgallen" 5 lines 31-MAY-1989 22:31 -< VAX instruction alignment is not necessary >- -------------------------------------------------------------------------------- Unlike data, instructions are usually accessed in order. I believe that all VAX processors will "save" the left-over bytes not used from the instruction stream for use in the next instruction fetch. Fancy processors have "pipelining" to actually do some preparation on the instructions before the actual operations. ================================================================================ Note 366.17 VAX Address Alignment for Speed 17 of 31 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 42 lines 31-MAY-1989 23:37 -< Data misalignment considered harmful. >- -------------------------------------------------------------------------------- As has been pointed out, testing for the effects of data [mis]alignemnt on a machine with a cache is [literally] a chancy proposition. I haven't run any tests myself. I can tell you that "align data on natural boundaries" is one of the efficiency hints given in various courses presented by DEC Ed. Svcs. I heard it from Lee Smith, who was literally the very first instructor to teach a VMS course, and who helped write many of the early course materials. He in turn had access to many of the architects of VMS and VAX. I have to assume that he knew what he was talking about. Also, there is ample evidence that DEC thoroughly believes in the benefits of both data and instruction alignment. Close examination of the code generated by the C compiler shows it inserting "useless" instructions (ie unreachable ones) with the effect of aligning a subsequent instruction (one that's always a branch target) on a longword boundary. Therefore when the VAX does a fetch for the first opcode at the branch target, the entire fetch obtains usable code. Also -- *outside of data structures which force a paticular ordering of elements*, all of the compilers I've looked at take some pains to align data on "natural" boundaries. They do this very simply, by starting out on an appropriate boundary and specifying the longwords (or quads, iif any) first, etc. In a previous note it was asserted that the insertion of "garbage" bytes in a structure for data alignment purposes buys nothing, because the garbage bytes will still be fetched and therefore will waste part of a fetch cycle. This is true only if the structure is accessed serially from beginning to end -- if your references "hop around" in the structure, and if you have a bunch of different data types, a bit of effort to arrange things so that all the data is "naturally" aligned will help a bit. Evidence? Look at all the data structure diagrams in the IDSM and in the back of the Device Support manual. Look at how all of the longwords and words line up so prettily, sometimes with the aid of an "unused" byte or word here or there. I seriously doubt that VMS Engineering would have gone to all that trouble if it didn't make any difference. Nor do I find that the structures in question, or the code that references them, are any harder to understand because of the alignment of data (indeed, if anything, proper alignment makes the diagrams easier to read). ================================================================================ Note 366.18 VAX Address Alignment for Speed 18 of 31 EISNER::KATZ "Mark Katz" 19 lines 1-JUN-1989 14:42 -< RISC machines are whole other animal >- -------------------------------------------------------------------------------- > an interesting point with the whizzer MIPS chips that > insist on data on quad word boundaries. Arnold, are the > instructions on quad word boundaries also? > Does RISC mean quad word alignment? I would suspect that this is irrelevant. The compiler for the RISC machines produces code in the RISC machine's machine language which is or isn't aligned in some way depending on the characteristics of the architecture, instruction set and compiler sophistication. I wouldn't expect the RISC machine's instructions to look anything like VAX instructions, so their size, alignment, look,feel,nature, etc. would be all different. The data structures, on the other hand, have a prayer of looking very similar from machine to machine due to the relatively low-level description we supply in our programming languages. Of course, since I don't have any RISC machines and haven't been reading that much about them, I may be all wet. ================================================================================ Note 366.19 VAX Address Alignment for Speed 19 of 31 EISNER::HARVEY "Jack Harvey" 26 lines 1-JUN-1989 22:50 -< Compiler Writers *Enjoy* Aligning Data >- -------------------------------------------------------------------------------- >> Does RISC mean quad word alignment? > > I would suspect that this is irrelevant. Ridiculous is a better word, Mark. (I find I should really use smiley faces more often. I guess my attempts at humor really aren't that funny.) I'm afraid people reading Arnold's .13 might leap to the conclusion (which Arnold did *not* put forward) that alignment is the secret of the performance of RISC machines. Jamie is a recognized expert we all have much respect for, and the points he makes in .17 and earlier notes in this thread may all be correct. But this is bad science. His argument in .17, over simplified, seems to me to be that since everyone says alignment is important, it must be so. The history of science and engineering is shot full of this sort of mass mis-think. And data alignment, like using lightning rods, could simply be advice given by "experts" to students eager for hints on how to become "experts". No one ever asks, "But how *much* does it help?" Apparently no one is concerned enough to even run a test. I had hoped that some data would come forward here. Alas, we have become philosophers. <:-( ================================================================================ Note 366.20 VAX Address Alignment for Speed 20 of 31 EISNER::TERRAZAS "Mike Terrazas" 35 lines 2-JUN-1989 09:48 -< Not so ridiculous >- -------------------------------------------------------------------------------- >>> Does RISC mean quad word alignment? >> >> I would suspect that this is irrelevant. > > Ridiculous is a better word, Mark. (I find I should really > use smiley faces more often. I guess my attempts at humor > really aren't that funny.) I'm afraid people reading > Arnold's .13 might leap to the conclusion (which Arnold did > *not* put forward) that alignment is the secret of the > performance of RISC machines. Not really so ridiculous. It's not the secret of their performance, but usually an artifact of the simplified instruction set (trying to get all instructions to execute in one processor cycle) that causes most RISC architectures to require eight-byte alignment of the more difficult data types, such as floating point. I have dealt with 3 different RISC chips, _*ALL*_ of them required floats inside of data structures to be aligned on eight-byte boundaries in order to take advantage of compiler optimizations. If you override the built-in compiler interlocks to prevent optimizations, you will get memory bus errors (which on two of the 3 machines cause a general hardware fault which, of course takes the whole machine south.) When you force compressed data structures (for data compatibility with the VAX, for example,) the compilers all handle it by doing byte-by-byte moves of the floating value to an aligned location for every reference to the element of the structure. As a result, performance goes out the window. One of the compilers has an optimization pass which recognizes this and does the transfer to temp store once and then replaces all references with that location, adding transfer back after each write. It's somewhat faster, but only marginally for our applications. > I had hoped that some data would come forward here. Alas, > we have become philosophers. <:-( Socrates would think that puts us nearer our eternal rest in Elysian. ;-) ================================================================================ Note 366.21 VAX Address Alignment for Speed 21 of 31 EISNER::KILGALLEN "Larry Kilgallen" 15 lines 2-JUN-1989 18:08 -< Experiments are for skeptics >- -------------------------------------------------------------------------------- > students eager for hints on how to become "experts". No one > ever asks, "But how *much* does it help?" > > Apparently no one is concerned enough to even run a test. Those of us who "know" in our heart-of-hearts that data alignment helps are supporting the conventional wisdom. Those who seek fame and glory by disproving it should be the ones to experiment. I have no doubt that you can come up with an example where it does not help, as I can come up with one where it helps a lot. I doubt, however, that you can come up with one where it hurts performance, so until you disprove my stubborn assumptions, I will continue to align my structure definitions out of habit. I realize there will be applications where that does not help me, but when one comes up where it does matter, my habits will lead me in the right direction. ================================================================================ Note 366.22 VAX Address Alignment for Speed 22 of 31 EISNER::HARVEY "Jack Harvey" 33 lines 2-JUN-1989 21:01 -< Challenge Accepted >- -------------------------------------------------------------------------------- > Those who seek fame and glory by disproving it should be the > ones to experiment. I have no doubt that you can come up with > an example where it does not help, as I can come up with one > where it helps a lot. Glory rejected. Fame not possible. Rather than get into an endless debate about what is a valid test, I suggest we attempt, jointly, to find what is the most god-awful case where alignment will really save the bacon. A max max worst case, where you are an absolute dummy not to align the data. If a consensus happens, or even a likely candidate baddy for worst case appears, I will attempt to test it. I'll publish the code here in fragments to give the gist, and in executable form on Pageswapper, so others can cross check me. And also give us their results. If we have a max max worst case result on a few machines, then we might leave it to individual interests to judge if the probability of occurrence of such a case (and of course easily understandable variations of likely occurrence) then our community might have a quantitative handle on this. Until this happens, we will remain philosophers. Now, if you accept this role, perhaps you can consider what would be a significant degradation in performance before you start yelling at your programmers to, "Dammit, align your data!" Or conversely, "Please don't waste your time and my money worrying about that." [I have been a bad Noter in this thread in forgetting to mark my humor and sarcasm with appropriate smiley faces. This note has no smileys. I am serious.] ================================================================================ Note 366.23 VAX Address Alignment for Speed 23 of 31 EISNER::KILGALLEN "Larry Kilgallen" 4 lines 3-JUN-1989 00:28 -< My suggestion >- -------------------------------------------------------------------------------- I would nominate successive reads of alternate quadwords from an array, backwards. The unaligned version should be off by one byte. Obviously multiple passes are in order to swamp startup effects without incurring pagefaults (you did say worst case). ================================================================================ Note 366.24 VAX Address Alignment for Speed 24 of 31 EISNER::GLEASON "CyberPunk" 5 lines 3-JUN-1989 19:12 -< The Socratic Method >- -------------------------------------------------------------------------------- >Socrates would think that puts us nearer our eternal rest in Elysian. ;-) Yeah, and look what happened to him... ================================================================================ Note 366.25 VAX Address Alignment for Speed 25 of 31 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 40 lines 4-JUN-1989 03:06 -< Yet more philosophy :-) >- -------------------------------------------------------------------------------- > the points [Jamie] makes in .17 and earlier notes in this > thread may all be correct. But this is bad science. His > argument in .17, over simplified, seems to me to be that > since everyone says alignment is important, it must be so. I never claimed it was "science", good or otherwise, so labelling it as "bad science" is a bit of a misnomer, don't you think? And I thought I made it clear that I was *not* simply restating "popular wisdom": My source [Lee Smith] had direct access to the folks who designed the VAX-11/780 and who got the first versions of VMS running on it. > The history of science and engineering is shot full of this > sort of mass mis-think. And data alignment, like using > lightning rods, could simply be advice given by "experts" to > students eager for hints on how to become "experts". No one > ever asks, "But how *much* does it help?" You picked a bad analogy; any insurance adjuster will tell you that lightning rods help a LOT. > Apparently no one is concerned enough to even run a test. I don't feel I need to run a test to see that misaligned data can result in two memory accesses where one would have served, and that two accesses will take longer than one. Of course, if you work mostly with code that spends most of its time waiting for I/O to complete, you may never notice the difference. I work mostly with code that responds to interrupts and tries to keep I/O devices as busy as possible, and, yes, I do worry about microseconds. > I had hoped that some data would come forward here. Alas, > we have become philosophers. <:-( Guilty [so far], but I hope you're wearing the same hat as the rest of us. After all, Jack, it was you who raised the question in the first place -- and on the basis of speculation rather than test results. But speaking of test results... see the next note. ================================================================================ Note 366.26 VAX Address Alignment for Speed 26 of 31 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.co" 134 lines 4-JUN-1989 03:54 -< Test results for data [mis]alignment >- -------------------------------------------------------------------------------- All right. I cobbled up a dozen different permutations of test programs, all based on Larry Kilgallen's suggestions a few notes back. Tests were run on an 8200. In order to eliminate the effects of the cache, I not only walked backwards through the arrays; I also turned off the cache. (This is just MTPR #1, #PR$_CADR in kernel mode. To turn it back on, move a 0 instead. If you don't know how to use the $CMKRNL system service you can poke the CADR from the console.) The differences between the permutations of programs were: Data type -- quadword vs. longword Alignment -- aligned or off-by-one-byte or no operation Operation -- write (CLRx) or read-write (MOVx with both operands in memory, both operands using same alignment) (Timings for "operation - none" were done so that the time required for the instruction could be isolated from the loop iteration, etc. It may look odd for "no operation" to be an option under "Alignment", but there seemed no point in running separate "noop" tests with the data aligned or not when we weren't touching it anyway, and different "operations" used different iteration counts, so this is the way it happened to fall out.) Here are the results. All times are seconds. The first test used ten "outer" iterations, each "inner" iteration hitting 50,000 different array elements. Every program was run five times and the results averaged to get the values shown here. Data type: Quadword; Operation: Write (instruction: CLRQ X[R0]) Misaligned Aligned Noop 12.03 10.42 4.36 With the data misaligned, the program took 15% longer (using "aligned" as the base). Subtracting noop times from the other runs, we get the following for the CLRQ instruction alone: 7.67 6.06 Misalignment caused the CLRQ instruction to take 26% longer. In all subsequent tests the inner loop was cut to a tenth its previous length (ie the instruction being tested was executed 5000 times per "outer" loop). This was done because the addition of the second array for the read-write test caused the program to blow the working set, resulting in huge numbers of page faults. Run times were still long enough and consistent enough to suit me. :-) Data type: Quadword; Operation: Read-write (instruction: MOVQ X[R0], Y[R0] Misaligned Aligned Noop 2.11 1.64 0.44 The program took 22% longer with the data misaligned. After subtracting noop time: 1.67 1.20 The instruction took 39% longer with the data misaligned. Data type: Longword; Operation: Write (instruction: CLRL X[R0] Misaligned Aligned Noop 0.95 0.87 0.44 The program took 9% longer with the data misaligned. After subtracting noop time: 0.51 0.43 The instruction took 19% longer with the data misaligned. Data type: Longword; Operation: Read-write (instruction: MOVL X[R0], Y[R0] Misaligned Aligned Noop 1.47 1.24 0.44 The program took 18% longer with the data misaligned. After subtracting noop time: 1.03 0.80 The instruction took 29% longer with the data misaligned. I also ran some tests where the program was reading successive (not alternate) longwords, stepping forward through the array. As Jack pointed out earlier, the cache would reduce the importance of data alignment, since prefetch would tend to keep the next operand available in the cache. And indeed, with the cache off, results were consistent with the "longword write" case above, but with the cache on there was almost no difference between the aligned and misaligned cases (with the data misaligned the program took 3% longer to run). (And the program took about three times as long with the cache off as with it on. Make DAMN sure you turn the cache back on when you're done testing!) Discussion: Clearly, for individual instructions where cache hits do not occur, the difference between aligned and misaligned data IS significant. Now the question is, how much does it matter in actual programs? Well, that will depend on how often you're referencing the misaligned data and on whether it can be counted upon to stay in the cache. Measuring this would be VERY tricky. Among other difficulties, it would be dependent on what else was running in the system: THe memory cache is system-wide, and if lots of other processes are being scheduled to run, the odds of finding your data in the cache when it's your turn again will be reduced. For my purposes, I'm satisfied that data alignment is worth the EXTREMELY trivial effort required. More speculation: It occurs to me that, since the VAX cache is organized like memory in that it's either longwords or quadwords (not individual bytes) that are cached, misaligned data can reduce the effective size of the cache: A datum that unnecessarily spans a "natural" memory unit will take two "cache slots" where one would have served. ================================================================================ Note 366.27 VAX Address Alignment for Speed 27 of 31 EISNER::HARVEY "Jack Harvey" 16 lines 4-JUN-1989 11:13 -< Good Science >- -------------------------------------------------------------------------------- >For my purposes, I'm satisfied that data alignment is worth the >EXTREMELY trivial effort required. Jamie, you just re-earned that gorgeous jacket with the DECUServe artwork on the back you were wearing on Thursday in Atlanta. (Congratulations, by the way.) In my opinion, the previous note is good science. (Still think it's a misnomer?) You gave your test procedures, summarized the results and discussed them. Others may now confirm and extend your results, building on what you have started. These results, by the way, are the *first* I have ever seen on this subject. One of us, at least, need not wear the philosopher hat. He has a wizard jacket instead. ================================================================================ Note 366.28 VAX Address Alignment for Speed 28 of 31 EISNER::KOZAM 12 lines 4-JUN-1989 12:32 -< WoW! >- -------------------------------------------------------------------------------- Wow! Thanks, Jamie. That's a great bit of work. I'd say that the results look pretty definitive. I wonder if other VAX implementations are similar. > One of us, at least, need not wear the philosopher hat. He > has a wizard jacket instead. This is an interesting comment from a historical prospective. For centuries, "science" was run by philosophers who basically sat around discussing the world. Experimental science is relatively new. ================================================================================ Note 366.29 VAX Address Alignment for Speed 29 of 31 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 30 lines 5-JUN-1989 16:39 -< More test results >- -------------------------------------------------------------------------------- Since it isn't clear that turning off the cache is equivalent to running with the cache but not getting cache hits, I ran one set of tests with the cache on. Here are the results, together with the corresponding cache-off results for comparison. Data type: Longword; Operation: Read-write (instruction: MOVL X[R0], Y[R0] --- Cache off ---------------------- --- Cache on ---------------------- Misaligned Aligned Noop Misaligned Aligned Noop 1.47 1.24 0.44 0.47 0.39 0.12 The program took 18% longer with the The program took 20% longer with the data misaligned. data misaligned. After subtracting noop time: After subtracting noop time: 1.03 0.80 0.35 0.27 The instruction took 29% longer with The instruction took 30% longer with the data misaligned. the data misaligned. Discussion: The "cache-on" results show the expected speedup in execution that comes from never having to go to memory to get executable code, but the difference between the aligned and nonaligned cases remains remarkably consistent. (Given these short run times, a percentage point or two is well within experimental error.) ================================================================================ Note 366.30 VAX Address Alignment for Speed 30 of 31 EISNER::LEDERMAN "Bart Z. Lederman" 30 lines 6-JUN-1989 07:58 -< Can we test branch destination effects? >- -------------------------------------------------------------------------------- I find the information and test methods very interesting. One factor which I think was discussed was the result of alignment on the destination of a branch operation: something a lot of the compilers appear to do. Can anyone think of a good way to test the effect of alignment here? Judging by the results posted, it would appear that it would be good for the destination of a branch (or, I suppose, an .ENTRY point) to be aligned as the memory read would pick up the largest amount of memory following the 'branched to' instruction, and should therefore execute the following instructions faster (or fill the pipeline faster). Unfortunately, the only way I can think of to test the effect is with a program that would look something like this: BR 40$ .BLKB 17 10$: BR 60$ .BLKB 17 20$: BR 50$ .BLKB 17 30$: BR 70$ .BLKB 17 40$: BR 10$ .BLKB 17 50$: BR 30$ .BLKB 17 60$: BR 20$ etc., or some such convoluted sequence, with or without padding between branches; and with a second version which has .ALIGN LONG in front of every label. ================================================================================ Note 366.31 VAX Address Alignment for Speed 31 of 31 EISNER::HARVEY "Jack Harvey" 36 lines 15-JUN-1989 05:39 -< Tests and Evaluations >- -------------------------------------------------------------------------------- The results and discussion so far have been useful and interesting. I've been able to get in some weekend tests also, generally confirming Jamie's results. Jamie may be able to give me more specific code from his tests that I can incorporate. One goal is a generalized machine timing procedure that can be used to easily evaluate any instruction or series of instructions on any machine and with other users active. I'm using LIB$INIT_TIMER and its related RTL routines to determine process CPU time. Since the RTL is a software issue, I've started topic xxx in VMS to ask about accuracy. Another goal is to attempt *valid* alignment effect measurements with cache enabled, since that's the way real programs run. I'm not questioning Jamie's tests with it disabled; I'm simply looking for a way around (if possible) this worry. Any ideas? The crudely rounded 25% slowdown due to non-alignment could be interpreted as turning a 785 into a 780, or a 780 into a MicroVAX II, or a 6240 into a 6230. But as Jamie pointed out, using the direct results of his specific instruction tests this way ignores that a real program uses many other instructions than the ones he tested. If software generally uses nine instructions not affected by alignment (CMPL #3659,R0 ;for example) for every one that is, then the overall slowdown is only a few percent. Take a look at some of your machine code. Of course, if you are in the number crunching business and invert arrays at the drop of a hat, then you may have software with predominately memory reference instructions. If so, you might be able to shorten a 100 hour job down to 70 or 80 hours by aligning the data. Do it.  ================================================================================ Note 367.0 MicroVAX I to II Conversion 25 replies EISNER::RENES "Stuart Renes, DFWLug Chair" 17 lines 24-MAY-1989 13:44 -------------------------------------------------------------------------------- I'm in need of infomation relating to conversion of old MicroVax I processors to MicroVAX II's. Didn't DEC once sell update kits? I presume you would also need to upgrade the disk so you could run VMS 5? I don't remember what size disk the old MicroVax I's had but I'm sure that they weren't big enuf! Anyone done this conversion? I also presume that if DEC once sold the kits, they might be available on the used market as well. Thanks in advance. Stuart. ================================================================================ Note 367.1 MicroVAX I to II Conversion 1 of 25 EISNER::HARVEY "Jack Harvey" 28 lines 24-MAY-1989 22:31 -< Check DECdirect for MicroVAX I Upgrade >- -------------------------------------------------------------------------------- Yes indeed, DEC sold an MV I to MV II upgrade kit. It was in DECdirect for several issues. (I'm still not unpacked from my recent move so I can't check which.) Since it seems to be DEC's policy to stamp out MicroVAX I's, I expect it's still available. The kit consisted of the new CPU board and cabinet kit for the console port and the MVII LED boot diagnostic, some MVII memory, an RQDX3 controller, a TK50 and controller, diagnostic tape, and an MVII license. There were options. I think it was available with an Ultrix license, and possibly no TK50. The cost is about $18,000 list. But that is less than the sum of the parts cost, so the kit is *very* tightly controlled. For example, you must *return* the old MV I cpu board, so that you can't just buy a few hundred kits and sell the parts on the street at a profit. I've never heard of a "used" market for this kit. It gets expended in the process of upgrading a system. You then have a used MVII system available. But the kit is gone. There was also a kit (from DECdirect) for upgrading a PDP-11/73 to a MicroVAX II - more bucks, of course. ================================================================================ Note 367.2 MicroVAX I to II Conversion 2 of 25 EISNER::JENSEN_M 6 lines 25-MAY-1989 01:36 -< Buy a used MVII >- -------------------------------------------------------------------------------- If it were me, I wouldn't worry about upgrading an MVI to II. The used market for MVII's has plummeted. It is possible to buy a BA23 with CPU for less than $5K. One source is Midwest systems in Minnesota. (no connection, monetary or otherwise). ================================================================================ Note 367.3 MicroVAX I to II Conversion 3 of 25 EISNER::RICE "Gary Rice" 11 lines 25-MAY-1989 09:55 -------------------------------------------------------------------------------- > ...market for MVII's has plummeted. It is possible to buy a BA23 with CPU > for less than $5K. That may be true, but a CPU in a cabinet is hardly a system. By the time you add memory, disk (& controller), and a tape drive (to load the thing with), not to mention a Terminal controller (and likely an Ethernet card), your system will STILL go out the door for more than $11K, and THAT assumes you buy the pieces from a whole range of vendors. Gary ================================================================================ Note 367.4 MicroVAX I to II Conversion 4 of 25 EISNER::JENSEN_M 9 lines 25-MAY-1989 13:01 -< Wheres my calculator ? >- -------------------------------------------------------------------------------- The thought on the MVII was that if somebody has an MVI, he probably has a disk controller and Ethernet board (the dreaded DEQNA). Only thing left to get is memory. Used to be able to get 8MEG for 1500 (lowest was 975 for us). If he needed tape for the MVII he probably needed it for the MVI. Still sounds a bit less than $11K. Mike ================================================================================ Note 367.5 MicroVAX I to II Conversion 5 of 25 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 8 lines 25-MAY-1989 15:59 -< SOFTWARE >- -------------------------------------------------------------------------------- > The thought on the MVII was that if somebody has an MVI, he probably has > a disk controller and Ethernet board (the dreaded DEQNA). Only thing left > to get is memory. Used to be able to get 8MEG for 1500 (lowest was 975 for > us). If he needed tape for the MVII he probably needed it for the MVI. > Still sounds a bit less than $11K. One small item VMS. The upgrade kit comes with an 8 user VMS license. ================================================================================ Note 367.6 MicroVAX I to II Conversion 6 of 25 EISNER::HUDGINS "Jerry Hudgins" 14 lines 25-MAY-1989 16:03 -< The MV-I was rather primitive... >- -------------------------------------------------------------------------------- > The thought on the MVII was that if somebody has an MVI, he probably has > a disk controller and Ethernet board (the dreaded DEQNA). Bear in mind that most MV-I's were probably shipped with an RQDX1 or RQDX2 and an RD52, and will probably need to be upgraded to an RQDX3 and an RD53 or RD54 to be useful with VMS. The upgrade kit did include at least the RQDX3, I think, and maybe an RD53. > If he needed tape for the MVII he probably needed it for the MVI. The MV-I predated the TK50, I believe. Most (if not all) were shipped with RX50's. I have one with this configuration (RQDX1/RD52/RX50). I don't think I've ever seen one with a TK50. Anyone else? ================================================================================ Note 367.7 MicroVAX I to II Conversion 7 of 25 EISNER::RENES "Stuart Renes, DFWLug Chair" 13 lines 25-MAY-1989 18:39 -< RD53 included? >- -------------------------------------------------------------------------------- > Bear in mind that most MV-I's were probably shipped with an RQDX1 > or RQDX2 and an RD52, and will probably need to be upgraded to an > RQDX3 and an RD53 or RD54 to be useful with VMS. The upgrade kit > did include at least the RQDX3, I think, and maybe an RD53. Good question - does the upgrade kit contain an RD53 and if so, is that large enuf for VMS 5? Stuart. (slighly less puzzled than before!) ================================================================================ Note 367.8 MicroVAX I to II Conversion 8 of 25 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 18 lines 25-MAY-1989 19:05 -< 630XR-HA >- -------------------------------------------------------------------------------- The MVI to MVII upgrade kit (630XR-HA) contains... RQDX3-M RD CTL QZ002-C5 8 USER LIC QZ002-H5 TK50 KIT TQK50-AB TK50 CTL TK50-DA DESK TOP TK50 KA630-AA MVII CPU CARD CK-KA630-AB CAB KIT FOR MVII CARD MS630-CA 8 MB ZNAAA-C5 HARDWARE DOC ================================================================================ Note 367.9 MicroVAX I to II Conversion 9 of 25 EISNER::HARVEY "Jack Harvey" 17 lines 25-MAY-1989 22:28 -< RD53 is Big Enough >- -------------------------------------------------------------------------------- > Good question - does the upgrade kit contain an RD53 and if so, > is that large enuf for VMS 5? Jeff's note confirms that the RD53 is not part of the kit. I have put V5.0-2 on an RD53 and had quite a significant amount of disk space left. I'll try to post the number of blocks for a basic V5.1 system with expanded libraries tomorrow. Putting all of DECwindows (on the V5.1 release tape) on an RD53 with V5.1 VMS is probably not possible. (Haven't tried it.) Be advised, that DEC warned that they expect VMS to keep getting bigger. Shoot for an RD54 equivalent if possible, for well under $2,000. RD53's are available used for a few hundred bucks. ================================================================================ Note 367.10 MicroVAX I to II Conversion 10 of 25 EISNER::HARVEY "Jack Harvey" 13 lines 26-MAY-1989 11:45 -< VMS V5.1 on an RD53 - see 367.17 also >- -------------------------------------------------------------------------------- Some numbers: Storage for VMS V5.1 with expanded libraries - 90,439 blocks. Capacity of an RD53 - 138,672 blocks Capacity of an RD54 - 311,200 blocks So, after V5.1 is installed on an RD53, you have 48,233 blocks (24.7 megabytes) free. ================================================================================ Note 367.11 MicroVAX I to II Conversion 11 of 25 EISNER::ROBERTS_D "don roberts #241" 17 lines 26-MAY-1989 13:11 -< Doesn't match my numbers >- -------------------------------------------------------------------------------- > Storage for VMS V5.1 with expanded libraries - 90,439 > blocks. > Capacity of an RD54 - 311,200 blocks I have just installed 5.1-1 on a uVAX II with an RD54, taking all default sysgen parameters (except SCSNODE) and decompressing the libraries. With NOTHING else on the disk, performing all purges along the way and doing all the things it said to do in the installation manual and NOT installing DECwindows, SHOW DEVICE DUA0 gives 201K free blocks. Therefore, 5.1-1 used 110K blocks. Before decompressing the libraries and at 5.1, I had 210K free blocks, or 101K blocks. Which of the above numbers is wrong? If 5.1 with expanded libraries is 90K blocks, where did my additional 11K blocks go? I did the purges and analyse/disk/repairs where installation notes requested. This would only leave 28K blocks on the RD53 at 5.1-1 ================================================================================ Note 367.12 MicroVAX I to II Conversion 12 of 25 EISNER::ROBERTS_D "don roberts #241" 4 lines 26-MAY-1989 13:14 -< More questions on the numbers >- -------------------------------------------------------------------------------- > So, after V5.1 is installed on an RD53, you have 48,233 > blocks (24.7 megabytes) free. Not to quibble, but wouldn't that be 24.11 Mb free? ================================================================================ Note 367.13 MicroVAX I to II Conversion 13 of 25 EISNER::RENES "Stuart Renes, DFWLug Chair" 10 lines 26-MAY-1989 16:36 -< Thanks to all! >- -------------------------------------------------------------------------------- Thanks to all who replied to my original .0 After looking at the used market, it may be cheaper in the long run to buy a used MicroVAX 2000 system, complete with 11 ports, 4 mb of memory, a TK50 and RD54, VMS and DECnet licenses for around $13k! Stuart. ================================================================================ Note 367.14 MicroVAX I to II Conversion 14 of 25 EISNER::HARVEY "Jack Harvey" 25 lines 26-MAY-1989 21:55 -< In Our World, 1024 = M/B >- -------------------------------------------------------------------------------- > > So, after V5.1 is installed on an RD53, you have 48,233 > > blocks (24.7 megabytes) free. > > Not to quibble, but wouldn't that be 24.11 Mb free? Don, I judge your quibble to be fact: 48,233 * 512 / 1000 = 24,695.296 bytes ( ~=24.7 mb) however... 48,233 * 512 / 1024 = 24,116.5, by the (in my opinion) accepted 1024 multiplier for K and M memory designations. Factual quibbles are not quibbles. Oxymorons? No. Now, gang, how many decimal bytes in a gigabyte of disk memory??? We are talking G here. (Real Question. I got it asked one day ago by a person with a checkbook.) Any reasonable engineer will agree that 1000 == 1024. But the bean counters must be recognized. And 1024/1000 is a useful fiddle factor among us engineers. (Shut UP!) ================================================================================ Note 367.15 MicroVAX I to II Conversion 15 of 25 EISNER::MCCORMICK_B "Virtual Bob" 7 lines 27-MAY-1989 17:39 -< Does calc inc page/swap/dump files? >- -------------------------------------------------------------------------------- For those who are stating sizes of VMS on disk: Did your calculations consider or exclude the memory management files (PAGEFILE & SWAPFILE) ... and how about a dump file? Just wondering ... they can be a significant amount of disk space on a 16mb MicroVAX II with RDxx for anything more than a couple hands full of processes. ================================================================================ Note 367.16 MicroVAX I to II Conversion 16 of 25 EISNER::HARVEY "Jack Harvey" 11 lines 27-MAY-1989 19:31 -< Page/Swap/Dump Files Included >- -------------------------------------------------------------------------------- > -< Does calc inc page/swap/dump files? >- Good point, Bob. My figure in .10 *includes* all three, the page, swap and dump files. I'll post their sizes when I get a chance to check them. This is especially important because Autogen varies their sizes depending on the system configuration and so the raw size figure. I did the install on a 4MB MV2000. Don's figure may differ if his MVII has more memory. ================================================================================ Note 367.17 MicroVAX I to II Conversion 17 of 25 EISNER::HARVEY "Jack Harvey" 17 lines 28-MAY-1989 13:51 -< Page/Swap/Sysdump File Sizes >- -------------------------------------------------------------------------------- The V5.1 block requirements described in .10 of this thread include the following: PAGEFILE.SYS 11,200 blocks SWAPFILE.SYS 1,700 SYSDUMP.DMP 10,245 Readers should note that the sizes of these files are established by Autogen and may vary depending on your actual system configuration. In general, the larger the system, the larger these files will be. These were generated for a 4MB MicroVAX 2000. Note also that SYSDUMP.DMP is usually optional. Unless you are into writing your own device drivers or other exec internals, you probably care little about the dump if it crashes... ================================================================================ Note 367.18 MicroVAX I to II Conversion 18 of 25 EISNER::BURNET "John Burnet" 13 lines 30-MAY-1989 18:28 -< Yes, the dump file is optional... why waste the space? >- -------------------------------------------------------------------------------- > Note also that SYSDUMP.DMP is usually optional. Unless you > are into writing your own device drivers or other exec > internals, you probably care little about the dump if it > crashes... This is a good point that small-disk users (that's "users of small disks", not "small users of disks" :-) should take note of. You might consider setting DUMPBUG and SAVEDUMP to 0, renaming SYSDUMP.DMP to something else, REBOOTING YOUR SYSTEM, and then (and only then) deleting SYSDUMP.DMP. That will free up the space previously taken by the dump file. Note that DUMPBUG has to be set to zero using SYSGEN (use current, set dumpbug 0, write current) since AUTOGEN knows nothing about it. ================================================================================ Note 367.19 MicroVAX I to II Conversion 19 of 25 EISNER::HARVEY "Jack Harvey" 4 lines 30-MAY-1989 21:53 -< Sysgen params DUMPBUG and SAVEDUMP >- -------------------------------------------------------------------------------- Thanks for pointing these out. I was not aware of them. A time or two, I've just deleted SYSDUMP.DMP when cramped for space. Wonder what would have happened if the system had crashed? ================================================================================ Note 367.20 MicroVAX I to II Conversion 20 of 25 EISNER::KENNEDY "Terry Kennedy" 9 lines 30-MAY-1989 22:36 -< Your files get trashed... >- -------------------------------------------------------------------------------- > Wonder what would have happened if the system had crashed? The crash file is located at system startup and its starting block number is recorded for later use. The crash dump driver will write the dump to the disk blocks it was told the crash dump file lives at, regardless of anything you have done to the directory or those blocks since. So, the answer is- will write over whatever files you place at those disk blocks. ================================================================================ Note 367.21 MicroVAX I to II Conversion 21 of 25 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 11 lines 31-MAY-1989 04:27 -< Your files ALWAYS get trashed. >- -------------------------------------------------------------------------------- > Thanks for pointing these out. I was not aware of them. > A time or two, I've just deleted SYSDUMP.DMP when cramped > for space. Wonder what would have happened if the system > had crashed? Terry is correct -- your files get trashed. In fact they get trashed even if the system doesn't crash: A small part of the dump file is written even during normal shutdowns. Once you delete the "live" dump file the safest bet is to power off (and, if you have battery backup for your memory, disable it). ================================================================================ Note 367.22 MicroVAX I to II Conversion 22 of 25 EISNER::TILLMAN "Brian Tillman, Smiths Industries." 10 lines 31-MAY-1989 12:29 -< AUTOGEN can do it >- -------------------------------------------------------------------------------- RE: < Note 367.18 by EISNER::BURNET "John Burnet" > -< Yes, the dump file is optional... why waste the space? >- |Note that DUMPBUG has to be set to zero using SYSGEN (use current, set |dumpbug 0, write current) since AUTOGEN knows nothing about it. Perhaps true, but AUTOGEN will set any system parameter you tell it to set, even if it knows nothing about it. I.E., if you put DUMPBUG=1 in your MODPARAMS.DAT, AUTOGEN will change the value of DUMPBUG. ================================================================================ Note 367.23 MicroVAX I to II Conversion 23 of 25 EISNER::MERRIMAN "George Merriman - CCA/NY" 5 lines 31-MAY-1989 16:19 -< VMS discussion should be in VMS conference >- -------------------------------------------------------------------------------- This thread seems to be turning into a VMS tuning discussion. In fact, the VMS conference has numerous notes on page, swap and dump files. Perhaps we should continue there. George Merriman, HARDWARE_HELP co-moderator. ================================================================================ Note 367.24 MicroVAX I to II Conversion 24 of 25 EISNER::BURNET "John Burnet" 9 lines 31-MAY-1989 16:20 -< autogen, dumpbug, savedump, surprise! >- -------------------------------------------------------------------------------- > Perhaps true, but AUTOGEN will set any system parameter you tell > it to set, even if it knows nothing about it. I.E., if you put > DUMPBUG=1 in your MODPARAMS.DAT, AUTOGEN will change the value of > DUMPBUG. Very interesting. Thanks... I had based my (too-hasty) assumption on the fact that "SEARCH SYS$UPDATE:AUTOGEN.COM DUMPBUG" turned up nothing, while searching for SAVEDUMP did turn up something. It's good to know that AUTOGEN is smarter than I had given it credit for. ================================================================================ Note 367.25 MicroVAX I to II Conversion 25 of 25 EISNER::ANGLEY "Kevin Angley" 6 lines 9-JUN-1989 22:18 -< Easy way out >- -------------------------------------------------------------------------------- Yes, deleting SYSDUMP.DMP will trash files, but you could rename it to TEST.TMP, then reboot and delete. Or you could do it the right way with the system parameters as described.~r  ================================================================================ Note 368.0 MicroVAX terminal mux's? 3 replies EISNER::RENES "Stuart Renes, DFWLug Chair" 15 lines 24-MAY-1989 14:00 -------------------------------------------------------------------------------- What is the best way to get more than 8 terminal lines on a MicroVAX II box WITHOUT the use of DECserver technology; i.e the lines should be locally attached. Are there third-party Q-bus terminal mux's that give you more than 8 lines? Can you use more than a single terminal mux on the same Q-bus backplane to accomplish this? Help from a non-microVAX person! Stuart. ================================================================================ Note 368.1 MicroVAX terminal mux's? 1 of 3 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 60 lines 24-MAY-1989 15:02 -< Multiple DHx cards/emulations possible >- -------------------------------------------------------------------------------- > What is the best way to get more than 8 terminal lines on a > MicroVAX II box WITHOUT the use of DECserver technology; i.e > the lines should be locally attached. Use more/bigger muxes. > Are there third-party Q-bus terminal mux's that give you more > than 8 lines? The Emulex CS04 we use supports 64 DHV (like) ports from a quad card. VMS thinks it is 8 DHVs each with private VECs + CSRs. It comprises a CC04 quad card, up to 50' (ours must be at least 30') of 34 wire ribbon cable, and 1 to 4 of their CP34 16 port panels daisy chained on the 34 wire ribbon cable. A terminator plugs into the last panel. The panels are 'smart', and also can provide (on a per port switched basis) EIA 422 for driving LONG lines @38.4, and such tricks as allowing out of band EIA lead flow control NOT seen by DEC's S/W, as well as some weird baud rates you won't believe (or probably ever need). The same panels are/were also used with the unibus CS11 system that (depending on proms) emulated DH11, DV11, or DMF32 (all ports modem control!), or in the unibus CS32 subsystem only emulated DMF32, but allowed 128 ports (8 CP34 panels) per controller card. The older CP11/12 panels (watch for them used @ < $100.) work on the U bus controllers but (darn it) not on the CC04 which only uses the newer CP34s. New, this stuff is MUCH MUCH more $s than getting a NEW DELQA + Databilities or even Emulexes terminal servers. DEC and Xyplex also make servers, but DEC is $s, and Xyplex (unforgivably in my book) treats each 8 port card in their box as a seperate source of ethernet packets. I.E. NO traffic to the same host from their box will share a packet UNLESS it happens to be from the same card. Think of it as a box that holds a lot of DS200s. I am curious why you can't use terminal servers. > Can you use more than a single terminal mux on the same Q-bus > backplane to accomplish this? Sure. 4 Emulex CS04s (each at MAX size) would get you 256 ports. Is that enough? If you are into smaller configs, try Dilog, Emulex, Able, and others for 16 ports (2 DHV or DHQ emulations) per card, with dumb distribution panels. Put several in as needed. The biggest problem is ribbon cable clutter, and rear bulkhead space for the distribution panels. This may only become a serious problem if you are 1) very tidy, and 2) observe FCC regulations. (Intentionally ignoring DEC's MMJ nonsense) If you can function on 3 twisted pair Modtap type distribution, you may be able to get each 8 ports on a 25 pair telco type connector. Failing going directly, Dilog will get you to a DENSELY packed RJ11 type panel, and octopus cords or harmonicas will get you onto 25 pairs from there easily. Last, but NOT LEAST - NEVER EVER EVER get a DZ. You may save a few dollars up front, but you will pay dearly every day thereafter in needless interrupts. ================================================================================ Note 368.2 MicroVAX terminal mux's? 2 of 3 EISNER::HARVEY "Jack Harvey" 10 lines 24-MAY-1989 22:42 -< Use DHQ11 for Eight More Ports >- -------------------------------------------------------------------------------- Use of third party multiplexers is the way to go if you really want to stuff as much as possible on the Q-Bus without going the terminal server route. However, if you just want to add, say, another eight ports, and stay pure DEC, you could put in a DHQ11, which is a dual board and not very expensive (relative to terminal server startup costs). Bulkhead panel space is a problem unless you can use the cabinet kits that put the DB25 RS232 connectors on the end of a cable remote from the system unit. ================================================================================ Note 368.3 MicroVAX terminal mux's? 3 of 3 EISNER::RENES "Stuart Renes, DFWLug Chair" 4 lines 25-MAY-1989 18:36 -< Thanks! >- -------------------------------------------------------------------------------- Thank you all for your help! Stuart :-)  ================================================================================ Note 369.0 TK50s outside DEC - Amiga 2000 via standard SCSI 4 replies EISNER::HASSINGER "Bob Hassinger" 12 lines 25-MAY-1989 10:11 -------------------------------------------------------------------------------- Saw something interesting last night at a meeting of the Worcester Amiga Computer Users Group (WACUG). A TK50 connected to an Amiga 2000. Seemed to be working just fine. It was connected via a normal GVP SCSI controller that also had a hard disk on it. The software was something being developed by the person showing it. It was backing up disks onto TAR format tapes. Incidentally, the economics of this may not be real attractive for the average person but the individual in question happens to work for DEC in the area where TK50s and so on are engineered. He finds being able to move 100Mb at a shot of Amiga files between home and work is very handy. ================================================================================ Note 369.1 TK50s outside DEC - Amiga 2000 via standard SCSI 1 of 4 EISNER::RICE "Gary Rice" 8 lines 26-MAY-1989 10:31 -------------------------------------------------------------------------------- > ... A TK50 connected to an Amiga 2000. > Seemed to be working just fine. It was connected via a normal GVP SCSI > controller that also had a hard disk on it. VERY Interesting. Was it a TK50-AA or something more like a TK50Z-FA? Gary ================================================================================ Note 369.2 TK50s outside DEC - Amiga 2000 via standard SCSI 2 of 4 EISNER::HASSINGER "Bob Hassinger" 27 lines 26-MAY-1989 11:31 -< As much as I know - raw drive and I/F card >- -------------------------------------------------------------------------------- Tell me the difference and I will try to guess (I don't know the model numbers or much else about TK50s) What I saw was a box that was intended to be used for an IBM system ("cheapest box I could find") with the half height hard drive and the TK50 drive mounted in it. The cover was off the TK50 so everyone could observe all the threading and head movement stuff. He had installed an extra fan inside. Apparently the fan in the box power supply was not getting the job done with the TK50. There was a little magic about the cable that connected the SCSI from the GVP controller. One end had something like a DB25 and the other end had what was referred to as a standard SCSI connector. Apparently the cable had not been hard to come by. He also had a TK50 in what looked like a DECmate or MicroVAX box. I think that is a standard package. It was not connected to any thing. I think the idea was to show what the standard thing you could buy from DEC was and to impress you with an "upwards of $5K" price tag. I take it the hardware side of it was fairly simple. Getting the software playing was another story. It seems he is using what is currently (AmigaDOS V1.3) an undocumented interface in the disk driver to directly issue his own SCSI commands. AmigaDOS does not really have any higher level support for tape. He said the undocumented interface will become a documented developers interface in AmigaDOS V1.4 (coming Fairly Soon Now (maybe) - right after the Fatter Agnis chips are readily available I guess ...). ================================================================================ Note 369.3 TK50s outside DEC - Amiga 2000 via standard SCSI 3 of 4 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 6 lines 26-MAY-1989 15:12 -------------------------------------------------------------------------------- If this is the DEC H/W type I think it is, (his last initial is "P", right?) he was doing this as of at least a symposium ago, maybe more. Actually, I don't remember seeing him in Atlanta. But if it is someone else, then there are at least 2 DECies from that group doing it. ================================================================================ Note 369.4 TK50s outside DEC - Amiga 2000 via standard SCSI 4 of 4 EISNER::HASSINGER "Bob Hassinger" 6 lines 26-MAY-1989 16:03 -< Amigas seem to be big in his part of DEC >- -------------------------------------------------------------------------------- It sounds like a hot bed of Amiga fans. He talks about he and his buddy having Amigas at work hooked back to back via the serial ports to play some of those telecom type games (on lunch hour no doubt). He has a number of other Amiga fans not far away too. Gets transfers over the E-net from Europe of large amounts of Amiga software like the 210 Fred Fish disks (in off hours only), etc. etc....  ================================================================================ Note 370.0 8820+8550 ?=? 8830 2 replies EISNER::JENSEN_M 17 lines 25-MAY-1989 13:08 -------------------------------------------------------------------------------- Does anybody know of the details concerning 88xx upgrades ? We might have access to an 8820 and already own several 8550's. Since the CPU board set and memory are interchangable between 8550,8700, and 88xx, how much backplane, ribbons, power supply, etc. might have to be swapped in order make an 8830 out of a 8820 and 8550 ? The software product that drives this proposal can make good use of multiprocessing power, but not Clustered CPU's. Any 88xx techies out there ? Thanks for any help. Mike ================================================================================ Note 370.1 8820+8550 ?=? 8830 1 of 2 EISNER::ECKERT "Jerry Eckert" 17 lines 3-JUN-1989 21:48 -------------------------------------------------------------------------------- The official Digital upgrade is to physically replace the 8820-N (nee 8800) with an 8830. I'm not aware that the backplanes, etc. are available individually. > Since the CPU board set and memory are interchangable between > 8550,8700, and 88xx, That's not quite true. The MCL (memory controller) is not forward compatible (the 8820/8830/8840 MCL will work on 85x0/8810 (8700)/ 8820-N (8800) systems with a jumper change, but not vice versa), the CIM (console interface) is different, and I'm almost certain one other module in the CPU module set was modified. In addition, the NBIA (the NMI side of the NMI-BI interface) is also different. [Some 85x0/8810/8820-N systems may have the new memory controller, but most don't.] set is different. ================================================================================ Note 370.2 8820+8550 ?=? 8830 2 of 2 EISNER::SHANNON "Terry C. Shannon, IDC" 4 lines 4-JUN-1989 12:18 -< New Math. >- -------------------------------------------------------------------------------- 6430 > 8830  ================================================================================ Note 371.0 TEK 4693dx color printer setup 8 replies EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 13 lines 26-MAY-1989 09:50 -------------------------------------------------------------------------------- One of our users bought a TEKTRONIX 4693DX color image printer. He hooked it up to his PC for a while, then brought it to us and said, 'Make this work as a VAX print device.' The 4693DX is apparently capable of either serial or parallel operation. It runs at 9600 baud. In serial mode on a DMF32, we send files to it and they vanish. The port does not show any errors. In parallel mode (port LCA0), we get RMS write errors and system abort messages. Has anyone ever done this? Tektronix support said, 'We only tested it on PC's. Let us know if you get it to work.' Anyone have any suggestions on port settings, etc.? ================================================================================ Note 371.1 TEK 4693dx color printer setup 1 of 8 EISNER::KENNEDY "Terry Kennedy" 5 lines 26-MAY-1989 17:23 -< Wrong type of parallel interface >- -------------------------------------------------------------------------------- Well, I can explain the parallel mode problems - DEC expects all (recent) printers to be Dataproducts interface. Your PC is Centronics interface. The two are not interchangable without some fiddling. The big issue is that the strobe line is active high on DP and active low on Centronics. Also, some status lines are named differently between the two. ================================================================================ Note 371.2 TEK 4693dx color printer setup 2 of 8 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 14 lines 30-MAY-1989 14:43 -< Running color plotters >- -------------------------------------------------------------------------------- We have several Tek color printers running off of print queues on our VAX. To debug your problem please post the output of the following DCL commands run on your system 1) SHOW QUE/FULL target_queue 2) SHOW FORM/FULL target_or_default_form_used 3) SHOW TERMINAL target_terminal_port 4) DIR/FULL sample_plot_file It is also important to know is how the plot file was transfered from the PC to the VAX (KERMIT can really screw up plot files !). ================================================================================ Note 371.3 TEK 4693dx color printer setup 3 of 8 EISNER::LANG 20 lines 31-MAY-1989 22:09 -< Further 4693 Woes >- -------------------------------------------------------------------------------- The site that I work at has a 4693 which is connected to another manufacturer's machine. Due to operating system inconpatabilities with this other machine, we decided to look into alternative methods of using the 4693 (mostly because the system is currently non functional, and we really need the color printer!!) We found a vendor that makes a Centronix compatible interface for the Q-Bus, which we installed and are currently trying to get going. The problem that we are having is that both Tek and the manufacturer of the interface (Computer Products 361-T) are not aware that their corresponding parts to the puzzle even exist. I am currently working with both vendors to try to rectify the situation, and I will post any success/failure stories here. Note, the product flyer for the 361 states that the board is made to work with the 4693...this set-up should work. Is anyone else familiar with this set-up??? Bob Lang MIT Lincoln Labs 617 981-3411 ================================================================================ Note 371.4 TEK 4693dx color printer setup 4 of 8 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 19 lines 31-MAY-1989 23:18 -< DEC's LPV11 should work >- -------------------------------------------------------------------------------- >We found a vendor that makes a >Centronix compatible interface for the Q-Bus, which we installed and >are currently trying to get going. (this has been a L o n g time ago, and I don't have prints handy) DEC's LPV11 and their LP11 controllers have BOTH Dataproducts and Centronics support. They just don't call it that. The Lx180 in parallel mode was probably centronics compatible, so see if you have docs for strapping a DEC card for that DEC printer. A LPV11 card should be very inexpensive these days. There was once one offered (free?) in SWAPMEET, but probably has timed-out and the note been deleted. A line printer interface in not a very complicated beast. Failing all that, there are several nifty tricks in a BLACK BOX catalog. The Serial -> Parallel ones work, too. ================================================================================ Note 371.5 TEK 4693dx color printer setup 5 of 8 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 11 lines 1-JUN-1989 12:07 -< Note 15.3 tells how to use an LPV11 with Centronics printer >- -------------------------------------------------------------------------------- >> DEC's LPV11 and their LP11 controllers have BOTH Dataproducts >> and Centronics support. They just don't call it that. The Lx180 >> in parallel mode was probably centronics compatible, so see if >> you have docs for strapping a DEC card for that DEC printer. Everything you need to know about connecting an LP11/LPV11 to a Centronics-type printer is right here in note 15.3: DEC cable part number, pin-outs, and connector diagram if you want to "roll your own". I am running a Toshiba P1351 printer through 30 feet of ribbon cable to an LPV11. If you have any problems, holler. ================================================================================ Note 371.6 TEK 4693dx color printer setup 6 of 8 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 92 lines 1-JUN-1989 12:45 -< Device and Queue specs. >- -------------------------------------------------------------------------------- >> our VAX. To debug your problem please post the output of the following<< >> DCL commands run on your system << >> 1) SHOW QUE/FULL target_queue << We are using two queues, one for color, one for mono, and starting and stopping the queues to change stock. The queues are restricted to the operators and to one project, hence the acl. Printer queue TEK_COLOR, stopped, on ASPEN::LCA0, mounted form DEFAULT. /BASE_PRIORITY=4 /DEFAULT=(FEED,FORM=DEFAULT) /OWNER=[SYSTEM] /PROCESSOR=PRTSMB /PROTECTION=(S:E,O:D,G:R,W:R) (IDENTIFIER=[OPERATOR],ACCESS=READ+WRITE+EXECUTE+DELETE) (IDENTIFIER=[PROJ,*],ACCESS=READ+WRITE+EXECUTE) The other queue is the same, except the name is TEK_MONO. >> 2) SHOW FORM/FULL target_or_default_form_used << %DCL-W-IVKEYW, unrecognized keyword - check validity and spelling \FORM\ >> 3) SHOW TERMINAL target_terminal_port << When we tried to use a serial terminal port, we used : Terminal: _TXA5: Device_Type: Unknown Owner:SYMBIONT_2 Terminal Characteristics : Passall Echo Type_Ahead No Escape No Hostsync TTSync Lowercase NoTab Wrap Scope No Remote Eightbit No Broadcast No Readsync No Form Fulldup No Modem No Local-Echo No Autobaud No Hangup No Brdcstmbx DMA No Altypeahd No Set-Speed No Line Editing Overstrike Editing No Fallback No Dialup Secure Server No Disconnect No Pasthru No Syspasswd No SIXEL graphics No Soft Characters Printer port Numeric Keypad No ANSI_CRT No Regis No Block-Mode No Advacned_Video No Edit-Mode No DEC_CRT No DEC_CRT2 No DEC_CRT3 When we tried to use a parallel port, we used : Printer LCA0:, device type DMF32, is online, record-oriented device, carriage control, error logging is enabled. Error count 0 Operations completed 0 Owner Process "" Owner UIC [0,0] Owner Process ID 00000000 Dev Prot S:RWLP,O:RWLP,G:RWLP,W:RWLP Reference Count 0 Default buffer size 132 Page Width 132 Page Length 66 No Carriage-return Formfeed Uppercase No Passall No Wrap No Printall No Fallback No Tab No Truncate NOTE : It didn't make any difference if the device was spooled or not. >> 4) DIR/FULL sample_plot_file << Directory DUA61:[IMAGES] TEK_OUT_STD1.DAT;2 Size: 1/3 Created: 23-MAY-1989:13:30:07.64 Revised: 23-MAY-1989:13:30:13.22 (1) Expires: 19-NOV-1989:13:30:13.22 Backup: File organization: Sequential File Attributes: Allocation: 3, extend: 0, global buffer count: 0, no version limit Record format: Stream_LF Record attributes: Carriage Return Carriage control RMS attributes: None Journaling enabled: None File Protection: System:RWED, Owner:RWED, Group:RWED, World:RWED Access Cntrl List: (IDENTIFIER=[PROJ,*],ACCESS=READ+WRITE+EXECUTE) This is a subset of the original file; when we tried to print the original file, downloaded from the PC, we got an RMS error that the system couldn't handle a record length of >3100000 bytes. >> It is also important to know is how the plot file was transfered from << >> the PC to the VAX (KERMIT can really screw up plot files !). << Dipped if I know; the file, like the machine, was handed to us. We are trying to use a serial-to-parallel converter from ADAPTA, with no noticeable sucess. When they had the device running on a PC, they found they had to set up the parallel/serial switches on the printer reverse of the way the the manual recommended. Anything else I can tell you? ================================================================================ Note 371.7 TEK 4693dx color printer setup 7 of 8 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 45 lines 2-JUN-1989 11:59 -< Answers ! >- -------------------------------------------------------------------------------- Okay, I see several problems 1) The queue should be set /DEFAULT=NOFEED. The /FEED qualifier means that the print symbiont sticks a formfeed after printing the number of line specified by the form length. Did you specify and qualifies on the PRINT command line ? 2) Okay, I screwed up (not a first). What I need to see is $ SHOW QUEUE/FORM/FULL target_or_default_form You should probably create a new form (called TEK_COLOR) using the DEFINE/FORM command. The length and width are **NOT** important. However, it is **IMPERATIVE** to specify /NOWRAP and /NOTRUNCATE. You may also want to specifiy a SETUP module on the form that reset the plotter. 3) I see several problems with your terminal port setup. The following **MUST** be changed : $ SET TERMINAL/PERMANENT/INTERACTIVE/HOSTSYNC/NOWRAP - /NOSECURE/SCOPE/PASSTHRU/FORM/TAB Everything else looks okay. I'm not an expert on printer (parallel) port setups so maybe someone else can comment. 4) This could be a **MAJOR** problem. You have to find out how the file got to the VAX from the PC. If you used DECNET-DOS to copy it you're in good shape, if not ... The combination of "Record Format :Stream_LF" and "Record Attributes: Carriage Return Carriage Control" looks vary strange. Go back and get the original file. What does a DIR/FULL say these attributes are on that file ? How did you create the "subset file" ? Create an FDL that specifies File organization: Sequential, Record Format: Variable Length and Record Attributes: None. Now try to CONVERT the original file to the new format (you may get the same RMS error, but try it). 5) We use a Black Box (412-746-5530) Mini Print Spooler (part # GL-PIC90B, $299) to convert from serial to parallel. This is one of the such device on the market **THAT REALLY WORKS** ! ================================================================================ Note 371.8 TEK 4693dx color printer setup 8 of 8 EISNER::LANG 9 lines 20-JUN-1989 15:10 -< Interface Update >- -------------------------------------------------------------------------------- This is an update to my previous reply. The Computer Products 361T interface seems to incompatible with the VAXStation 3200. The vendor is working on getting a MicroVAX III to rectify this problem. Incidentally, has anyone seen other interfaces not work on a MicroVAX III?  ================================================================================ Note 372.0 BA32x2/RQDX3/RQDXE/Errors 9 replies EISNER::KRULEWICZ 24 lines 26-MAY-1989 09:50 -------------------------------------------------------------------------------- I know I've seen something about this before but I can't seem to find it. I have an RD54 in the second box of a dual BA32 MVII system that gets occasional disk errors. The controller is an RQDX3 to an RQDXE in the first box and then to the two distribution panels in each BA23. I think I heard something about there is an ECO about changing the cables to shielded cables. Has anyone done this? Are the cables to be replaced all the 50 conductor cables? What kind of shielded cables are they? I took a look at the 50 conductor cables and it appears to me that if one uses a ground_plane cable in this application, the ground appears on either pin 1 or pin 50. Pin 1 is designated as MFMWRTDT1 and Pin 50 is READY. Can either of these really be ground? Help. - Jerry ================================================================================ Note 372.1 BA32x2/RQDX3/RQDXE/Errors 1 of 9 EISNER::KENNEDY "Terry Kennedy" 29 lines 26-MAY-1989 17:34 -< Lots of info, some possibly useful >- -------------------------------------------------------------------------------- > I think I heard something about there is an ECO about changing the cables > to shielded cables. That FCO is RD53-specific. RD54's don't care. > Are the cables to be replaced all the 50 conductor cables? RQDX3 to RQDXE, RQDXE to internal BA23 distribution panel. The cable from the RQDXE to the bulkhead is Ok. The 50-pin round cable between the boxes is also Ok. > Can either of these really be ground? Check your source. Much of the pinout info on RQDXn is just plain *wrong* in the books. Even my fiche are wrong! Other things to look at: 1) Is your RQDXE up to rev? If not, you will write garbage on power-down. 2) RD54 in BA23 is, I belive, unsupported due to power draw. Check +5, +12 levels. (Yes, I know it works - mostly. The gotcha is the 'mostly'). Also, there are 2 types of BA23 power supplies +12 at 6A or 7A. You'd better have the 7A variety to do this. 3) Check the BA23 power harness (power supply to backplane, under the fan shroud (if no fan shroud, get professional help - you need *lots* of FCO kits. Carefully look for signs of burning or discoloration. Replace if needed (sorry, I'm at home and don't have the FCO number). ================================================================================ Note 372.2 BA32x2/RQDX3/RQDXE/Errors 2 of 9 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 12 lines 30-MAY-1989 10:03 -< BA23/RD54 *not* supported? >- -------------------------------------------------------------------------------- < Note 372.1 by EISNER::KENNEDY "Terry Kennedy" > >2) RD54 in BA23 is, I belive, unsupported due to power draw. Check +5, +12 > levels. (Yes, I know it works - mostly. The gotcha is the 'mostly'). > Also, there are 2 types of BA23 power supplies +12 at 6A or 7A. You'd > better have the 7A variety to do this. Are you sure about this? I got me a VAXstation 3200 (BA23) with two RD54's (one in the BA, the other external), and I'm sure DEC would be very upset if they shipped us an unsupported configuration..... ;-) Ed ================================================================================ Note 372.3 BA32x2/RQDX3/RQDXE/Errors 3 of 9 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 4 lines 30-MAY-1989 14:51 -------------------------------------------------------------------------------- For a short while DEC had a "special" BA23 that had power supply with a lower rating than the standard BA23. This was an attempt to come up with a "low cost" uVAX II long before the days of the 2000 ! ================================================================================ Note 372.4 BA32x2/RQDX3/RQDXE/Errors 4 of 9 EISNER::KENNEDY "Terry Kennedy" 22 lines 30-MAY-1989 20:14 -< Well, err, waffle... >- -------------------------------------------------------------------------------- > Are you sure about this? I got me a VAXstation 3200 (BA23) with two > RD54's (one in the BA, the other external), and I'm sure DEC would be very > upset if they shipped us an unsupported configuration..... ;-) Well, part RD54A-AA has a description of "159-Mbyte fixed disk drive with cables for BA23 enclosure", so it *is* possible, at least in new configur- ations. The same spec sheet lists the DC power requirements at 1.3A of +5 and 1.34A of +12. The RD53 requires 0.9A of +5 and 2.5A of +12, but has a slightly lower startup current (64W for RD53 vs 66W for RD54). I found the following note about the RD54 in the BA123 spec sheet: "Because of startup current limitations in the BA123 power supply, if an RD54 type disk drive is connected to the same 12-volt power supply regulator as another RD-type disk drive, then only 5 of the 7 amperes provided by that regulator can be used for powering the two disks and any additional options." It would seem to me that since it is noted that the RD54 is the only drive noted as having the problem, the culprit is not the BA123 but the startup surge current of the RD54 (since the BA23 does not support 2 full-height disk drives, no such note appears in its spec sheet). In any event, I would carefully sum up the power draw of all options in the box, derate by 20% to be safe, and then see if you are safe by comparing with the rating sticker on the BA23 power supply. ================================================================================ Note 372.5 BA32x2/RQDX3/RQDXE/Errors 5 of 9 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 11 lines 31-MAY-1989 12:14 -< I don't think it was a "special" >- -------------------------------------------------------------------------------- >> For a short while DEC had a "special" BA23 that had power >> supply with a lower rating than the standard BA23. This was an >> attempt to come up with a "low cost" uVAX II long before the >> days of the 2000 ! If you mean the VAXstation-II/RC (the "RC" stands for Resin Cured, since three backplane slots were filled with epoxy to prevent system expansion), I have a counter example. The BA23 that my VAXstation-II/RC came in has the full 7.0 amp 12v power supply. Maybe some were made with a different vendor's power supply? ================================================================================ Note 372.6 BA32x2/RQDX3/RQDXE/Errors 6 of 9 EISNER::HARVEY "Jack Harvey" 16 lines 31-MAY-1989 19:43 -< Short Lived Marketing Ploy >- -------------------------------------------------------------------------------- >< Note 372.5 by EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" > > -< I don't think it was a "special" >- I recall that BA23, Alan, and it *was* special. It was not the Workstation with the gummed up slots. I think it was offered only to OEMs, or some other limited market, and for only a short time. DEC explained that its low price was due to the smaller power supply. I think the backplane was only four slots, too, and not by plugging up the other four. Actually, the system price was about (as I recall) $2,000 lower than an equivilent normal system. Now *that's* cutting power supply cost. I never saw one, so I can't comment on the construction. ================================================================================ Note 372.7 BA32x2/RQDX3/RQDXE/Errors 7 of 9 EISNER::MAYHEW "Bill Mayhew" 16 lines 1-JUN-1989 17:58 -< MicroPDP-11/SV had smaller backplane, same power >- -------------------------------------------------------------------------------- I believe the "special" BA23 referred to in the past few notes was sold with MicroPDP-11 systems, only -- never saw any listings for MicroVAXen based on it. The *only* difference was the backplane -- the "special" box contained a 4-slot H9278-B instead of the normal 8-slot H9278-A. For quite a while you could buy the H9278-A from DEC for about $150. Then some "helpful" person printed this information in one of the trade rags, and DEC yanked the H9278-A from the price list. Now you can still buy it, but only as part of an upgrade kit that costs something like $3K and includes field service deinstall/reinstall. The MicroPDP that came with this box, btw, was the MicroPDP-11/SV. It was widely underappreciated and misunderstood. It was also my biggest-selling system in its day, and DEC has yet to introduce anything that can compete with it on a price-per-user basis. ================================================================================ Note 372.8 BA32x2/RQDX3/RQDXE/Errors 8 of 9 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 14 lines 2-JUN-1989 13:19 -< Third party vendors to the rescue! >- -------------------------------------------------------------------------------- >> For quite a while you could buy the H9278-A from DEC for about >> $150. Then some "helpful" person printed this information in >> one of the trade rags, and DEC yanked the H9278-A from the >> price list. Now you can still buy it, but only as part of an >> upgrade kit that costs something like $3K and includes field >> service deinstall/reinstall. I was able to get DEC to sell me one, but it was a royal pain in the bottom. You can still (I believe) get a third-party "H9278" backplane for <$300 from Zoltech Corp. in Van Nuys, CA. I bought one when I wasn't sure if DEC would honor their quote. In my opinion, it is of much higher quality than DEC's. ================================================================================ Note 372.9 BA32x2/RQDX3/RQDXE/Errors 9 of 9 EISNER::ROECKEL 15 lines 10-JUN-1989 14:04 -< Maybe you need a 'jump'er too >- -------------------------------------------------------------------------------- >I have an RD54 in the second box of a dual BA32 MVII system that gets >occasional disk errors. The controller is an RQDX3 to an RQDXE in the >first box and then to the two distribution panels in each BA23. We had some problems with our RD54's right after DEC installed them. Nobody could explain it until our local Field Service Engineer was doing some PM and said "Wow, how come these jumpers were not installed on your RQDX3 when your disks were upgraded?" Seems that these jumpers are required to keep static electricity (?) from causing disk errors on an RD54. (Jumpers not required for an RD53!) While he was there, he also changed out some cables -- couldn't understand why they were not upgraded with the disks either!  ================================================================================ Note 373.0 Digital Dealers Association 1 reply EISNER::RICE "Gary Rice" 7 lines 26-MAY-1989 10:47 -------------------------------------------------------------------------------- I am in the process of buying a used DEC computer. Several of the used equipment vendors add a logo to their ad that contains the letters "dda". Is this affiliation something that I should look for when selecting a vendor to deal with? Gary ================================================================================ Note 373.1 Digital Dealers Association 1 of 1 EISNER::HARVEY "Jack Harvey" 33 lines 26-MAY-1989 12:21 -< dda Ethics Committee >- -------------------------------------------------------------------------------- The Digital Dealers Association (dda) is a group of 100+ companies, mostly small, that strive to enforce a code of ethics on its members dealing in DEC equipment. Membership is not easy to come by. Any firm having a reputation for not following through on its promises won't get accepted. There is an effective and working procedure for expelling members for violations. Anyone doing business with a member firm may formally complain to dda if they are unhappy with the way they are treated. This triggers an investigation and hearing procedure. If the customer (you, for instance) can make a good case for bad treatment (failure to deliver equipment in the condition promised, for example, or failure to honor a promised warranty) and the member firm can't disprove it, the firm can lose its membership. It has happened. In other cases, the member firm made good on its claims or refunded the money in order to retain membership DEC "recognizes" dda inasmuch as they name a delegate to the group. If you want details of the code of ethics, contact any member or: Ray Schildknecht National Data Systems 201 744-4900 Ray is a former chairman of the dda ethics committee.  ================================================================================ Note 374.0 TK50-AA --> TK50Z-FA 10 replies EISNER::RICE "Gary Rice" 21 lines 28-MAY-1989 08:16 -------------------------------------------------------------------------------- I am in the market for a TK50 that is suitable for use with a VAXstation 2000. The DEC part number for this unit is: TK50Z-FA. The -FA sells new for over $5K. Used, it seems to run between $3500 and $4000, if you can find it. However, "all over town" you can find TK50-AAs for $1500. And according to one reference that I have seen, what you need to convert the TK50-AA to the -FA equivalent is a BA40B-AA. I have found that the BA40B-AA comes neatly packaged with an RD53 in it for as low as $1000. Thus, it is chaeper to "build" the -FA than buy it...IF I have my part numbers correct. +------------+ +----------+ | | | | | TK50-AA | | | Is a: | + | = | TK50Z-FA | ? | BA40B-AA | | | | | | | +------------+ +----------+ Gary ================================================================================ Note 374.1 TK50-AA --> TK50Z-FA 1 of 10 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 12 lines 28-MAY-1989 08:59 -< TK50Z >- -------------------------------------------------------------------------------- > However, "all over town" you can find TK50-AAs for $1500. And according to > one reference that I have seen, what you need to convert the TK50-AA to the > -FA equivalent is a BA40B-AA. I have found that the BA40B-AA comes neatly > packaged with an RD53 in it for as low as $1000. Thus, it is chaeper to > "build" the -FA than buy it...IF I have my part numbers correct. When I sold my first TK50 for a MV2000 the first thing I did was to disassemble it. There is an extra logical board in the TK50Z box that converts the SCSI bus to the standard TK50 interface. The board has a 50 pin Berg connector that takes the SCSI in - and a 26 pin Berg connector for cabling out to the TK50. ================================================================================ Note 374.2 TK50-AA --> TK50Z-FA 2 of 10 EISNER::HARVEY "Jack Harvey" 26 lines 28-MAY-1989 12:17 -< What's a BA40B-AA? >- -------------------------------------------------------------------------------- >And according to >one reference that I have seen, what you need to convert the TK50-AA to the >-FA equivalent is a BA40B-AA. I am unable to find any mention of the BA40B-AA in the DEC Price List, or VAX Systems and Options. As Jeff points out, there is a *major* difference between the TK50-AA and the TK50Z-FA. The logic board he described is not a minor cable connector adapter. It has lots of chips and is about the size of a Q-Bus dual-size board. For what it's worth, the "SCSI" bus on the MV2000 apparently ain't quite pure SCSI. The TK50Z-GA appeared for the MicroVAX 3100 (PVAX) which reportedly has a true SCSI bus. Sure enough, DEC offers an FCO to convert the TK50Z-FA to the -GA. But no help for the TK50-AA. Also, don't confuse the BA40B-AA with the BA40A-AA. The latter is simply a 1.5" high expansion that screws to the bottom on the MV2000 system box, with connectors for the TK50Z-FA, the external disk box, and communication options. If you can better identify the BA40B-AA, you may do us all a service. The whole world is trying to hang a TK50-AA in a MicroVAX 2000, following precisely your line of thinking. ================================================================================ Note 374.3 TK50-AA --> TK50Z-FA 3 of 10 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 0 lines 28-MAY-1989 13:41 -< BTW - DON'T FORGET THE TK30 IS IN THE PIPELINE >- -------------------------------------------------------------------------------- ================================================================================ Note 374.4 TK50-AA --> TK50Z-FA 4 of 10 EISNER::RICE "Gary Rice" 17 lines 29-MAY-1989 07:50 -------------------------------------------------------------------------------- My information is sketchy (ie: a one line entry in a parts list), but the part number (BA40B-AA) is what's listed as a component of the TK50Z-FA as well as the RD53-FA. The implication was that if you have a RD53-FA (and you remove the RD53) what is left is the box, power supply and some cabling. Those remains are a BA40B-AA. As Jeff so carefully points out, the "extra circuit board" in a TK50Z-FA may not be part of the BA40B-AA. THAT is the real qustion here. Anyway, the BA40B-AA is not available from DECdirect, but according to part number assistance, it IS available from "your local DECrep". Gary ================================================================================ Note 374.5 TK50-AA --> TK50Z-FA 5 of 10 EISNER::HARVEY "Jack Harvey" 1 line 29-MAY-1989 07:58 -< Terrific. Who has an RD53-FA? >- -------------------------------------------------------------------------------- I have a screw driver. ================================================================================ Note 374.6 TK50-AA --> TK50Z-FA 6 of 10 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 7 lines 29-MAY-1989 09:20 -< NOPE >- -------------------------------------------------------------------------------- > -< Terrific. Who has an RD53-FA? >- > > I have a screw driver. The RD53-FA is *NOT* a SCSI drive. It is connected directly via an ST506 interface - therefore the SCSI board would *NOT* be in that box. ================================================================================ Note 374.7 TK50-AA --> TK50Z-FA 7 of 10 EISNER::HARVEY "Jack Harvey" 31 lines 30-MAY-1989 13:52 -< Logic Board Description >- -------------------------------------------------------------------------------- The logic board found inside the MicroVAX 2000 compatible TK50Z-FA tape drive is as follows: The board is 5-5/8 x 8 inches. There are several dozen chips mounted on it. On one end, it has a 26 pin Berg connector to a ribbon cable that connects to the TK50-AA assembly. On the other end, there is a power connector and a 50 pin Berg connector. The ribbon cable in this goes (in parallel) to the two 50 pin SCSI bus connectors on the back of the cabinet. The circuit board itself is marked as follows: TZK50 side 1 - 3428-5225 side 2 - 5017638-01-B1-P2 (Comment: the latter two markings probably identify the etched circuit art work.) There is a ROM with the following printed on its label: 23-254E6-00 @DEC 12-21-88 Wyle Labs ================================================================================ Note 374.8 TK50-AA --> TK50Z-FA 8 of 10 EISNER::RICE "Gary Rice" 23 lines 3-JUN-1989 09:10 -------------------------------------------------------------------------------- > The circuit board itself is marked as follows: > TZK50 Armed with this info (along with the original part number that I have: BA40B-AA), I assalted my DECrep for a quote on the 2 parts. After much knashing of teeth, he returned with: "The 2 parts are not on AQS (DEC's Automated Quote System). I called around and finally located them thru field service. They are stocked as spare parts only and are not available as non-exchange items." Well, I thought we had something here. However, all is not lost. During the course of this investigation, I called Trimarchi. They told me that their BA40A expansion box clone supports a vanilla TK50 (no SCSI interface required). Also, the price is the lowest it has ever been: $400. This appears to allow me to buy a TK50-AA, a power supply from the local PC store and their box. The combination should come in under $2000. Can you shoot holes in my plan? Gary ================================================================================ Note 374.9 TK50-AA --> TK50Z-FA 9 of 10 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 6 lines 3-JUN-1989 10:46 -< I REALLY DOUBT THIS - ASK FOR IT IN WRITTING >- -------------------------------------------------------------------------------- > Well, I thought we had something here. However, all is not lost. During > the course of this investigation, I called Trimarchi. They told me that > their BA40A expansion box clone supports a vanilla TK50 (no SCSI interface > required). Also, the price is the lowest it has ever been: $400. This > appears to allow me to buy a TK50-AA, a power supply from the local PC > store and their box. The combination should come in under $2000. ================================================================================ Note 374.10 TK50-AA --> TK50Z-FA 10 of 10 EISNER::HARVEY "Jack Harvey" 20 lines 3-JUN-1989 17:41 -< Someone may be confused >- -------------------------------------------------------------------------------- >During the course of this investigation, I called Trimarchi. They told >me that their BA40A expansion box clone supports a vanilla TK50 (no >SCSI interface required). Also, the price is the lowest it has ever >been: $400. Jeff and I think there is a misunderstanding with Trimarchi. They have general purpose boxes for mounting 5.25" devices. The TK50-AA is a 5.25" device, therefore... Now ask about the SCSI cable that connects to the MV2000. Where does the other end connect on the Trimarchi box? This may disclose the problem. Tom Trimarchi is a pretty aggressive advertiser. And there certainly would be a *big* market for a $400 item that would do what you hope. If Tom had it, I'm sure he would be yelling about it in full page ads. On the other hand, if he does have it and doesn't realize how many people want it, please come back and let us know!  ================================================================================ Note 375.0 TECSTOR model 3/315 information wanted. 3 replies EISNER::LEDERMAN "Bart Z. Lederman" 11 lines 2-JUN-1989 10:30 -------------------------------------------------------------------------------- When we picked up some extra tape drives, one of them had a disk drive in the same cabinet. We are told that it's a 300 MB disk, and that's about all we know about it. It is physically about the same size as an RA81 (19" rack mount, about 8" high or so), and has flat cables coming out: one 60 line, one 26 line. Does anybody know anything about this drive: capacity, specs., company still in business, drives/company reliable or not, what kind of interface, can we hook it up to a Microvax-II and if so, what kind of controller, etc. Or does someone have a current address for Tecstor? ================================================================================ Note 375.1 TECSTOR model 3/315 information wanted. 1 of 3 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 1 line 2-JUN-1989 11:06 -< THAT IS AN SMD INTERFACE >- -------------------------------------------------------------------------------- > has flat cables coming out: one 60 line, one 26 line. ================================================================================ Note 375.2 TECSTOR model 3/315 information wanted. 2 of 3 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 31 lines 2-JUN-1989 12:48 -< try the Dilog DQ256 SMD controller >- -------------------------------------------------------------------------------- > what kind of controller, etc. Sounds like an SMD drive. Try the Dilog DQ256 (or DQ3256 for BA213 type boxes). That controller is TOTALLY flexible about drive geometry. You simply type in all relevant numbers, how many (1-4) virtual drives you would like it to appear as, and it will format it for you. If some needed info is missing, you can grope for the edges easily. Start by specifying geometry very small. Stretch each dimension individually until the built in format utility complains. This will get HEADS, and CYLINDERS easily (binary search, if necessary). The trickier one may be sectors per track. You typically need some dip switch setting in the drive to define sector pulses in terms of dibit counts. The controller's onboard diagnostics/formatter can even help here. You will be told how many sector pulses were seen IF your specified count was NOT = or one less than (to cater to runt sectors) the count actually seen. Also it will fail if bits / sector gets too low. The Dilog manual has much help for tuning for MAX space on a drive, and MANY popular drives are already included with geometry and drive switch settings detailed. Your drive may be there, and failing that, they may have data on it if you call and ask. AND, if you are using dual ported drives, or movable packs, there IS a compatible U bus controller that formats disks identically (DU256). The Dx256 controllers both do 4 drives, and can hack the 3.0 mbyte xfer rate of something like a big Sabre. Index and Sector pulses must be on the A cable. ================================================================================ Note 375.3 TECSTOR model 3/315 information wanted. 3 of 3 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 21 lines 2-JUN-1989 14:42 -< Tecstor address and phone >- -------------------------------------------------------------------------------- > Or does someone have a current address for Tecstor? The Electronic Industry Telephone Directory, 87/88 [latest I could find] edition, gives: Tecstor Corporation 16161 Gothard Huntington Beach CA 92647 714-842-0077 And, since someone is sure to ask, the EITD is published by Harris Publishing Co. 2057 Aurora Rd. Twinsburg OH 44087 216-425-9000 If you buy it outright it is $50 -- and would be dirt cheap at many times that price. We get ours free from one of the electronics components distributors we do business with.  ================================================================================ Note 376.0 Who sells 7-track tape drives today? 4 replies EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 14 lines 2-JUN-1989 13:30 -------------------------------------------------------------------------------- I just got a call from a friend who is trying to bid on a Navy contract. The contract calls for the ability to read tapes (using a PC or whatever other CPU is proposed) written by an old 7-track, 556 BPI tape drive. The last 7-track drive I saw was a modified Cipher 910 which the company doesn't sell any more. Furthermore, the modification was for 800 BPI only. (OK, OK, I'll confess. The drive is in my garage, just in case I ever need it. However, that doesn't help my friend, who needs more than one, preferably new.) Does anyone know where you can buy a 7-track, 556 BPI drive today? ================================================================================ Note 376.1 Who sells 7-track tape drives today? 1 of 4 EISNER::GARDNER "Tim Gardner" 16 lines 2-JUN-1989 15:09 -< CDC is a possible source >- -------------------------------------------------------------------------------- Re: < Note 376.0 by EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" > -< Who sells 7-track tape drives today? >- > Does anyone know where you can buy a 7-track, 556 BPI drive > today? The only possible source I know of is CDC. They sell* a 7 track drive for their Cyber 200 class machines. I don't have any idea what it would take to build a controller for this device, but it may be a place to start looking. *Last new one I saw was purchased 3 years ago. It's entirely possible that CDC no longer sells these beasts. tg ================================================================================ Note 376.2 Who sells 7-track tape drives today? 2 of 4 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 9 lines 2-JUN-1989 15:37 -< use TU10s >- -------------------------------------------------------------------------------- >Does anyone know where you can buy a 7-track, 556 BPI drive >today? Our TU10s went into the dumpster, but TU10s are still out there at some sites, and DEC still maintains them. Some Dealers might still be willing to sell them on a nothing-guaranteed basis. Since TU77 Masters are about $700, TU10s should be VERY inexpensive. ================================================================================ Note 376.3 Who sells 7-track tape drives today? 3 of 4 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 22 lines 4-JUN-1989 01:09 -< TU10 maintenance costs would eat them alive >- -------------------------------------------------------------------------------- >> Some Dealers might still be willing to sell them on a >> nothing-guaranteed basis. Since TU77 Masters are about $700, >> TU10s should be VERY inexpensive. This would be a last-resort solution. The ultimate customer (the government, of course) doesn't like to buy used equipment. If CDC has a standard interface, that might be a decent solution. So far, the best bet looks like buying refurbished Cipher 910's (a *very* good drive, in my opinion) and having them converted to 7-track. It strikes me as slightly ludicrous that the government is trying to buy new special systems to process tapes made on 20-year old 7-track tape drives. They would probably be money ahead to replace the old 7-track drives that produce the data with 9-track drives. It reminds me of how the Air Force had to buy vacuum tubes from Czechoslovakia to keep our air-defense computer (SAGE) running (1980-1983). RCA, GE, and Sylvania had quit making them in 1975 and the Air Force had only bought a 5-year supply. ================================================================================ Note 376.4 Who sells 7-track tape drives today? 4 of 4 EISNER::LEDERMAN "Bart Z. Lederman" 8 lines 4-JUN-1989 12:24 -< Maybe a pointer >- -------------------------------------------------------------------------------- The manual that came with our Digi-Data 1740 drive (see another note stream in this conference) says that there are 7 track models. You'd have to contact Digi-Data directly for more information. I agree that 7 track is VERY obsolete. I'd even be a little supprised that any tapes that old are still useable. Maybe they should check, and if some system is still producing 7 track tapes they should buy new 9 track drives for that system and do an upgrade.  ================================================================================ Note 377.0 What is an LVP16? 1 reply EISNER::KOZAM 10 lines 2-JUN-1989 21:26 -------------------------------------------------------------------------------- What is the Hewlett-Packard designation for the LVP16? It only supports the HPGL graphics language, looks like an HP plotter, works like one. I'd like to use it for output from a package that supports HP plotters, but it wants to know the HP model number. Has DEC "modified" it so that it isn't totally equivalent to its HP look-alike? While I'm at it, anyone know what a Dilog DQ100 controller is? I stumbled upon it while going through a stack of old boards. ================================================================================ Note 377.1 What is an LVP16? 1 of 1 EISNER::BURNET "John Burnet" 0 lines 2-JUN-1989 23:25 -< The LVP16 is an unmodified HP 7475A >- --------------------------------------------------------------------------------  ================================================================================ Note 378.0 RD52/RQDX1-E compatibility? 1 reply EISNER::MAYHEW "Bill Mayhew" 12 lines 8-JUN-1989 15:16 -------------------------------------------------------------------------------- I have a system (11/73) with an RQDX1 and two RD52s (the second one being a -DA, connected via, I believe, an RQDX1-E). I want to upgrade one of the RD52s (the internal one) to an RD53. I have the required RQDX2. My question is, will the RQDX1-E work, or do I also need to go to an RQDXE? (I'll check for the shielded cables while I'm in there. I know there is some relevant discussion to this in topic 6, but that topic became *so* wide-ranging, and is now so dated, that I opted to pose the question again here; it wasn't clear from 6.* that this was ever definitively answered.) ================================================================================ Note 378.1 RD52/RQDX1-E compatibility? 1 of 1 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 6 lines 8-JUN-1989 17:05 -< OK >- -------------------------------------------------------------------------------- > I want to upgrade one of the RD52s (the internal one) to an RD53. > I have the required RQDX2. My question is, will the RQDX1-E work, or > do I also need to go to an RQDXE? The RQDXE is only needed for RQDX3's. I have RQDX2's with RQDX1-E's  ================================================================================ Note 379.0 DHV/DZV/DZQ distribution panels 5 replies EISNER::MAYHEW "Bill Mayhew" 10 lines 8-JUN-1989 15:22 -------------------------------------------------------------------------------- I have a BA23 that I just acquired that has a DHV11 in it, but the rear bulkhead has only the cab kit for a DZV11 (DZQ11?) in it -- just 4 ports. (The box has been collecting dust for quite some time and various stuff has been moved into/out-of it and other BA23s at the site, hence the disarray.) Is there any difference between the DZV/DZQ distribution panel and one of the normally-two DHV distribution panels? Can I reliably use this distrib panel with half of the DHV? ================================================================================ Note 379.1 DHV/DZV/DZQ distribution panels 1 of 5 EISNER::DAVIS "Steven P. Davis" 4 lines 9-JUN-1989 12:10 -< Yes, it works. >- -------------------------------------------------------------------------------- To the best of my knowledge they are the same. I have used distribution panels between the various controllers without difficulty. I did not specifically check every feature of the controllers. ================================================================================ Note 379.2 DHV/DZV/DZQ distribution panels 2 of 5 EISNER::HUDGINS "Jerry Hudgins" 6 lines 12-JUN-1989 09:19 -< Well, maybe, sort of >- -------------------------------------------------------------------------------- I believe that I have experienced difficulty in using the DHV w/ DZV cab kit combination; I seem to remember that modem control is not passed properly. I think that DCD wasn't making it out of the cab kit. There may have been other problems. For a data-leads-only application, though, it might work. Anyone else tried this a bit more carefully? ================================================================================ Note 379.3 DHV/DZV/DZQ distribution panels 3 of 5 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 0 lines 12-JUN-1989 10:21 -< The part # for the DHV panel is 50-15588-01 / what's the DZ >- -------------------------------------------------------------------------------- ================================================================================ Note 379.4 DHV/DZV/DZQ distribution panels 4 of 5 EISNER::MAYHEW "Bill Mayhew" 5 lines 12-JUN-1989 12:30 -< Seventysomething >- -------------------------------------------------------------------------------- The DZV panel I have is etched on the board with 50-15419, but I believe the 50-series numbers are PCB numbers only. The part number for the distribution panel for a DZQ is listed as 70-19964-00 in the DZQ User's Guide. The DHV Technical Manual lists *its* panels as H3173-As. ================================================================================ Note 379.5 DHV/DZV/DZQ distribution panels 5 of 5 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 21 lines 12-JUN-1989 12:51 -< Cheap DHV cabinet kits from DEC >- -------------------------------------------------------------------------------- >> I have a BA23 that I just acquired that has a DHV11 in it, but the >> rear bulkhead has only the cab kit for a DZV11 (DZQ11?) in it -- I just got a DHV11 cab kit (2 panels) direct from DEC for only $55 (plus tax). I ordered it from DEC's "The Outlet" flyer dated Spring 1989. If you call quickly, they might still have some at that price. The info you need is: 1. Part number is CK-DHV11-AC (Cabinet kit for DHV/EIA, 30" cable) 2. Specify "Source code UN" and "The Outlet, Spring 1989" in your order. 3. Place order through DECdirect at 800-344-4825 or mail to DEC, P.O. Box CS2008, Nashua, HN 03061. 4. If you prepay by check, DEC will pay shipping. Don't forget to include sales tax. 5. Quantities are limited and orders MUST be placed by 30-Jun-89.  ================================================================================ Note 380.0 Relation of SID jumpers to SID value 6 replies EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 11 lines 8-JUN-1989 16:02 -------------------------------------------------------------------------------- We are in the process of installing the latest version of the 8600/8650 console media, firmware, and diagnostics. One of the things our field service organization has said is that they need to change the SID to reflect the proper FCO level for the machine before the console will load the microcode. I assume that this will change the value returned by the F$Getsyi( "SID" ) lexical or the $GETSYI system call for the SID. Is it possible to determine what the new SID value will be if I can find out the jumper configuration before and after the change? Please, go to SOAPBOX if you want to discuss the problem of using the SID for software licensing, I have a system we will need to "fix". ================================================================================ Note 380.1 Relation of SID jumpers to SID value 1 of 6 EISNER::RENES "Stuart Renes, DFWLug Chair" 11 lines 8-JUN-1989 17:53 -< Yes. >- -------------------------------------------------------------------------------- Yes it is possible. The FCO should have such information for you. This approach is not new. It started during the 780 to 785 upgrade days. BTW, any code that uses the SID should ONLY use the lower part (I cant remember how many bits but the hardware handbooks have it all described nicely). That FCO should only affect the upper part and then your software package won't be affected. Stuart. ================================================================================ Note 380.2 Relation of SID jumpers to SID value 2 of 6 EISNER::KENNEDY "Terry Kennedy" 26 lines 8-JUN-1989 17:53 -< DEC software looks at SID, too! >- -------------------------------------------------------------------------------- > I assume that this will change > the value returned by the F$Getsyi( "SID" ) lexical or the $GETSYI > system call for the SID. Yes, it will. > Is it possible to determine what the new > SID value will be if I can find out the jumper configuration before > and after the change? Field Service should be able to get that from Colorado for you. It may or may not be correct, depending on their understanding of what upgrade your local rep is doing. I would suggest you keep your old console media and have the rep build a new one for the new rev. Also have him show you the jumper which changes the rev, so you can go back if you have to. By the way, you will get stuck on three things: 1) Re-run AUTOGEN as it will get very confused otherwise (it will say this is a different CPU). Verify all your parameters are still Ok and reset as needed. 2) Your error log will now think you have a different CPU. Have your service rep look at the existing entries and decide if he wants it saved under another name or just deleted. Then rename/delete it ans start fresh. 3) Some cluster software will get confused (VAXsim, etc.). Have your rep verify all diag- nostic tools are still functioning properly after the update. ================================================================================ Note 380.3 Relation of SID jumpers to SID value 3 of 6 EISNER::FULLER_S "Funky cold medina" 19 lines 8-JUN-1989 23:07 -< The jumpers ARE the SID! >- -------------------------------------------------------------------------------- I've not heard of anything in the console media that causes or requires a change in the SID for the 86xx line. In the SID, there is the cpu type code (hi byte = 04). Bit 23 is 1 for 8650, 0 for 8600. Bits 16-22 reflect the hardware revision level. Bits 0-15 are the cpu serial number. There are jumpers for bits 0-23, but bits 24-31 are set by the console software, and will always be '04'x anyway. The console will check the hardware revision against a text file on the RL02 (KA865.REV), and will complain if the hardware rev is low, but will load the microcode and continue anyway. VAXsim(+) and SPEAR only look at the low 12 bits of the SID for licensing purposes. But, Seton's with CDC for maintenance, anyway, so the point is moot. Stu (DEC Field Service) ================================================================================ Note 380.4 Relation of SID jumpers to SID value 4 of 6 EISNER::KENNEDY "Terry Kennedy" 16 lines 9-JUN-1989 03:05 -< Generic title >- -------------------------------------------------------------------------------- > Bits 16-22 reflect the hardware revision > level. This is what they want to change, in all probability. Whether it is required or not is an interesting question, but not changing the jumpers while changing the hardware is probably an invitation to disaster in the future... > VAXsim(+) and SPEAR only look at the low 12 bits of the SID for > licensing purposes. But, Seton's with CDC for maintenance, anyway, > so the point is moot. I wasn't concerned so much with the licensing as I was with the programs thinking it was a different CPU. Back when DEC upgraded our 780 microcode and FPLA, one of these guys thought we had another CPU (and that both were eligible for the code to run on). The cute little maps they draw showed *2* CPU's, both the same node name. This may have been fixed, though... ================================================================================ Note 380.5 Relation of SID jumpers to SID value 5 of 6 EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 7 lines 9-JUN-1989 09:19 -< Jumper config will be checked >- -------------------------------------------------------------------------------- I plan on getting the exact new jumper configuration before I the techs change anything and then pass the new SID by any vendors that are doing things with it. Our hardware service vendor does have some software on the system that probably checks the SID, so they are going to have to deal with that as well... Seton ================================================================================ Note 380.6 Relation of SID jumpers to SID value 6 of 6 EISNER::GRANZOW 14 lines 9-JUN-1989 11:46 -------------------------------------------------------------------------------- < Note 380.5 by EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" > -< Jumper config will be checked >- If the console pack is upgraded to REVISION 9.13, the jumpers will be checked by the console software and if found to be below revision, the system will not come up. The minimum revision can be found on the 9.13 pack by using Exchange under VMS and doing a exc> type csa1:ka860.rev or ka865.rev. At the >>> prompt you can do a >>> show ka860.rev/ascii or show ka865.rev/ascii. Breakdown of the SID can be found in 86xx INSTALLATION MANUAL EK-8600I-IN in chapter 10. (normally page 10-11)  ================================================================================ Note 381.0 RQDX1+RX50+MicroVAX II??? 15 replies EISNER::OSUDAR "John Osudar" 13 lines 8-JUN-1989 16:45 -------------------------------------------------------------------------------- We've got a MicroVAX II with an RQDX2 and a MicroPDP 11/23 with an RQDX1. The RQDX2 has an RD53 and two RX50's on it -- but only the RX50's are ever used for anything. The RQDX1 has something ridiculous (a 10 megabyte DEC disk???) on it. We'd like to put the RQDX2 and the RD53 into the 11/23. There doesn't appear to be any problem with this. We'd also like to keep the RX50's on the MicroVAX II -- so we were wondering whether the RQDX1 could control the two RX50's, and whether the RXDX1 is supported in a MicroVAX II under the current version of VMS. Does anybody know the answers? ================================================================================ Note 381.1 RQDX1+RX50+MicroVAX II??? 1 of 15 EISNER::KENNEDY "Terry Kennedy" 16 lines 8-JUN-1989 17:59 -< Where dou you want the unsupportedness? >- -------------------------------------------------------------------------------- > We'd like to put the RQDX2 and the RD53 into the 11/23. There doesn't > appear to be any problem with this. Should be ok. Just make sure you have the shielded cable between the RQDX2 and the signal distribution board in the backplane assembly. Also watch your +12 loading carefully - the RD53 draws 2.6A. > We'd also like to keep the RX50's on the MicroVAX II -- so we were > wondering whether the RQDX1 could control the two RX50's, and whether > the RXDX1 is supported in a MicroVAX II under the current version > of VMS. Supported, no; works, yes. If this is a real problem for you, you could get a set of RQDX2 ROMs and put them in the RQDX1; then the RQDX1 will control an RD53 and be unsupported, but you'll have a supported RQDX2 in your MicroVAX... ================================================================================ Note 381.2 RQDX1+RX50+MicroVAX II??? 2 of 15 EISNER::OSUDAR "John Osudar" 6 lines 9-JUN-1989 09:49 -< if it works, it works >- -------------------------------------------------------------------------------- Thanks... I should have been more explicit in what I meant by "supported" -- "it works" is what I was looking for (i.e. that the ancient RQDX1 works with the current VMS disk driver); I take it that the answer is "yes"... (I don't care if it's "supported" in the "endorsed by DEC" sense -- if only they *knew* some of the things we do... :-) ================================================================================ Note 381.3 RQDX1+RX50+MicroVAX II??? 3 of 15 EISNER::CROWELL "Shefth of the Fourth Order" 8 lines 10-JUN-1989 18:08 -< Is this the last bus to DMAland? >- -------------------------------------------------------------------------------- The RQDX1 should work in the MicroVAX and the VMS driver should work with it - though unsupported. CAVEAT!!! Remember that the RQDX1 has the "feature" that it doesn't pass DMA grant signals. When you install it in the MicroVAX, be certain that it is the last DMA device on the bus. ================================================================================ Note 381.4 RQDX1+RX50+MicroVAX II??? 4 of 15 EISNER::KENNEDY "Terry Kennedy" 7 lines 11-JUN-1989 03:46 -< The RQDX2 has the same problem... >- -------------------------------------------------------------------------------- > Remember that the RQDX1 has the "feature" that it doesn't pass DMA > grant signals. When you install it in the MicroVAX, be certain > that it is the last DMA device on the bus. The RQDX2 is no champ here either- they're really just different artwork revs of the same board. Of course, what do you do with this and a VCB01, *both* of which want to be last? ================================================================================ Note 381.5 RQDX1+RX50+MicroVAX II??? 5 of 15 EISNER::MAYHEW "Bill Mayhew" 4 lines 11-JUN-1989 13:48 -< RQDX2 was SUPPOSED to fix that problem >- -------------------------------------------------------------------------------- DEC installation docs (no flames please on how much THOSE are worth!) state that the RQDX2 does NOT have to be last... this is one of the things that the RQDX2 was introduced to solve. Are you saying that experience has shown otherwise? ================================================================================ Note 381.6 RQDX1+RX50+MicroVAX II??? 6 of 15 EISNER::KENNEDY "Terry Kennedy" 11 lines 11-JUN-1989 16:43 -< Yet more on the RQDX2 >- -------------------------------------------------------------------------------- > Are you saying that experience has shown otherwise? Well, the RQDX2 at least *tries* to pass bus grant - but it still has some bugs in it. If it isn't last, you can still hang the bus with cer- tain operations. Of course, the RQDX3 is supported everywhere the 2 is, does RX33 as well, is 1/2 the size and power, does 1-1 interleave on recent RD-type disks (all but RD51, I believe), and is available *cheap* on the used market - so why stay with a 1 or 2, especially when you can buy an RQDX3 with 2 Maxtor 2190's (RD54's) for the cost of 1 RD54 from DEC? ================================================================================ Note 381.7 RQDX1+RX50+MicroVAX II??? 7 of 15 EISNER::MAYHEW "Bill Mayhew" 8 lines 12-JUN-1989 12:24 -< RQDX2 is still much less expensive >- -------------------------------------------------------------------------------- RQDX3s seem to be running $800-1K (or so) in the used market. Not all RQDX3s will support RX33s (ROM upgrade may be required). OTOH, RQDX2s are much, much cheaper, particularly if one has clients that are upgrading from DX2/RD53s to DX3/RD54s, and hence can buy them at the _wholesale_ used price. And I've picked up RD53s at as little as $325 (including freight) recently, which is less than 1/4 the best price I've seen for a 2190. ================================================================================ Note 381.8 RQDX1+RX50+MicroVAX II??? 8 of 15 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 12 lines 12-JUN-1989 12:39 -< I disagree about the VCB01 >- -------------------------------------------------------------------------------- >> Of course, what do you do with this and a VCB01, *both* of >> which want to be last? Gee, my VCB01 has the following after it and nobody has complained yet: RRD50 CDROM DEQNA Ethernet DQ-132 Tape TQK50 TK50 RQDX3 Disk DZQ11 Terminals ================================================================================ Note 381.9 RQDX1+RX50+MicroVAX II??? 9 of 15 EISNER::SMITH_PA "Paul Smith" 7 lines 20-JUN-1989 22:24 -< RQDX1 >- -------------------------------------------------------------------------------- I didn't realise anybody was still clamouring for RQDX1's. I have a shelf full of them that I can't give away. Anybody want to prove me wrong? ================================================================================ Note 381.10 RQDX1+RX50+MicroVAX II??? 10 of 15 EISNER::OSUDAR "John Osudar" 5 lines 21-JUN-1989 13:00 -< who wants one? not me... >- -------------------------------------------------------------------------------- Not clamoring for RQDX1 (do I *look* crazy?!) but happen to have one that I don't need (and also can't give away) -- and I wanted to see if I could make it earn its keep. If you have a shelf full of them, you've got a lifetime supply of *wonderful* dust collectors... ================================================================================ Note 381.11 RQDX1+RX50+MicroVAX II??? 11 of 15 EISNER::LEDERMAN "Bart Z. Lederman" 6 lines 21-JUN-1989 16:47 -< We could use one (maybe?) >- -------------------------------------------------------------------------------- We happen to have a Micro-VAX II, an empty slot, an empty drive slot (no 5 1/4 " drives on this unit, it uses RA81s), and can get some RX50 drive units off of some Rainbows which went up in flames. Don't know if we can get cables. If we got all the parts for free, does it make sense to bother putting one in as was implied by the original note? ================================================================================ Note 381.12 RQDX1+RX50+MicroVAX II??? 12 of 15 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 26 lines 22-JUN-1989 12:25 -< RQDX1s are still useful. >- -------------------------------------------------------------------------------- >> I didn't realise anybody was still clamouring for RQDX1's. >> I have a shelf full of them that I can't give away. >> Anybody want to prove me wrong? Only a few weeks ago, a customer with a dying RD52 sent it to me to recover anything I could. I had to borrow a system with an RQDX1 in order to do the recovery. Just in case I ever have to do this again, I would like to have one of those "dust collectors" on the shelf. The recovery? I got back all but four or five blocks of the customer's data and all of the important stuff. It involved careful temperature control (translation: I put the drive on the table in front of an air conditioner after soaking it in the refrigerator for a few hours), orientation (some blocks could be read with the drive on its side, others with it on end), and lots of patience. I found that certain controller/drive errors could cause the system (PDP-11/73 w/RSX-11M) to lock up in such a way that I had to reboot it many times. I also had to delete a 4000-block error log file when I was done. This also took three days to do. DEC had told them that the drive was a total loss and it would be impossible to recover the files. It turned out to be a good example of what you can do if you throw away the rule book. ================================================================================ Note 381.13 RQDX1+RX50+MicroVAX II??? 13 of 15 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 9 lines 22-JUN-1989 13:58 -< NOPE - THIS IS AN EXAMPLE OF PLAYING BY THE RULES >- -------------------------------------------------------------------------------- > DEC had told them that the drive was a total loss and it would > be impossible to recover the files. It turned out to > be a good example of what you can do if you throw away the > rule book. 1. Always question what DEC F.S. tells you 2. The golden rule - always depend on 3rd party consultants ;-) ================================================================================ Note 381.14 RQDX1+RX50+MicroVAX II??? 14 of 15 EISNER::SMITHJR "Charles T. Smith, Jr." 9 lines 23-JUN-1989 00:22 -< Free boards? >- -------------------------------------------------------------------------------- I would also be interested in taking up to three of the boards off your hands. I have three microvax systems with four hard disks and nothing to hook up the RX50's to. (Systems have an RQDX3 and 4 RD53's or 54's). Of course, I'll cheerfully pay shipping. Drop me a line in mail if you really would like to clear a few units off the shelf. ================================================================================ Note 381.15* RQDX1+RX50+MicroVAX II??? 15 of 15 EISNER::MERRIMAN "George Merriman - CCA/NY" 5 lines 23-JUN-1989 22:13 -< Try SWAP_MEET for hardware bargains >- -------------------------------------------------------------------------------- I think any discussion on the exchange of hardware should take place either in SWAP_MEET or off-line via VMS mail. Use now to get to SWAP_MEET. Thanks. George Merriman -- HARDWARE_HELP co-moderator.  ================================================================================ Note 382.0 diskless VAXstation 3100? 10 replies EISNER::OSUDAR "John Osudar" 9 lines 8-JUN-1989 17:06 -------------------------------------------------------------------------------- We just came to the horrible realization that the diskless VAXstation 3100 systems are not only diskless, but disk *controller*-less... I assume that those who are buying VS3100's aren't paying DEC the ridiculously high prices they charge for ridiculously low-capacity SCSI disks. Right? So what are you doing -- buying systems with the 52-megabyte disks just to get the controller? Or are we missing something here? ================================================================================ Note 382.1 diskless VAXstation 3100? 1 of 10 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 5 lines 8-JUN-1989 21:48 -< Yup >- -------------------------------------------------------------------------------- We have 11 3100s only one of which is diskless (for cheapness reasons). The rest come with the 52MB drive and floppy. (I noticed the controller-less nature of the beast just before ordering). ================================================================================ Note 382.2 diskless VAXstation 3100? 2 of 10 EISNER::KILGALLEN "Larry Kilgallen" 2 lines 8-JUN-1989 22:08 -< I'm glad they offer it that way! >- -------------------------------------------------------------------------------- It will make it more difficult for someone to attach an unauthorized disk drive. ================================================================================ Note 382.3 diskless VAXstation 3100? 3 of 10 EISNER::JENSEN_M 9 lines 9-JUN-1989 02:54 -< 3100 Bare SCSI controller ? >- -------------------------------------------------------------------------------- I believe a recent DEC Sales update made mention of new options for the 3100. One was the SCSI controller - for about $1150. Notes are in the office. > It will make it more difficult for someone to attach an unauthorized > disk drive. Unauthorized from whose perspective ? The seller, the owner, the user, or ? Just curious. ================================================================================ Note 382.4 diskless VAXstation 3100? 4 of 10 EISNER::OSUDAR "John Osudar" 5 lines 9-JUN-1989 09:54 -< unauthorized disks on missing controllers >- -------------------------------------------------------------------------------- Knowing Larry, "unauthorized" from the perspective of the owner (who is not the user) *IF* the controller is available as an add-on option, this is OK (>$1000 is somewhat steep?!); if it's *NOT*, this is ridiculous! ================================================================================ Note 382.5 diskless VAXstation 3100? 5 of 10 EISNER::ALDRICH "Jeff Aldrich" 8 lines 9-JUN-1989 12:48 -< Why do they do this to us? >- -------------------------------------------------------------------------------- We noticed this early enough but are still upset. We just want a workstation with a local paging disk and to get that from DEC right now we also have to buy a floppy drive. We would prefer that our students not be able to bring unknown code and put it onto our systems. (Grumble, grumble!!!) Let's hope that the rumors of a SCSI controller option are true!! ================================================================================ Note 382.6 diskless VAXstation 3100? 6 of 10 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 18 lines 9-JUN-1989 13:36 -< The floppy isn't welded in place, is it? >- -------------------------------------------------------------------------------- > We noticed this early enough but are still upset. We just want > a workstation with a local paging disk and to get that from DEC > right now we also have to buy a floppy drive. We would prefer that > our students not be able to bring unknown code and put it onto our > systems. (Grumble, grumble!!!) To address this particular problem, why not remove the floppy? Of course, if you REALLY want to prevent them from bringing in unknown code, you'll also have to do something about the serial port... not to mention the keyboard. As as been said in other discussions in other conferences: You can either trust VMS security measures to keep the non-priv'ed users non-priv'ed... or you can not let them on your systems at all. Most attempts at compromise are (a) very inconvenient for the honest folks and (b) ineffective against the one trying to do you dirt. ================================================================================ Note 382.7 diskless VAXstation 3100? 7 of 10 EISNER::MAYHEW "Bill Mayhew" 6 lines 9-JUN-1989 16:08 -< Part numbers >- -------------------------------------------------------------------------------- VS42D-Hx SCSI/ST506 Floppy controller VS42D-Jx SCSI/SCSI controller (i.e. 2 SCSI "busses") where x = 'G' for factory-installed, 'F' for field-installed. Field Service installation for the latter is $370. Price of all variants is $1000 (LIST price). ================================================================================ Note 382.8 diskless VAXstation 3100? 8 of 10 EISNER::KILGALLEN "Larry Kilgallen" 25 lines 9-JUN-1989 16:15 -< "unauthorized" local media >- -------------------------------------------------------------------------------- > Note 382.4 > EISNER::OSUDAR "John Osudar" > Knowing Larry, "unauthorized" from the perspective of the owner > (who is not the user) And indeed, John seems to know me well. There are two major issues which can lead to this situation: a) Owner wants to prohibit unauthorized removeable media to prevent unauthorized import or export of programs or data. (Import either with a view toward viruses (previously mentioned) or a view toward unlicensed use of programs. (As a vendor, I won't get much by sueing Jane Programmer, but boy I might make out in an action brought against Megabucks Corporation.) As Jamie mentioned, you can't protect against all possibilities through the use of hardware tricks. You can, however, discourage the undesired behaviour. The ultimate loophole is that someone can look at a screen and carry information out of the building on a piece of paper, or even more insidious just by remembering it. b) Owner wants to prohibit the use of local hard disks because of the belief (well founded, in most cases) that the owner's data kept there will not get backed up. ================================================================================ Note 382.9 diskless VAXstation 3100? 9 of 10 EISNER::ALDRICH "Jeff Aldrich" 6 lines 19-JUN-1989 15:20 -< Such is life... >- -------------------------------------------------------------------------------- I certainly understand that we cannot stop the really determined dirt-doers. I just don't want to make it easy for them. I did consider disconnecting the floppy drive. It's not a real big deal, but we are working on a fixed grant budget and would prefer not to pay for things we don't need. Oh well... ================================================================================ Note 382.10 diskless VAXstation 3100? 10 of 10 EISNER::CHADWICK "Keith Chadwick, Fermilab" 14 lines 20-JUN-1989 11:16 -< A software solution >- -------------------------------------------------------------------------------- > < Note 382.9 by EISNER::ALDRICH "Jeff Aldrich" > > > I certainly understand that we cannot stop the really determined > dirt-doers. I just don't want to make it easy for them. I did > consider disconnecting the floppy drive. You can make the floppy drive inaccessable to a non-privileged user by placing the following ACL on it during your SYSTARTUP: SET DEVICE/ACL=(ID=*,ACCESS=NONE) ddcu We us this in order to protect the console disks from our users. -Keith.  ================================================================================ Note 383.0 CI-CI without Star Coupler (2 nodes) 5 replies EISNER::WRIGHT_S "Sim Wright" 6 lines 11-JUN-1989 07:21 -------------------------------------------------------------------------------- CI - CI link without Star Coupler. Does anyone know if this is possible, and or HOW I could do this, with only two nodes. I realize that the sending node and channel has to see its own packets? Is there any available information? Thanks... ================================================================================ Note 383.1 CI-CI without Star Coupler (2 nodes) 1 of 5 EISNER::FULLER_S "Funky cold medina" 20 lines 11-JUN-1989 11:52 -< Nope, no way, no how - gotta use SC! >- -------------------------------------------------------------------------------- You must have the star coupler - period. However, you might be able to get away with not having the SC cabinet. The cabinet just contains 2 (or 4) coupler panels, mounted on _plastic_ (ie. non-conductive) mounting brackets. You could remove thepanels from the SC cab and mount them in some convenient place in/on your cpu cabinet (or the HSC cab, if you have one). You must still use the plastic mounting brackets, tho, since the panels should not be grounded. But, you must have the cables plug into the coupler, and not direct cpu to cpu. Whether this is supported or not is probably up to whomever is going to support you . . . Stu ================================================================================ Note 383.2 CI-CI without Star Coupler (2 nodes) 2 of 5 EISNER::LEDERMAN "Bart Z. Lederman" 7 lines 11-JUN-1989 14:46 -< DEC will even show you how. >- -------------------------------------------------------------------------------- > However, you might be able to get away with not having the SC cabinet. > The cabinet just contains 2 (or 4) coupler panels, mounted on _plastic_ The latest TEMPEST products catalog specifically states that you CAN do this, and it is in fact the supported method of setting up a cluster. They show an HSXx0 controller in a 5 foot shielded cabinet with the star coupler mounted above it. ================================================================================ Note 383.3 CI-CI without Star Coupler (2 nodes) 3 of 5 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 20 lines 12-JUN-1989 10:19 -< It should work... >- -------------------------------------------------------------------------------- Actually, if all you have are two nodes (1 VAX <-> 1 HSC, or 2 VAX <-> 0 HSC) then you should be able to. I'm basing this statement on the fact that I've seen Field Service do loopback testing on the CI with loopback connectors. Each of the two loopback connectors used (one for path A and one for path B) consists of two hunks of coax with TNC connectors, and an attenuator (I want to say 5dB, but it's been a while) in the middle. Come to think of it, I remember seeing a video tape of Stu Farnham (one of the VMS engineers) where he discussed clusters in general. He mentioned how one can make the loopback connectors, and something was said about using them during development to do CI bandwidth testing. Since the current star coupler is strictly a passive device, I can see no reason why this setup wouldn't work. Of course you have diddly for expansion capability. Just ask you friendly FS person to let you take a look at their CI loopback connectors, and make some yourself. Just make one of the cables a bit longer... ;-) Ed ================================================================================ Note 383.4 CI-CI without Star Coupler (2 nodes) 4 of 5 EISNER::HARVEY "Jack Harvey" 4 lines 12-JUN-1989 20:11 -< Watch Ground Loops >- -------------------------------------------------------------------------------- I believe one of the functions the passive star coupler provides is power isolation. If you have a ground potential difference between cpu cabinets, it might cause a problem even if you get the cables to mate. ================================================================================ Note 383.5 CI-CI without Star Coupler (2 nodes) 5 of 5 EISNER::FULLER_S "Funky cold medina" 19 lines 14-JUN-1989 22:29 -< You gotta have a SC! >- -------------------------------------------------------------------------------- > Actually, if all you have are two nodes (1 VAX <-> 1 HSC, or 2 VAX ><-> 0 HSC) then you should be able to. I'm basing this statement on the >fact that I've seen Field Service do loopback testing on the CI with >loopback connectors. Each of the two loopback connectors used (one for The scenario described is just as it mentions: loopback testing only. The HSC, as part of its self test, wants to see its own packets. If it can't, then the CI port's self test fails in test 11. If you have the Xmit port of the HSC directly to the VAX's Recv port, (and vice versa), then the HSC won't see its packets and will fail the self test. I don't know what the HSC's response is -that is, will it attempt to use the CI port anyway, or just sit there and vegetate? Bottom Line: You gotta have a star coupler, even if it's just the panel stuck in the back/top of an HSC or in the bottom of the CI750 cab (as once was sold with a 4 port SC).  ================================================================================ Note 384.0 Source and Meaning of device errors? 3 replies EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 13 lines 14-JUN-1989 18:59 -------------------------------------------------------------------------------- When a disk or tape served by an HSC has an error and the HSC/device logs errors to the error log (well at least in VMSland) where do the various messages come from? For example, one can get the following MSCP Messages: SDI Error Disk Transfer Error Bad Block Replacement Each of these can have different flags along with them. In addition when running shaddow sets you get error log entries that don't seem to have a message type but do have a bunch of information about the state of the unit, along with some usage information. ================================================================================ Note 384.1 Source and Meaning of device errors? 1 of 3 EISNER::FULLER_S "Funky cold medina" 21 lines 14-JUN-1989 22:23 -------------------------------------------------------------------------------- I'm not sure I understand what you mean by "where do the various messages come from?". They come from the HSC. The DUDRIVER simply copies the message as-is, along with some unit info to an error log buffer, and the ERRFMT process adds things like the date/time and a sequence number and writes it to the error log file. There are about 10 error log message types defined, including the 3 types you mentioned already, plus: "Controller error" "Host access error" "Small disk error" (typically, RDxx, RC25 and I think RFxx) and 3 or 4 varieties of tape errors. The format byte (MSLG$B_FORMAT) determines how the error log report program (ERF or ANA/ERR) should handle the rest of the packet. Stu ================================================================================ Note 384.2 Source and Meaning of device errors? 2 of 3 EISNER::DROPPERS "Seton Droppers EXCOM Mem @Large" 10 lines 14-JUN-1989 23:33 -< But where do they come from? >- -------------------------------------------------------------------------------- Well I was a bit vague, but I wanted to be... Part of what I was asking you answered when you said it comes from the HSC, handled by DUDRIVER. This brings a couple more questions: 1) What causes the HSC to send the error (some are result of software conditions in the HSC, some because the drive itself to send an error return, etc.?) 2) Is this "path" of error messages documented for "the rest of us" anywhere? ================================================================================ Note 384.3 Source and Meaning of device errors? 3 of 3 EISNER::BATES "Ken Bates" 15 lines 15-JUN-1989 21:16 -------------------------------------------------------------------------------- > 1) What causes the HSC to send the error (some are result of software > conditions in the HSC, some because the drive itself to send an > error return, etc.?) The HSC will send an error log to the host whenever it detects an error. If the error is as a result of a request from a host (e.g., an MSCP operation), it will include information which relates the error to the specific command which caused the error. If an error occurs without a command from the host causing it, the HSC will report that also, albeit in a slightly different format. As Stu pointed out, there are several general classes of error logs, with each type having a specific format. - Ken  ================================================================================ Note 385.0 Creating DTE_DF224 for dialing No replies EISNER::MCCORMICK_B "Virtual Bob" 116 lines 17-JUN-1989 20:43 -------------------------------------------------------------------------------- I was cleaning out my directory and came accross this file, which may have been published a few years back in PageSwapper... maybe its of use posted here (for electronic searching...) Of course, all current DEC modem offerings (DF242, DF124-C) are DMCL and hence you wouldn't have to consider the following. But there was no Scholar Plus at the time, so this is what I had to do. Note the VMS part is extremely outdated... maybe someone who has access to the VMS fiche for V5.X can provide the proper update to this patch!  This document shows how you can take the DTE_DF112.EXE image distributed with VAX/VMS V4.2 or MicroVMS V4.2 and modify it to support the DEC DF224 Scholar modem. The first part applies a patch to change the module name within the old image to the name of the new image. Since the DF224 expects to see the "!" character at the end of the number, we have to replace the code that checks for the "#" used in the DF112 dialing in the second part. Since most of the modems I dial are either DF03's or DF112's, I have set my DF224 to work at 1200 baud. Two important things should be set inside your DF224. First, your characters should NOT echo. Second, you should have the full responses set. My switch pack settings in the DF224 are: SWITCH PACK #1 1 2 3 4 5 6 7 8 --- --- --- --- --- --- --- --- OFF ON ON ON ON ON OFF ON | | | | | | | | | | | | | | +---+------ Bell 212 mode | | | | | +-------------- Async | | | +---+------------------ Internal modem timing | | +-------------------------- Full modem response | +------------------------------ Auto answer +---------------------------------- Input character echo disabled SWITCH PACK #2 1 2 3 4 --- --- --- --- OFF ON OFF OFF | | | | | | | +------------ Interface speed select disabled | | +---------------- MI/MIC ground disabled +---+-------------------- Force DTR on disabled In addition, I have my DF224 plugged into TTA2: on my VAX. This is the way I initialize the line upon system startup: $ SET TERMINAL TTA2:/PERM/NOAUTO/MODEM/SYS/DIALUP/SPEED=1200 You may also want to add a line in your startup file to install the image for the DF224 dialer, and any other dialers you have on your system. The following is how I used the VMS patch utility to modify the image to work with my DF224. If you desire further insight, look in the fiche. The DTE_DF112 listing can be found on 621 C02 of the V4.2 fiche. Login to the SYSTEM account, or something with equiv privs...then... $ SET DEFAULT SYS$LIBRARY ! Let's work in here $ COPY DTE_DF112.EXE DTE_DF224.EXE ! Copy (create) new file $ PATCH/ABSOLUTE DTE_DF224.EXE ! Now lets patch the image PATCH Version 4-00 15-Sep-1984 %PATCH-I-NOGBL, some or all global symbols not accessible %PATCH-I-NOLCL, image does not contain local symbols PATCH>replace 005F ! Change the ascii OLD> 00323131 OLD> exit NEW> 00343232 NEW> exit old: 0000005F: 00323131 ! Was '112' new: 0000005F: 00343232 ! Now '224' PATCH>replace 080E ! Change the ascii here, too OLD> 07323131 OLD> exit NEW> 07343232 NEW> exit old: 0000080E: 07323131 ! Was '112' new: 0000080E: 07343232 ! Now '224' PATCH>update %PATCH-I-WRTFIL, updating image file SYS$SYSROOT:[SYSLIB]DTE_DF224.EXE;2 PATCH>exit $! $ PATCH DTE_DF224.EXE ! Now patch the code change PATCH Version 4-00 15-Sep-1984 %PATCH-I-NOLCL, image does not contain local symbols PATCH>replace/instr 00AE OLD> 'CMPB #23, (R4)' ! Change check for '#' OLD> exit NEW> 'CMPB #21, (R4)' NEW> exit old: DIAL_ROUTINE+1C: CMPB #23,(R4) ! Was '#' new: DIAL_ROUTINE+1C: CMPB #21,(R4) ! Now '!' PATCH>replace/instr 00FF OLD> 'CMPB #23, (R4)' ! Change this '#' check, too OLD> exit NEW> 'CMPB #21, (R4)' NEW> exit old: DIAL_ROUTINE+6D: CMPB #23,(R4) ! Was '#' new: DIAL_ROUTINE+6D: CMPB #21,(R4) ! Now '!' PATCH>update %PATCH-I-WRTFIL, updating image file SYS$SYSROOT:[SYSLIB]DTE_DF224.EXE;3 PATCH>exit $! $ purge DTE_DF224.EXE ! Clean up previous versions $ install add SYS$LIBRARY:DTE_DF224.EXE ! Got to be installed $! $! $! the following DCL command will do the connection $! if your phone system is DTMF, start number with T, else P (for pulse) $! end the number with the (required) '!' $! replace the TTan: with the correct tty line $! which should be set to the right speed, and modem, etc $! $ set host/dte/dial=(number:"T5551212!",modem:DF224) TTan:  ================================================================================ Note 386.0 RRD40 CDROM compatibility 2 replies EISNER::BUCK "Michael D. Buck" 16 lines 18-JUN-1989 12:57 -------------------------------------------------------------------------------- THe current Systems & Options catalog lists the following on page 5.63: ------------------------------------------------------------ RRD40 Compact Disk Drive Features: Format: Phillips/Sony CDROM standard Unifile data format ------------------------------------------------------------- I am not familiar with this standard, or for that matter, any CD standards... Is this at all compatible with CD's that can be read by PC CD machines? More exactly, can any PC CD's be read by the RRD40 ?? ================================================================================ Note 386.1 RRD40 CDROM compatibility 1 of 2 EISNER::BURNET "John Burnet" 46 lines 18-JUN-1989 13:22 -< CD-ROM formats, High Sierra/ISO, Unifile >- -------------------------------------------------------------------------------- I'm not a CD-ROM expert, so this is just a very rough summary of some aspects of CD-ROM recording formats: DEC's and the industry's. If I've botched anything, someone please correct me. The Philips/Sony recording format, which is a de-facto industry-wide standard, specifies the *physical* characteristics (pit and land size and spacing) of the recording. The Unifile standard (which was developed by DEC; I don't know whether DEC considers it proprietary or not) specifies a *file-system organization* for the information on the disk. Unifile assumes that the physical stuff is taken care of by a lower "layer" -- the drive and controller. There is, or will be, an ISO standard for file-system organization on CD-ROMs (can't remember the number offhand); it's based on the High Sierra Group proposal. This will be incompatible with Unifile; I think that the people participating in the High Sierra group considered Unifile to be "too DEC-like" (tied too closely to ODS-2). However, I believe that the existing Philips/Sony physical recording standard will continue to be used, even though higher recording densities are probably possible with today's technology if a new recording scheme were to be worked out and agreed upon. To compare this to 9-track tape standards, the Philips/Sony spec is analogous to a physical tape recording method: NRZI (800 bpi), PE (1600 bpi), or GCR (6250 bpi), while using Unifile or the ISO/High Sierra standard would be analogous to using ANSI-standard labels and file organization on the tape. The best collection of CD-ROM information I've seen anywhere so far is the set of books published by Microsoft Press (three volumes): "CD-ROM: The New Papyrus"; "CD-Rom: Optical Publishing"; and (I think) "Interactive Multimedia". The second volume in particular has information about the High Sierra standard. DEC has said that they will be supporting the ISO standard when it becomes official. That, presumably, means that they'll be junking Unifile in the not-too-distant future. (That's just my guess.) OTOH, existing RRDx0 drives will still be able to read discs that are physically laid out according to the ISO standard; only the software will need to change. As for RRDx0/PC compatibility and media interchangability, the situation is pretty chaotic right now, with a number of incompatible file-organization methods being used by various publishers of CD-ROM databases and such for use with PCs. Even though two drives (e.g., an RRDx0 on a VAX and something else on a PC) might use the Philips/Sony recording format, there's no guarantee that the software on either system will be able to interpret the file-system organization of the other system's disc. I assume that there will be some sort of market shake-out after the ISO standard appears. ================================================================================ Note 386.2 RRD40 CDROM compatibility 2 of 2 EISNER::DUFF "Steve Duff (Ergodic Systems)" 8 lines 21-JUL-1989 19:18 -< From the silk purse to the sow's ear... >- -------------------------------------------------------------------------------- I have a current application that would be helped greatly by the ability to spool data to a CD WORM disk, for subsequent retrieval and review on a PC. Can anyone tell me whether this is possible using present off-the-shelf technology (ies), or is this wait-for-the-future time? Thanks in advance.  ================================================================================ Note 387.0 NEED PRACTICAL INFO on DISK DRIVES 6 replies EISNER::CLOSE 7 lines 20-JUN-1989 15:18 -------------------------------------------------------------------------------- Where would it be appropriate for me to obtain information and other comments from users of EMC, EMULEX, System Technology, etc disk drives? I want to swap 2 or 3 RA81's and ...gulp... 2 RP07's for newer, smaller, faster disks. How can I obtain suggestions and warnings? I reviewed HARDWARE_HELP conference and got technical information of EMC MAXPORT drives. Looking for more practical info on these and other drives. Thanks. Ed Close ================================================================================ Note 387.1 NEED PRACTICAL INFO on DISK DRIVES 1 of 6 EISNER::KENNEDY "Terry Kennedy" 36 lines 20-JUN-1989 19:08 -< Random ramblings >- -------------------------------------------------------------------------------- > Where would it be appropriate for me to obtain information and other > comments from users of EMC, EMULEX, System Technology, etc disk drives? You're in the right conference. A quick look did not show an all-inclusive topic like you request, so I'll continue here... > I > want to swap 2 or 3 RA81's and ...gulp... 2 RP07's for newer, smaller, faster > disks. How can I obtain suggestions and warnings? You didn't give a lot of background information, so I'll make some guesses and you can tell me where I went wrong... I'll assume VAX CPU's, probably unclustered, UDA and RA81's, as well a RH7xx and RP07. The maintenance on a RA81 + UDA50 for 8x5 Basic coverage is $148, with additional drives being $95 each. Since you've got equipment you feel is old, you don't need gigs and gigs of disk (probably). This maintenance cost is about 1/2 the price of a new drive per year, which isn't really too bad. The space per meg isn't too bad either. In my environment (educational) it is hard to justify 20K for a new drive from EMC or somebody when you can get the same capacity for 1/5 the price as used RA81's. If you really want to be up-to-the minute, con- sider some of the new DEC offerings (mixed-media SA units). I would stay away from Emulex, as they have a long-standing reputation for charging for continuing support (firmware upgrades to fix problems) or just aban- doning the product (as they did with the SC31 when VMS V5 came out). I also had a recent bad experience with EMC, but I *think* that was just at the local level. While other vendors offer great deals (does "we integrate 256 IBM PC- style hard disks in a big box that only uses one SDI port *and* we offer a lifetime warranty!" sound familiar?), remember that the company's life- time may be shorter than that of the product. ================================================================================ Note 387.2 NEED PRACTICAL INFO on DISK DRIVES 2 of 6 EISNER::HARVEY "Jack Harvey" 10 lines 20-JUN-1989 19:49 -< Questions for Questions >- -------------------------------------------------------------------------------- Is floor space a factor? Do you do self maintenance? Since you are considering third party disks, do you consider or have third party maintenance? If you do, your maintainer probably has an opinion. If you are planning on a major increase in storage (x4, perhaps) what about tape backup? Unless you already have more than adequate tape drives now, they may become an issue if your data outgrows them and/or the available hours for backups. ================================================================================ Note 387.3 NEED PRACTICAL INFO on DISK DRIVES 3 of 6 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 14 lines 21-JUN-1989 11:29 -< more details needed >- -------------------------------------------------------------------------------- Ypu didn't say if you MUST have disks that connect to the DEC UDA50 style controller, or simply any controller the you could plug into your Q bus, U bus (what DO you have?). Even if on a Q bus, do you want SCSI drives with an eye to moving them to a workstation in the future? Or do you simply want many big (eg CDC 1.23 gig Sabres) SMDs? Getting 3rd party disks thet plug into a DEC HSC (or UDA50, etc) COSTS. Buying the raw drive with an SMD interface is MUCH less $s. 4 CDC 1.23g Sabres fit in the rack space of an RA81. Please give more details about your situation. ================================================================================ Note 387.4 NEED PRACTICAL INFO on DISK DRIVES 4 of 6 EISNER::CLOSE 15 lines 21-JUN-1989 13:21 -< More details >- -------------------------------------------------------------------------------- I apoligize for being skimpy on details: Here it goes: We have a 780/8350 CI cluster with five RA81 and 3 RA82 on HSC50. One RP06 and two RP07 on 780 MASSBUS adapter. I would like to trade-in the rp06, both rp07's, and two or three RA81's (approx 2 GB) for 2 or 3 GB of diskspace that would not require more than 2 or 3 HSC disk channels (to avoid purchasing third HSC disk interface). We want to place these on the HSC. Other than the restriction of HSC disk channels, the disks must be on the GSA schedule (for US Dept of Transportation), we would prefer DEC maintenance, and other factors that are always important (price, reliability, expanibility, etc.). Floor space is not a big issue (since we are getting rid of all these washing machines). I am also evaluating 8mm tape (which is a separate topic), so I don't expect backups to be an issue too far in the future. Thanks again for your help. Any additional assistance would be appreciated. Eddie Close ================================================================================ Note 387.5 NEED PRACTICAL INFO on DISK DRIVES 5 of 6 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 3 lines 22-JUN-1989 17:17 -< Digital has it now ! >- -------------------------------------------------------------------------------- DEC has (had?) a pretty good trade in/up program for their RA90 class disks. I know they aren't the cheapest, but they meet all your requirements ! ================================================================================ Note 387.6 NEED PRACTICAL INFO on DISK DRIVES 6 of 6 EISNER::LANG 30 lines 28-JUN-1989 22:56 -< RA Disk Alternatives >- -------------------------------------------------------------------------------- There are several companies that will sell you "RA" series equivalent drives to connect to your HSC. From a recent experience in looking for new drives, I'd like to offer this information. If you lean toward third party offerings, read the fine print about what you are gettng for your money (i.e. cables...warranty etc.) and investigate the financial backgrounds of the companies that you are likely to deal with. Also, your local DEC field service unit manager can help you decide about what stuff to use that DEC will service (if you prefer DEC service) or talk to your field service provider to see what items they may support. Now, when I was looking into these drives, I was most impressed by the Imprimis 8 in. drives, I think they are designated as 9720's and come in all sorts of capacities. The problem that I had was that to connect them to your HSC, you must convert SDI (standard disk interface) commands to SMD. This means you need a protocol converter. Because of recent litigation between DEC and SI over this type of interface this winds up being a solution you might not like to get into, but at least two companies (SI excluded) make these protocol converters (they also have ESDI conversion boxes also if you like 5.25" drives.) The two companies are Lago Systems and Exsys. I wound up going with RA90's in my case, but there were other factors involved in the decision. I hope the informatio provided here is of some use. DISCLAIMER I have no affiliation with any company mentioned here.!!  ================================================================================ Note 388.0 Tape lifetime - how to determine 19 replies EISNER::MCCREADY "Gary, former DEC, DCL is Fun!" 12 lines 22-JUN-1989 20:31 -------------------------------------------------------------------------------- When is a tape "bad"? Sure, if the drive will not mount it, but how about all those "recoverable media errors" that show up after some reels of tape in backup? Are those a valid indicator of "tape fitness", or what? So how do you tell that you should trash a tape - before you are forced to? BTW - we have some recycled about 15+ times - any comparisions? --Gary McCready ================================================================================ Note 388.1 Tape lifetime - how to determine 1 of 19 EISNER::COY "Dale E. Coy" 12 lines 22-JUN-1989 23:17 -< Bathtub curve >- -------------------------------------------------------------------------------- Obviously tape lifetime depends on the "breed" of the tape, except for occasional random occurrances (eaten by drive, etc...) In general, a tape will get BETTER for about the first 5 read/write passes. From that point, it gets worse until it's unusable. In addition, tapes much prefer to be exercised regularly. My "rule of thumb" is that 4 months on the shelf is worth 1 read/write pass. My other "rule of thumb" is to throw 'em out after 13 cycles (that equates to a year if I'm rotating a set of 4 weeklys). I figure that's BEFORE they're going to die. ================================================================================ Note 388.2 Tape lifetime - how to determine 2 of 19 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 4 lines 23-JUN-1989 14:19 -------------------------------------------------------------------------------- Back when I didn't know any better I used some "daily" over and over again on a weekly rotation for several **YEARS**. I figure that some had to have seen 200 - 300 cycles and yet I could read and write them no problems ! ================================================================================ Note 388.3 Tape lifetime - how to determine 3 of 19 EISNER::COY "Dale E. Coy" 2 lines 24-JUN-1989 00:17 -< The request was for advice >- -------------------------------------------------------------------------------- Yes, they can be used for hundreds of cycles. Would you recommend it? ================================================================================ Note 388.4 Tape lifetime - how to determine 4 of 19 EISNER::CONROY "Alan Conroy" 19 lines 24-JUN-1989 04:32 -< The good, the bad, and the questionable >- -------------------------------------------------------------------------------- > When is a tape "bad"? Sure, if the drive will not mount it, but A tape is bad when you get parity errors that bomb backup. It is possible to write a tape with no hard errors, but for the tape to be unreadable at some later date. My understanding of VMS is that if it encounters a write error, it tries to rewrite the record several times. If that fails, it writes a long Inter Record Gap (hopefully OVER the entire bad spot) and continues. At least, that's how RSTS did it. However, unlike some IBM systems which will just write longer and longer IRGs, VMS only tries this once and then gives up. (If I'm wrong, I trust that someone "in the know" will set me straight.) Of course, the real answer depends upon how redundant your backups are. If you include lots of CRC blocks, and do lots of backups, then you're probably safe using the tapes right up until they bomb the software. Of course, if you're not being overly cautious, this might be a bad way to go. For myself, I've NEVER lost any data because of a bad tape. Incidentally, keeping the tape drive heads and path clean will cure about 50% of your potential I/O problems, and stripping the ends of "bad" tapes will cure another 45%. The other 5% should just be thrown out. And of course, any long-term archive should be on the most reliable media that makes sense. ================================================================================ Note 388.5 Tape lifetime - how to determine 5 of 19 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 35 lines 24-JUN-1989 05:15 -< "I tell you three times..." >- -------------------------------------------------------------------------------- >. My understanding of VMS is that if it encounters a write error, it >tries to rewrite the record several times. If that fails, it writes a long >Inter Record Gap (hopefully OVER the entire bad spot) and continues. At least, >that's how RSTS did it. However, unlike some IBM systems which will just write >longer and longer IRGs, VMS only tries this once and then gives up. (If I'm >wrong, I trust that someone "in the know" will set me straight.) During normal writes to the tape drive the drive will indeed try to rewrite the record several times, and if that fails writes a 6" IRG and tries again. This is a function of the drive and controller, not the operating system. The drives I have seen try this for several iterations, but not forever, before giving up. (By the way, "write errors" are detected by the second set of heads: The data is read immediately after being written, and the read-back data is compared to what the drive was told to write.) VMS BACKUP does things a bit differently. Above I said "during normal writes". Backup does its writes with the inhibit retry QIO modifier; this causes the driver to tell the drive to simply report write errors and not try to correct them by backspace-and-rewrite. Instead, When such an error is detected, BACKUP simply keeps going, and repeats the write of the same block. So now you have two (or maybe more) copies of the same block on the tape; the last one is the one that got written correctly. The idea is that if a write doesn't work the first time, we don't want to depend on that section of tape at all -- after all, there's plenty more tape available; better to waste a bit of tape than to depend on a flaky section. The blocks are numbered so that Backup can tell them apart, and backup's CRC check tells backup which edition of a particular block is good. VMSTPC users, please take note: When you make a copy of a BACKUP saveset with VMSTPC, you are defeating this safety mechanism. VMSTPC does "ordinary" writes to the tape, allowing the drive's retry mechanism to attempt to correct write errors by retrying on the same spot on the tape. Probably okay for DECUS software distribution -- after all there are lots of copies of it available -- but I wouldn't trust vital data to it. ================================================================================ Note 388.6 Tape lifetime - how to determine 6 of 19 EISNER::KENNEDY "Terry Kennedy" 20 lines 24-JUN-1989 18:54 -< Well, sort of, most of the time... >- -------------------------------------------------------------------------------- > VMS BACKUP does things a bit differently. Above I said "during normal writes". > Backup does its writes with the inhibit retry QIO modifier; this causes the > driver to tell the drive to simply report write errors and not try to correct > them by backspace-and-rewrite. Instead, When such an error is detected, BACKUP > simply keeps going, and repeats the write of the same block. So now you have > two (or maybe more) copies of the same block on the tape; the last one is the > one that got written correctly. This behavior is modified when you use the /INTERCHANGE qualifier. When it is used, behavior reverts to the 'normal' mode you describe (although still controlled by BACKUP). These duplicate blocks caused me no end of problems when I was writing my VMS BACKUP for IBM PC software product. I finally figured it out, but I still think it's bizarre. Backing up, writing a long gap, and then rewriting gives you a logically cleaner program. The only reason I can see for doing what they do is because the TK25 did it that way, so they left it in. It certainly can't be due to the need to stream the drive, after all we're talking less than 50 or so repositions for a tape with 20 errors, and we all know how efficient VAX BACKUP is with streamers anyway... 8-{ ================================================================================ Note 388.7 Tape lifetime - how to determine 7 of 19 EISNER::KILGALLEN "Larry Kilgallen" 4 lines 24-JUN-1989 21:23 -< The reason they made Backup not backspace >- -------------------------------------------------------------------------------- is that given problems with the tape drive, they don't necessarily trust that a backspace is going to go back the right amount and not wipe out something before the block which had the problem (and is going to be rewritten). ================================================================================ Note 388.8 Tape lifetime - how to determine 8 of 19 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 24 lines 28-JUN-1989 18:54 -< We do hundreds with no dread... >- -------------------------------------------------------------------------------- >> In general, a tape will get BETTER for about the first 5 read/write >> passes. From that point, it gets worse until it's unusable. True for most tapes. If you get tapes which have been polished before shipment (e.g., Memorex), you will not see much of this effect. We have some tapes that have been through hundreds of cycles, but what we do is to keep track of recoverable errors on our Backups and when they begin to get out of hand (6 or so, but we run large blocks), we clean and verify the tape. This will nearly always get the error count back down to 1 or less. For our non-Backup tapes (plots, off- site transfers, etc.) we insist that these tapes be cleaned and veri- fied after 25 uses. Our cataloged tapes, containing user data which is accessed infrequently, are cleaned and verified when they are released by the user if they have been in storage more than 6 months. We dump any tape that does not show a reduction in errors after cleaning and verification. If you are not into tape cleaning (yourself or a service), then a policy of dumping tapes that show too many errors on Backup, or after a fixed number of usages, like 25 or so, would make a lot of sense. But, figure out how many tapes you would then reasonably dump in a year, and price a cleaning machine or service. ================================================================================ Note 388.9 Tape lifetime - how to determine 9 of 19 EISNER::SMITH_PA "Paul Smith" 7 lines 28-JUN-1989 22:31 -< 3M Blackwatch Tapes? >- -------------------------------------------------------------------------------- I have heard mixed reviews of the 3M Blackwatch tape. Some FEs blame it for all drive problems; others say its OK. I get it because it comes in 3600' reels, which does my full backups nicely on two reels. Any qualified opinions? ================================================================================ Note 388.10 Tape lifetime - how to determine 10 of 19 EISNER::KENNEDY "Terry Kennedy" 2 lines 29-JUN-1989 18:38 -< Look here in topic 213; much information >- -------------------------------------------------------------------------------- > Any qualified opinions? ================================================================================ Note 388.11 Tape lifetime - how to determine 11 of 19 EISNER::PROVOST 11 lines 7-JUL-1989 14:22 -< SKINNY tape >- -------------------------------------------------------------------------------- re: < Note 388.9 by EISNER::SMITH_PA "Paul Smith" > > -< 3M Blackwatch Tapes? >- >I get it because it comes in 3600' reels, which does my full backups nicely on >two reels. I think my comments on back-coated tape are posted elsewhere, but I have heard problems over the years regarding 3600' reels on a variety of tape drives. The media is very thin. It tends to stretch, as mentioned elsewhere, but it also tends to slip on the capstans. If you have friction capstans you can get errors unless the capstans are excessively cleaned. ================================================================================ Note 388.12 Tape lifetime - how to determine 12 of 19 EISNER::PRIGOT "Jonathan M. Prigot" 9 lines 10-JUL-1989 10:32 -< Timeout errors >- -------------------------------------------------------------------------------- The other problem that I have heard of with 3600' reels is that the tape is too long! I seem to remember reading somewhere that either the drive's microcode (for TU8x and TA8x drives) or the device driver (for just about every drive) knows how long it would take to go from one end of a 2400' reel to the other. If that time is exceeded, the drive reports some kind of error. We use 2400' reels on our VAXen and 3600' reels on our DECsystem which really doesn't care. ================================================================================ Note 388.13 Tape lifetime - how to determine 13 of 19 EISNER::MADISON "Matt Madison" 9 lines 11-JUL-1989 19:49 -< DEC says "no" to 3600', Black Watch >- -------------------------------------------------------------------------------- On the subject of Black Watch and 3600' tapes... A letter just came the other day from our local FS group "reminding" us that DEC only supports ANSI X3.whatever-compliant tapes in their tape drives and mentions back-coated and 3600'-reel tapes as being specifically excluded from being supported. I was rather mystified by this unprompted "reminder", since we don't use either. Thought it was interesting though. ================================================================================ Note 388.14 Tape lifetime - how to determine 14 of 19 EISNER::HASSINGER "Bob Hassinger" 15 lines 12-JUL-1989 11:08 -< Really? Maybe a first! >- -------------------------------------------------------------------------------- Was that actually identifiable as having come from DEC? With a signature, letter head, and so on? I have been trying to get DEC to actually take a documented stand on the backcoated part of this for a number of years. Every time I try to pin them down they slip away. The best I was ever able to do was to get a piece of paper that said things like that but which had no identification as to where it came from, who was saying it, etc. I appeared they where never prepared to take a stand that someone could call them on - 3M I suppose. Until DEC gives me something like that I have no way to go back to the corporate purchasing people and show them why they need to switch the bulk order or to go to the 3M and ask them to settle the issue with DEC (and maybe ANSI too). ================================================================================ Note 388.15 Tape lifetime - how to determine 15 of 19 EISNER::MADISON "Matt Madison" 2 lines 12-JUL-1989 19:44 -< Yes >- -------------------------------------------------------------------------------- > Was that actually identifiable as having come from DEC? With a > signature, letter head, and so on? ================================================================================ Note 388.16 Tape lifetime - how to determine 16 of 19 EISNER::KENNEDY "Terry Kennedy" 51 lines 12-JUL-1989 21:01 -< It's been here for a year >- -------------------------------------------------------------------------------- > I have been trying to get DEC to actually take a documented stand > on the backcoated part of this for a number of years. > Until DEC gives me something like that I have no way to go back > to the corporate purchasing people ... Bob, that information has been on this system for almost a year (8 days short of a year, actually). As a matter of fact, it was posted in response to your request! Since that time I've seen additional material in DEC manuals for the TU/TA 78/79, as well as other drives. But, the one reference should be enough... <<< EISNER::DUA0:[NOTES$LIBRARY]HARDWARE_HELP.NOTE;1 >>> -< HARDWARE_HELP >- ================================================================================ Note 182.9 Religious wars, Mag Tape division 9 of 22 EISNER::KENNEDY "Terry Kennedy" 31 lines 20-JUL-1988 17:38 -< DEC document # disclaiming back-coated tapes... >- -------------------------------------------------------------------------------- > Until DEC is willing to stand behind their claims that Blackwatch (or > any back coated) tape is bad, we will have no choice except to continue > using it. See the following extract from the Pageswapper system: <<< NODE::DISK$PAGESWAPPER:[000000]IO.NOTE;3 >>> -< Submissions to the Pageswapper INPUT/OUTPUT Column >- ================================================================================ Note 971.29 Master versus slave! 29 of 32 NODE::US192903 "Terry Kennedy" 26 lines 30-APR-1988 07:26 -< I hope you have a large dumpster! >- -------------------------------------------------------------------------------- [material not related to back-coated discussion deleted] > We *always* use back-coated tapes. Two references: o "Tape Requirements for IBM 1/2" Tape Units at 800, 1600 and 6250 BPI", IBM Pub. # GA32-0006-5. Page 6, Paragraph 9 - "Fractional Drag - Magnetic Coating to Back Surface" - BW tape fails this specification. Page 7, Paragraph 12 - "Coating In- tegrity Failure" - 3 of 25 reels of BW exhibited failure at or immediately before load point after 100 hours of use o "TU80 Subsystem User's Guide", DEC Pub. # EK-0TU80-UG-002. Page 1-2, "Tape Media Requirements" - "...To date, there are no back-coated magnetic tapes that meet DIGITAL's specification..." ================================================================================ Note 388.17 Tape lifetime - how to determine 17 of 19 EISNER::HASSINGER "Bob Hassinger" 7 lines 13-JUL-1989 11:26 -< No, never resolved - for my need anyway. >- -------------------------------------------------------------------------------- > -< It's been here for a year >- Yes, I recall that thread well. If you review it further you will note 182.12 and 182.17. We never got to where we had a documented statement from DEC that 1) clearly applies to my TU77, and 2) is available to the normal customer. I think we still ended up talking about a DEC spec which we could not identify a viable source for. ================================================================================ Note 388.18 Tape lifetime - how to determine 18 of 19 EISNER::KENNEDY "Terry Kennedy" 13 lines 13-JUL-1989 23:29 -< You asked for it - you got it! >- -------------------------------------------------------------------------------- > 1) clearly applies to my TU77, and 2) is available to the normal customer. The TU77 User's Guide (EK-OTU77-UG-02, 1984) states the same quote on page 1-2 of the manual, and ends "As a result of these problems, back-coated mag- netic tapes are not recommended for the TU77 tape drive." That *should* be explicit enough for your purchasing folks. I also found the earliest reference to the problem. It is on fiche EP-TTMIS- TI-F. Reference "Tapes TT-11 (Tech Tip 11)". The Tech Tip is marked "Company Confidential" so I won't post it here, but it is dated 7-9-79. You might ask your service rep to show it to you (or your purchasing agent), but I really don't see why the TU77 manual wouldn't be enough. ================================================================================ Note 388.19 Tape lifetime - how to determine 19 of 19 EISNER::HASSINGER "Bob Hassinger" 4 lines 14-JUL-1989 10:51 -< Thanks >- -------------------------------------------------------------------------------- > The TU77 User's Guide (EK-OTU77-UG-02, 1984) states the same quote on page As it happens we got our TU77 in 1982 so we don't have this one but I will see what we do have.  ================================================================================ Note 389.0 DMF32 Flow Control 2 replies EISNER::DOHERTY "Bob Doherty" 11 lines 22-JUN-1989 23:43 -------------------------------------------------------------------------------- Does anyone know whether the full modem control lines ( 0 and 1) on a DMF32 are capable of hardware flow control? I think they carry through both RTS and CTS. The reason for the question is we are setting up a Courier HST DS modem for dialout, and would like to fix the DTE speed at 19200, and let the modem negotiate withe the HST at the other end for the appropriate speed. According to the manual, this setup requires hardware flow control to prevent the DTE from overrunning the modem. ================================================================================ Note 389.1 DMF32 Flow Control 1 of 2 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 37 lines 23-JUN-1989 01:36 -< CTS yes, RTS so-so. >- -------------------------------------------------------------------------------- > Does anyone know whether the full modem control lines ( 0 and 1) > on a DMF32 are capable of hardware flow control? I think they carry > through both RTS and CTS. Yes. These ports, in concert with the VMS terminal driver, will handle CTS the way you want. That is, when the modem drops CTS, the VAX will stop sending; the VAX will resume sending when the modem raises CTS. We're using a Telebit Trailblazer this way and it works great. Somebody is sure to ask about async DDCMP. It doesn't work. async DDCMP is implemented in an alternate terminal class driver. The port driver reports modem control signal state transitions to the class driver, and it's up to the class driver to interpret them. The regular terminal class driver (TTDRIVER) does what we want here; regrettably, the async DDCMP terminal class driver does not. RTS is implemented in a less useful fashion, albeit in a manner appropriate for full-duplex lines. Which is to say that the terminal driver keeps RTS up just about all the time -- it does not raise it to warn the modem that it wants to start sending. That's okay with modern modems like the Courier HST and the Telebit, since they are smart and buffered and basically give you a full-duplex link; the modem figures out that the VAX wants to send whenever the VAX starts sending. You may find a reference in your modem manual to another use of RTS, in which the DTE raises RTS to indicate that it can take data from the DCE, and drops RTS to tell the DCE to back off. As far as I can determine, this was invented by authors of certain PC-based comm software, and some modem manufacturers have picked up on it. Whatever, the VAX does not do this. Locking the interface speed to faster than the line can provide data (which you said you were doing; that's how we use the Trailblazer, too) is part of the answer here, but if the VAX turns its back on the port for a bit (ie, stops posting reads) you can still overrun the typeahead buffer. If this is a problem, use XON/XOFF from the VAX, and make sure the modems are set up to carry this transparently to the far end DTE. ================================================================================ Note 389.2 DMF32 Flow Control 2 of 2 EISNER::DOHERTY "Bob Doherty" 27 lines 25-JUN-1989 15:23 -< Thanks, Jamie!!! >- -------------------------------------------------------------------------------- Thanks, Jamie. That is the answer I hoped for. > Locking the interface speed to faster than the line can provide data (which > you said you were doing; that's how we use the Trailblazer, too) is > part of the answer here, but if the VAX turns its back on the port for a > bit (ie, stops posting reads) you can still overrun the typeahead > buffer. We use a large alternate typeahead buffer(2k, at least I think that's large) and there shouldn't be any problem with overrun. I have transferred files from our Mac to the VAX over a hardwired line at 19200 (using ZTERM on the MAC end and rz/sz on the VAX) and have seen no problem with overruns(provided ALTYPEAHD is enabled) with effective baud rates of ca. 11,000 baud. When ALTYPEAHD is turned off on the line, all that occurs is retransmission of ZMODEM data packets. > If this is a problem, use XON/XOFF from the VAX, and make sure the modems > are set up to carry this transparently to the far end DTE. Unfortunately, since we will be using the high speed primarily for ZMODEM file transfers, we can't use XON/XOFF flow control unless we use 8 bit quoting, which would slow down the transfer greatly.  ================================================================================ Note 390.0 Q-Bus Co-Processors? 3 replies EISNER::ROBITAILLE "MRobitaille@PaxRv-NES.Arpa" 24 lines 23-JUN-1989 19:46 -------------------------------------------------------------------------------- We're (our company) is putting together the initial architecture and requirements for a network server widget for good ole Uncle Sam. We've pretty much solidified everything, but a new requirement that just came in the door (Murphy strikes again!) will cause us to really stretch our performance budgets. In trying to find a quick way out, I wondered if, on a 3400 platform, one could have a co-processor board that appeared to the main 3400 CPU as a device, accessible via device driver QIOs. Specifically, we may need to generate "encrypted checksums" for select data that goes out the network wire. If the server receives the data, and can replicate said checksum, then the server can beleive that the data hasn't been modified. Trouble is, we think that putting the software to support this in-line with the rest of the code on the 3400 may break our throughput estimates, thus our wonderment if there's a way to farm this encrypted checksumming off to an asynchronous process hidden behind a QIO call with buffer address, length, and where to write the answer. So, back to the hardware question - are there any q-bus boards out there (DEC or third party) that can be used as a slaved co-processor? Note: Yes, we've figured that the 4MB of memory on the 3400 CPU card is probably not addressable from the q-bus...one problem at a time! ;-} ================================================================================ Note 390.1 Q-Bus Co-Processors? 1 of 3 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 21 lines 23-JUN-1989 20:47 -< KXJ, KXT, Simpact >- -------------------------------------------------------------------------------- There is the KXJ card that has at least 2 possibilities. You can run CPRSX (CoProcessor RSX) which is M+ on it, or you can simply get the micro-power pascal + related tool kits, etc and have a dedicated (15 mhz J11) engine without M+. The card itself also has a external comm ports, that aren't really supported under CPRSX (you ARE welcome to write a driver and load it). CPRSX isn't expecting you to use the card as a front end comm card with comm port, but really as an RSX M+ engine using VMS for terminal and other device I/O. There is a KXT (T11 rather than J11 chip) that will run MPP, but not CPRSX. It costs less, does less. There is someone's smart-ass comm support chip on there, too. It does CRC and related stuff, and might be of use. Actually the RSX group ISN'T the only 11 op sys using KXJs as coporcessors in a VMS world. Either or both of RSTS and RT should be checked. I just don't remember. Ask Simpact what they have, if you want a 3rd party board. ================================================================================ Note 390.2 Q-Bus Co-Processors? 2 of 3 EISNER::TINKELMAN "Bob Tinkelman - CCA" 8 lines 24-JUN-1989 08:56 -< KXJ supported by MPP, RSX and RT >- -------------------------------------------------------------------------------- RE: < Note 390.1 by EISNER::BRUCE "Barton F. Bruce - CCA/MA" > > Actually the RSX group ISN'T the only 11 op sys using KXJs > as coporcessors in a VMS world. Either or both of RSTS and RT should be > checked. I just don't remember. The RT group supports the KXT/KXJ - and up to 14 of them (the hardware limit), unlike the RSX product. ================================================================================ Note 390.3 Q-Bus Co-Processors? 3 of 3 EISNER::CROWELL "Shefth of the Fourth Order" 26 lines 25-JUN-1989 18:45 -< Probably not 14 >- -------------------------------------------------------------------------------- > The RT group supports the KXT/KXJ - and up to 14 of them (the hardware > limit), unlike the RSX product. Actually, while the CoProcessor/RT-11 product (when it comes out) can support 14 boards on VMS or 8 on RT-11 hosts, official support is likely to be limited to 4 in "known" VAX systems because of power supply and cooling problems. For this application, though, you might be more interested in the BOOT/IMAGE mode of CoProcessor/RT-11 (as described by one of the RT-11 team a couple of DECUS's ago) whereby instead of booting the RT-11 operating system on the KXJ, you "boot" a completely independent RT-11 .SAV image which runs without benefit of any operating system. This way a dedicated application can be run without the overhead of an operating system, boot device, etc. The most likely scenario is the BOOT/STANDALONE mode, wherein RT-11 is booted on the CoProcessor with it's own memory serving as a RAM disk system device (VM:). Then your application on the VAX with QIO's to the driver, talks back and forth to the encryption application on the KXJ through the corresponding handler on the RT side. It all just sorta works. Although they've described the product and its use at DECUS, the RT-11 guys have not made any predictions of when it'll be available.  ================================================================================ Note 391.0 SYSTEMS INDUSTRIES (SI) 6 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 16 lines 24-JUN-1989 13:34 -------------------------------------------------------------------------------- This is a question for a friend. He is a software person who does not get involved with hardware. His hardware people are installing SI products. 1. What is your opinion of SI products? Why and what experiences have you had? 2. How is their customer support? 3. How does the product work from OS version to OS version 4. What DEC hardware are you running it on? 5. What other 3rd party did you consider? and why did you pick SI products? ================================================================================ Note 391.1 SYSTEMS INDUSTRIES (SI) 1 of 6 EISNER::CONROY "Alan Conroy" 36 lines 25-JUN-1989 02:26 -< SI: Mixed feelings >- -------------------------------------------------------------------------------- > 1. What is your opinion of SI products? Why and what > experiences have you had? Though told by SI that the 9400 disk controller was faster than Emulex SCxx disk controllers, my own benchmarks proved otherwise. This destroyed any faith I had in SI's claims (either they lie or they just don't know - either way I don't trust what they say). I don't know how the 9900 stacks up against other controllers. The hardware LOOKS klugey to me, but in all fairness I have to say that (when working) they do seem to go on and on... > 2. How is their customer support? Do they have customer support?!? As I have explained elsewhere, my MAJOR problem with SI is that they refuse to be of any service to our third-party maintenance people - THEY want all the money for support, and if that means being unhelpful to CDC, Sorbus, or whomever, then that is exactly how they are. On the other hand, any time I've had a question about their hardware (version numbers, etc) they have always been helpful. However, my overall impression is that SI is a Mickey Mouse operation (my apologies to Disney). > 3. How does the product work from OS version to OS version Couldn't say we've had any problems thus far. I'll let you know when we go to VMS V5 on those machines... > 4. What DEC hardware are you running it on? VAX-11 series, PDP-11s. > 5. What other 3rd party did you consider? and why did you pick > SI products? *I* didn't choose them. They were installed before I arrived (otherwise they wouldn't be). Frankly, I'd go with Emulex if I could do it over. ================================================================================ Note 391.2 SYSTEMS INDUSTRIES (SI) 2 of 6 EISNER::HAMPARIAN "Don Hamparian - IDI" 6 lines 26-JUN-1989 12:13 -< SI's Field service >- -------------------------------------------------------------------------------- We have had trouble getting their field service to be responsive. They don't seem to understand that if a disk is down for 48 hours that that effects anyone. Disclaimer: This is my own personal opinion - not that of my company. ================================================================================ Note 391.3 SYSTEMS INDUSTRIES (SI) 3 of 6 EISNER::CLOSE 22 lines 26-JUN-1989 14:46 -< Worth somw consideration >- -------------------------------------------------------------------------------- < 1. What is your opinion of SI products? Why/what experiences have you had? In general, SI delivers quality products. I wouldn't necessarily believe all "pipe-dreams" a sales rep may try to sell you. If a product is delivered and you can check references you should be safe. < 2. How is their customer support? From reading the other replies, I must have been lucky. SI doesn't have the resources that DEC has (mainly Field Service), but spares were not a problem (we had several 9900 disk controllers, several EAGLES, STC tape drives, and QDA50 disk controllers). < 3. How does the product work from OS version to OS version Before DSA disks, we had to patch the disk/tape drivers but now I don't think thats a problem. SIMACS (a software "clustering" product prior to VMS 4) was great under VMS 3 but had its problems under VMS 4. I don't think SI has any bigger problem with OS compatibility than EMULEX (especially comm gear) or other third-party vendors. I have not used SI equipment in over 6 months since I switched jobs. But in general, I would recommend that your friend entertain SI as a alternative to DEC. The law suits (oops? did I scare someone) would not have any impact on my decision making. Ed Close ================================================================================ Note 391.4 SYSTEMS INDUSTRIES (SI) 4 of 6 EISNER::JENSEN_M 35 lines 27-JUN-1989 02:15 -< my $0.02 >- -------------------------------------------------------------------------------- > 1. What is your opinion of SI products? Why and what > experiences have you had? Haven't purchased any equipment, but evaluated their product on several rounds of purchases. In general my impression has been that they do not technically measure up to EMULEX. Their latest concoction (SI95C) is two CDC 9720-1130 drives. They chose to deepen the cylinders rather than concatenate them. This is a more complex system than decoding the cylinder address and routing to the proper drive (ala Emulex,MTI,LAGOS,etc). That solution also causes both drives to actuate on every seek - more wear and tear. The other solution allows for pre-emptive seeking (pre-placing in an anticipated location). This, I believe will be in the next generation of controllers/firmware. The only thing going for the SI solution is the savings when an IO operation crosses the normal cylinder boundary, and because of spindle- syncronization there is no rotational latency hit. The real kicker on this system is the packaging. These guys enshroud the drives in another skin(sheet metal) for no other apparent reason than to create the impression of a single drive. I visited their headquarters in April. I was told that they would be shipping their 8mm/HSC solution in a couple of weeks. As of last week, the local rep said maybe July. I believe there are severe organizational difficulties within SI. I have heard from other sources that there is a very high turnover in the technical ranks. They just laid off a signifigant portion of the sales force. The new CEO (Emery) is probably good for them. I met Duncan and was not overly impressed. For all of the above reasons, I did not move SI to my short list on my last round of disk purchases (10GB). ================================================================================ Note 391.5 SYSTEMS INDUSTRIES (SI) 5 of 6 EISNER::TIHOR "Stephen Tihor;SITE SIG,VAX SIG" 14 lines 3-JUL-1989 19:33 -< removed to avoid hassles >- -------------------------------------------------------------------------------- ================================================================================ Note 391.6 SYSTEMS INDUSTRIES (SI) 6 of 6 EISNER::HOPKINS "Al Hopkins" 7 lines 8-JUL-1989 11:31 -< Our SI Field Engineer has been responsive >- -------------------------------------------------------------------------------- | -< SI's Field service >- | | We have had trouble getting their field service to be responsive. For us, our SI FE has responded always within 4 hours and has worked round the clock when necessary with the best possible attitude, competence, and corporate support.  ================================================================================ Note 392.0 Courier HST's and DECUServe 1 reply EISNER::DOHERTY "Bob Doherty" 6 lines 25-JUN-1989 15:28 -------------------------------------------------------------------------------- Having just set up a Courier HST Dual Standard modem, I thought it might be nice to use it to control the noise on the FTS circuits I use to log on to DECUServe. No dice!! Has anyone successfully connected to DECUSERVE with Courier HST's and got MNP to come up?? Are there some magic settings?? ================================================================================ Note 392.1 Courier HST's and DECUServe 1 of 1 EISNER::KENNEDY "Terry Kennedy" 11 lines 25-JUN-1989 15:43 -< HST (but not Dual Standard) OK here >- -------------------------------------------------------------------------------- > Has anyone successfully connected to DECUSERVE with Courier HST's > and got MNP to come up?? Are there some magic settings?? I've been using the HST (original, not DS) for some time now here. In fact, we used one at the last Exec meeting. No special settings here, although that may have changed in the DS. I have about 50 DS on order; when they arrive I'll test for this. DEC modems fail the Microcom MNP compliance tests at this time, there are known problems with them. You might look in the HWST manual under the settings for S15=64 and see if that helps.  ================================================================================ Note 393.0 UNIBUS CARD INSTALLATION HELP 5 replies EISNER::CLOSE 7 lines 27-JUN-1989 13:46 -------------------------------------------------------------------------------- I need to take out an IVECS term controller module from the UNIBUS on our 780. I was told I needed to replace it with a grant continuity card (DEC part# G727). I got the card and it seems to be dual sized card. In which slot would I place it and why? There seems to be three locations on this one backplane slot that I could put it. I notice that there is a lot of smaller grant cards in the center slot. Is this correct and why? Does any of this make sense? As always, thanks for your help. Ed ================================================================================ Note 393.1 UNIBUS CARD INSTALLATION HELP 1 of 5 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 34 lines 27-JUN-1989 15:14 -< G727 goes into slots C+D in Unibus backplane >- -------------------------------------------------------------------------------- > I need to take out an IVECS term controller module from the UNIBUS on our >780. I was told I needed to replace it with a grant continuity card (DEC >part# G727). I got the card and it seems to be dual sized card. Maybe you were told to use a 727 (the old smaller ones you found there), but you have a G7273 probably. It is handier. >In which >slot would I place it and why? There seems to be three locations on this >one backplane slot that I could put it. I notice that there is a lot of >smaller grant cards in the center slot. The smaller 727s are in slot D. Note the etch on the left side of these. These pass the different priority grants on through. Note the same pattern on the lower finger of your G(for Green)7273. Does the same thing. Now note on the top right of the upper finger that there is one little jumper. This passes the NPG grant. Remember the wire wrapped jumper you have to REMOVE if you are installing a card that uses DMA, and that MUST be replaced when that card is removed? Well this jumper 'REPLACES' it. Some F/S folks insist on re-wrapping the backplane lest someone later stuff a non-DMA card in there that lacks that jumper, or even stuffs an old style 727 in. Some other folks simply take note which slots have jumpers, and only change them when absolutely necessary, but always use g7273 grant cards. Your IVECS card is for Bridge terminal servers, and may have been made by ABLE, but certainly emulated DEC DMA terminal MUXes, and so you need to use the G7273 card unless you wire-wrap and use a 727. Clear as mud! Right. Anyway, slots C+D (the center 2) are where you stuff it. Having the handle is the greatest boon. No more scratch knuckles trying to access 727s in a crowded box with a rasp grade card back adjacent. ================================================================================ Note 393.2 UNIBUS CARD INSTALLATION HELP 2 of 5 EISNER::KENNEDY "Terry Kennedy" 14 lines 27-JUN-1989 21:26 -< Modify M9202 to ease Unibus fault isolation >- -------------------------------------------------------------------------------- > Having the handle is the greatest boon. No more scratch knuckles trying > to access 727s in a crowded box with a rasp grade card back adjacent. Well, here's another Unibus hint (not too off-topic, I trust). When you miss the NPR wire or think you have a bad controller in a Unibus segment, you normally grab a 10' piece of Unibus cable and rip your hands to shreds trying to bypass the backplane with the problem. The solution: Get a spare M9202 (the backplane segment jumper card) and drill out the 4 rivets holding the 2 boards together. Remove the rivets and the spacers and you now have 2 full-height handles on a 2' piece of Unibus cable. PS - I've been told I should patent this and become rich - oh well... ================================================================================ Note 393.3 UNIBUS CARD INSTALLATION HELP 3 of 5 EISNER::GLEASON "CyberPunk" 3 lines 28-JUN-1989 01:37 -< Just curious >- -------------------------------------------------------------------------------- What will you be using for terminal connections after you remove the IVECS? ================================================================================ Note 393.4 UNIBUS CARD INSTALLATION HELP 4 of 5 EISNER::CLOSE 7 lines 28-JUN-1989 14:12 -< Fast correct help is appreciated! >- -------------------------------------------------------------------------------- Thank you all for your expeditious help. I placed the G727 in the C/D slot in replace of the IVECS board and presto.....no problem. Thanks again. > What will you be using for terminal connections after you remove the IVECS? Virtual connections via TCP/IP (Process Software) over Ethernet. We are getting the IVECS board fixed. Thanks. Ed Close ================================================================================ Note 393.5 UNIBUS CARD INSTALLATION HELP 5 of 5 EISNER::PROVOST 15 lines 7-JUL-1989 14:19 -< F-F UNIBUS jumper >- -------------------------------------------------------------------------------- > Well, here's another Unibus hint (not too off-topic, I trust). When you >miss the NPR wire or think you have a bad controller in a Unibus segment, >you normally grab a 10' piece of Unibus cable and rip your hands to shreds >trying to bypass the backplane with the problem. > The solution: Get a spare M9202 (the backplane segment jumper card) and >drill out the 4 rivets holding the 2 boards together. Remove the rivets >and the spacers and you now have 2 full-height handles on a 2' piece of >Unibus cable. Another hint I have found useful. We have several external boxes with UNIBUS running between. In isolating faults we would like to jumper around a box. The above solution uses the wrong sex for this. We have used two OMNIBUS connector blocks (from PDP-8 days) as a U-U connector. Watch out for inversion of one of the cables.  ================================================================================ Note 394.0 Dual port HSC50 to KDB50 1 reply EISNER::ROBERTS_D "don roberts #241" 10 lines 27-JUN-1989 14:06 -------------------------------------------------------------------------------- We just got our 6310 and HSC50 (yes, I know, HSC40 is better). We currently have an 8350 with KDB50 and RA series disks. The question is, can we dual port between the HSC50 and the KDB50? Seems kinda wasteful to have those KDB50's and not use them, and it would give us a fail-safe for the HSC. I realize we would probably lose cluster-wide access to the disks, but we would retain some access to the disks with this configuration. If this can be done, what manuals detail the setup or what sources can I turn to to learn about how to do this? ================================================================================ Note 394.1 Dual port HSC50 to KDB50 1 of 1 EISNER::FULLER_S "Funky cold medina" 26 lines 27-JUN-1989 15:49 -------------------------------------------------------------------------------- > We currently have an 8350 with KDB50 and RA series disks. The question > is, can we dual port between the HSC50 and the KDB50? Seems kinda Yes, as of VMS V5, you can dual port between the HSC and KDB with automatic failover. However, one thing to keep in mind is that the RA drive will only talk to one controller at a time. Therefore, the 6310 would have to serve the drive(s) to the 8350. I'm not sure you'd be able to boot the 8350 via the 6310. Anyway, if you have the 6310 serving the drives to the 8350 (via LAVC), if the 6310 crashes and the 8350 fails the drives over to its KDB, then the 6310 will not be able to boot, since the HSC won't be able to talk to the drive(s). In other words, yes you can dual port, but you probably won't want to. You might be better off getting rid of the KDB and replacing it with a CI port so the 8350 can directly access the HSC. The HSC is actually very reliable. Stu > wasteful to have those KDB50's and not use them, and it would give > us a fail-safe for the HSC. I realize we would probably lose > cluster-wide access to the disks, but we would retain some access > to the disks with this configuration. >  ================================================================================ Note 395.0 Bad Case of Screen Jitters Solved 32 replies EISNER::RENES "Stuart Renes, DFWLug Chair" 49 lines 29-JUN-1989 09:15 -------------------------------------------------------------------------------- What follows here is a tip on solving a rather simple but obscure problem with a "jittery" gray-scale monitor: I found to my horror that my brand new MAC IIx system, complete with Apple (Sony) gray-scale monitor had a case of screen jitters depending on where in my house it was sitting. After trying my monitor on a friends MAC IIx system and finding it stable as a rock, I became suspicious and moved my entire system to the exact spot where the test machine was. My system was also stable. I found out that the instability was only evident in one end of the house (right where the machine NEEDED to be , naturally!). Sometimes, the problem was almost non-existent, other times it was very noticable. To shorten this story, the problem has turned out to be due to a voltage drop across the neutral and safety ground of the AC circuit that feed that end of the house. My measurement revealed the drop to be about .8 volts (as compared to .1 or .2 on circuits at the other end of the house). This, I believe, causes several amps of ac current to flow in the wonderful metal chassis surrounding the Sony monitor guts (due to FCC shielding reqts.) which beats against the refresh frequency (78 hz.?) of the monitor causing the edges of the screen to "jitter". I understand that some home electrical wiring contractors take short-cuts and do not run a separate neutral from each circuit breaker but sometimes daisy-chain the neutral from adjacent circuits, causing a high-resistance neutral to ground condition. This condistion wouldn't be noticed by most appliances (although I have had ground loop problems with music amplifiers on this circuit in the past). Any hints from you seasoned electricans out there (remembering the great thread on AC power) on how to "fix" this problem? I guess I could re-wire the circuit but that seems like a lot of work, given attic crawl-space, etc. Could it be that the neutral or ground at the service entrance on that circuit is just corroded or loose? Stuart. ================================================================================ Note 395.1 Bad Case of Screen Jitters Solved 1 of 32 EISNER::LEDERMAN "Bart Z. Lederman" 29 lines 29-JUN-1989 09:51 -< Are you sure your wiring is like this? >- -------------------------------------------------------------------------------- > Any hints from you seasoned electricans out there (remembering > the great thread on AC power) on how to "fix" this problem? If in fact there is not a separate neutral wire running from your outlet to the circuit breaker/fuse/distribution panel, you have a very serious code violation (not to mention fire hazard), and your insurance company might not reimburse you if your house burns down. It must be fixed. If, as I suspect, there really are separate neutrals and grounds from the outlet boxes to the panel, but one or the other is high resistance because of a bad connection in the box, or because the ground connection goes through the outside of a BX cable, or cheap cable was used and the ground connection wire broke somewhere, it would be a good idea to fix it by running a good ground. You can drop an extra wire throught the wall, or ground to a cold water pipe or something like that. If the connections are daisy chained within the distribution box, fix it immediately. You would not have to re-wire the whole house, just the box. However, every box I've ever seen has a long bus bar to hold all of the neutrals and grounds, and don't understand why an electrician would go to MORE trouble to not use it. Just how old is your house and wiring? Watch how the neutrals and grounds are connected together in the distribution box. It's o.k. for them to be connected together at the feed point into your house, but sometimes they connect them together inside the box. ================================================================================ Note 395.2 Bad Case of Screen Jitters Solved 2 of 32 EISNER::RENES "Stuart Renes, DFWLug Chair" 33 lines 29-JUN-1989 12:40 -< More details... >- -------------------------------------------------------------------------------- Thanks Bart. I suspect you are correct. I have had several problems in the past with intermittent circuit breakers (have had to toggle them several times to get them working again - a large one for the A.C. which sat all winter with no load). I'll open the distribution box and inspect the neutral and ground buss bars. Maybe the connection(s) are just bad in one of the outlet boxes in that circuit string. Guess I'll have to locate all the boxes and check each one with a meter to see where the voltage drop starts? The info on the daisy-chained neutral came from a friend who had this very problem in a commercial bldg. He had a PC monitor that would jitter on one side of a room and not on the other. An electrician found the neutral problem and ran a new circuit all the way back to the panel and fixed the problem. I have an earth ground (6 ' copper ground rod) that's tied to a VAX 725 system ground in this same room. I believe that outlet is on this circuit but I'll have to verify. At any rate, the neutral/ground potential on THAT outlet was 0.7 volts too! Guess I've got some work to do. Stuart. :-( ================================================================================ Note 395.3 Bad Case of Screen Jitters Solved 3 of 32 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 47 lines 29-JUN-1989 18:00 -< a few power thoughts >- -------------------------------------------------------------------------------- A common neutral may be used for what is called a multi-wire branch circuit that has 2 (or even 3 in a 3 phase environment) hot wires that are from different phase legs. This, with symetrical loading, reduces the neutral loss to 0 volts. Some loss across the hot wire and the neutral between where you are using the power and where these wires are fed is legal, and normal. Without grabbing a code book, I am sure your loss is legal. You are allowed a few % loss on feeders and a slightly different few % on branch circuits. You simply can't have NO loss while you draw any current. The only question is how much. Presumable the ground wire is carrying NOTHING, and so reads the voltage at the source end of the neutral where the neutral and ground are legally joined. If you are fed from a branch panel, there is feeder loss to it from the main panel (where the neutral and ground are common) that might be in addition to your branch ckt loss (depending on phase). With NO load on your branch ckt, any neutral to ground voltage is from feeder loss and/or stray current flowing in the ground wiring. Grounds can be and are bonded to all sorts of building metal, and there ARE ground potential differences within ANY large building. Is your terminal equipped with a dual voltage powersupply (120/240)? Is the neutral on the longer slot and the hot on the shorter at the outlet? Is something ELSE on that circuit that 1) accounts for the drop you see 2) could be putting noise (low freq, even) on the branch ckt? Is this display connected to something powered from a different branch ckt? Stringing a long wire from one point to another in a room to try to read voltage from one point to another gets all sorts of RF garbage picked up, too. Interferance problems are often best tracked with a broadband inductive pickup (like a clamp on ammeter) and a spectrum analyser. Looking at your display's various wires and, grounds (seperated from normally adjacent signal or power wires) might prove interesting. But you shouldn't be doing the manufacturer's engineering for them. What do they say? To get virtually 0 volts between neutral and ground, put an isolation transformer right at the unit, and the secondary neutral will be grounded right there. You can get fancier ones that have more shielding and have a regulated voltage output. It will be interesting to hear what else you discover. ================================================================================ Note 395.4 Bad Case of Screen Jitters Solved 4 of 32 EISNER::MERRIMAN "George Merriman - CCA/NY" 33 lines 30-JUN-1989 22:36 -< I'll bet on an external field of a ground-loop >- -------------------------------------------------------------------------------- > This, I believe, causes several amps of ac current to flow in the > wonderful metal chassis surrounding the Sony monitor guts Not likely, unless the Sony attempts to connect the neutral conductor of the mains circuit to the ground through the chassis, in which case I would avoid operating the system without rubber gloves and gum-soled shoes. > I found out that the instability was only evident in one end of the > house (right where the machine NEEDED to be , naturally!). Is the problem a function of the location or the circuit? It is possible something in the area (an air conditioner? some kind of power supply?) is creating an alternating magnetic field at that location which would have the same effect. It doesn't take much. I have seen cases where one terminal's power supply was giving the jitters to another next to it. Same with PC power supplies, etc. I suspect that if it is a circuit problem it is more likely to be the result of a ground loop than the ground-neutral potential you site. I don't have a copy of the NEC handy, but I remember they allow something like a 3% voltage drop in feeders between the service entrance and the branch circuit, which is 3.6V on a 120V circuit. You would see half of this between neutral and ground, assuming they were well bonded at the entrance. It is possible to measure ground-circuit currents by running BOTH the neutral and live conductors of an AC circuit through a clamp-on ammeter. Any current flowing in the live which does not return in the neutral will flow back through the ground and encircle the ammeter current transformer. ================================================================================ Note 395.5 Bad Case of Screen Jitters Solved 5 of 32 EISNER::HARVEY "Jack Harvey" 19 lines 6-JUL-1989 22:21 -< Try Correlating the Changes >- -------------------------------------------------------------------------------- > Sometimes, the problem was almost non-existent, other times > it was very noticeable. Since we're brainstorming and a wild idea might help, the above quote might be worth exploring. Can you correlate the variations with anything else nearby that is varying? You did mention a 725 nearby. How close? That beast had a couple of disk drives with external magnetic fields that "jittered" about quite a bit when the heads were moving. Any correlation? Here's another one to test the power ground theory. Get a healthy long extension cord and plug it into an outlet on the other side of the house. Drag it through to the preferred location and plug your monitor into clean power from the other side. If the problem goes away, you've probably got it identified. (Now how do we get you to not use the extension cord permanently?) ================================================================================ Note 395.6 Bad Case of Screen Jitters Solved 6 of 32 EISNER::RENES "Stuart Renes, DFWLug Chair" 36 lines 11-JUL-1989 10:27 -< ...investigation continued... >- -------------------------------------------------------------------------------- > Here's another one to test the power ground theory. Get > a healthy long extension cord and plug it into an outlet > on the other side of the house. Drag it through to the > preferred location and plug your monitor into clean power > from the other side. If the problem goes away, you've > probably got it identified. (Now how do we get you to > not use the extension cord permanently?) Last week I tried just that, Jack! The problem STILL occurred. I have taken my monitor to a friends home miles from the "scene" and it performed flawlessly for hours. It is beginning to become more obvious that POWER may not be the issue. My 11/725 is 8 feet away and NOT running during these tests. A MAC-PLUS and MAC-SE had no problems in these rooms, but of course their monitor circuitry was quite different than the Sony Gray scale one. I can find no direct correlation between weather or time-of-day and this problem, either. (After detailing ALL the switches and outlets on this branch circuit, I did find one missing ground and one loose neutral wire, but of course fixing these didn't solve the problem.) Right now, I've stopped investigating the problem in order for my brain to start working on the problem in the background. (My foreground brain processing of this issue is in a loop!) Thanks to all of you who offered help. Any further help will be GREATLY appreciated. I intend to solve this problem and report the solution (which will probably be sickeningly obvious) here. Stuart. ================================================================================ Note 395.7 Bad Case of Screen Jitters Solved 7 of 32 EISNER::PRIGOT "Jonathan M. Prigot" 4 lines 11-JUL-1989 11:16 -< Magnetic fields >- -------------------------------------------------------------------------------- It doesn't take much of a magnetic field to throw some monitors. Our first experience was with a guy whose battery powered "smoke eater" jittered his terminal until we moved it to the filing cabinet behind him. ================================================================================ Note 395.8 Bad Case of Screen Jitters Solved 8 of 32 EISNER::LEDERMAN "Bart Z. Lederman" 4 lines 11-JUL-1989 13:56 -< Second on Magnetic Fields >- -------------------------------------------------------------------------------- Now that you mention it, I had a DEC monitor go crazy because a CODEX modem was under it. Moving the modem 6 to 8 inches lower helped a lot. The problem only occurred when the modem power was turned on, but even so it took a while for me to realize it. ================================================================================ Note 395.9 Bad Case of Screen Jitters Solved 9 of 32 EISNER::GLEASON "CyberPunk" 4 lines 12-JUL-1989 00:08 -< (electro) Magnetic Fields >- -------------------------------------------------------------------------------- How 'bout your cable tv lead in? Anywhere near the system? Maybe a diode junction is working as a detector in the monitor. (I agree, a long shot - but maybe worth thinking about). ================================================================================ Note 395.10 Bad Case of Screen Jitters Solved 10 of 32 EISNER::RICE "Gary Rice" 6 lines 12-JUL-1989 12:44 -------------------------------------------------------------------------------- > It doesn't take much of a magnetic field to throw some monitors. > Our first experience was with a guy whose battery powered "smoke > eater" jittered his terminal until we moved it to the filing cabinet > behind him. I agree. We had similar problems with a florescent light being too close. ================================================================================ Note 395.11 Bad Case of Screen Jitters Solved 11 of 32 EISNER::RENES "Stuart Renes, DFWLug Chair" 6 lines 12-JUL-1989 17:28 -< Could be anything. >- -------------------------------------------------------------------------------- I'll check out the suggestions in .9 and .10 this weekend. (I'm considering wrapping my entire home computer room in aluminum foil! Just waiting for a sale at the local grocery store.) Stuart :-) ================================================================================ Note 395.12 Bad Case of Screen Jitters Solved 12 of 32 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 8 lines 13-JUL-1989 10:28 -< mu-metal is *lots* more expensive than aluminum foil >- -------------------------------------------------------------------------------- >> (I'm considering wrapping my entire home computer room in >> aluminum foil! Just waiting for a sale at the local grocery >> store.) Aluminum foil will help with electric fields, but not with MAGNETIC fields. For that, you would need mu-metal (used for CRT yoke shields, etc.), but don't expect a sale on that anytime soon. ================================================================================ Note 395.13 Bad Case of Screen Jitters Solved 13 of 32 EISNER::CONROY "Alan Conroy" 1 line 15-JUL-1989 03:39 -< What is mu-metal? >- -------------------------------------------------------------------------------- >with MAGNETIC fields. For that, you would need mu-metal ================================================================================ Note 395.14 Bad Case of Screen Jitters Solved 14 of 32 EISNER::MERRIMAN "George Merriman - CCA/NY" 9 lines 15-JUL-1989 20:00 -< AL foil might do something (CU would be better) >- -------------------------------------------------------------------------------- > Aluminum foil will help with electric fields, but not with MAGNETIC > fields. I believe a good conductive shield is very effective against ALTERNATING magnetic fields, which is probably the case here. The theory is that the conductive shield acts as a shorted-turn inductor an creates a self-induced field which counteracts the intruding field. ================================================================================ Note 395.15 Bad Case of Screen Jitters Solved 15 of 32 EISNER::LEDERMAN "Bart Z. Lederman" 16 lines 16-JUL-1989 17:47 -< Magnetic fields not the same as electrostatic >- -------------------------------------------------------------------------------- > I believe a good conductive shield is very effective against > ALTERNATING magnetic fields, which is probably the case here. The Sorry, in this kind of instance aluminum foil probably won't help. If the jitter in the screen is a magnetic field affecting the beam in the CRT, only a material with a high magnetic permeability will help, and aluminum foil doesn't count. Mu-metal is a special metal alloy designed for just such shielding purposes. The 'best' source is usually a discarded old Color TV, some of the better sets having a shield around the CRT. But mu-metal is VERY sensitive to stress: bending it can destroy most of it's shielding ability. Plain old sheet steel (tin plate) is not bad for this. If you want a mu-metal shield, there are companies which make them for specific applications. Or maybe look for Tempest shielding: expensive, but effective. ================================================================================ Note 395.16 Bad Case of Screen Jitters Solved 16 of 32 EISNER::HARVEY "Jack Harvey" 13 lines 16-JUL-1989 19:55 -< Mumetal >- -------------------------------------------------------------------------------- Mumetal is so called because the Greek letter mu is commonly used in engineering for the magnetic quality permeability. The alloy is 77% nickel, with some copper and chromium. A very similar alloy goes by the name Permalloy. Then there is Hypernik, Supermalloy and Hymu 80. Bert is right. These are highly specialized engineering materials that require care and knowledge to use. For most magnetic shielding, a tin can will work as well. If waving a large empty tin can around the monitor alters the jitter, then you can be pretty sure it is a magnetic field that's causing it. ================================================================================ Note 395.17 Bad Case of Screen Jitters Solved 17 of 32 EISNER::RENES "Stuart Renes, DFWLug Chair" 5 lines 17-JUL-1989 17:15 -< ! >- -------------------------------------------------------------------------------- I'll try the tin can trick tonight and report the results. Thanks. Stuart. ================================================================================ Note 395.18 Bad Case of Screen Jitters Solved 18 of 32 EISNER::RENES "Stuart Renes, DFWLug Chair" 15 lines 18-JUL-1989 14:46 -< Power Feed Path >- -------------------------------------------------------------------------------- I just discovered that I may have underground electrical service running from a power pole across the street from my house, between my house (on the computer room side!) to the alley terminating in a large (4 foot cube) transformer serving several homes. I'm now wondering if this could have any effect on my situation? I'm sure the feed is in grounded conduit. Or is it? I've seen the local power company working on that pole when we've had fuse problems in our transformer, so I'm pretty sure the feed must run between our homes from pole to xfmr. Not sure if this could or should cause problems, however. Stuart. ================================================================================ Note 395.19 Bad Case of Screen Jitters Solved 19 of 32 EISNER::KENNEDY "Terry Kennedy" 11 lines 19-JUL-1989 00:34 -< If recent, not metal; also other problems >- -------------------------------------------------------------------------------- > I just discovered that I may have underground electrical service > running from a power pole across the street from my house, > between my house (on the computer room side!) to the alley > terminating in a large (4 foot cube) transformer serving several > homes. If it's at all recent, it probably is *not* in grounded conduit, but in PVC with ground wires. Each phase is run in a seperate conduit. Also, if this is *under* you house, you should probably wonder why your title search didn't discover this (as if it fails, they *may* have the right to dig thru your house (and certainly your yard) to replace it... ================================================================================ Note 395.20 Bad Case of Screen Jitters Solved 20 of 32 EISNER::HARVEY "Jack Harvey" 20 lines 19-JUL-1989 20:14 -< Magnetic Fields are Pretty Short Range >- -------------------------------------------------------------------------------- I got the impression the wires were *under* his house, too, but on re-reading... Weeeel, it's a possibility. If they spaced the wires far apart, like a couple of feet, you might get a significant field. But they would have had to dig a wide trench and that's no fun. That transformer is a possibility too. But I'm guessing all this hardware is at least 20 feet from you, and I would be surprised if you can get any effect at that range. Of course, if they have a bad phase imbalance which they are dumping into a ground rod - no, it can't be *that* dumb. A magnetic field from the wires themselves will vary with the amount of power being used by the homes. However, the magnetic field of the transformer is pretty independent of the power going past you. Gee, I hope this isn't the cause - it would be damned difficult to shield against. ================================================================================ Note 395.21 Bad Case of Screen Jitters Solved 21 of 32 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 4 lines 20-JUL-1989 00:27 -< Just a small comment >- -------------------------------------------------------------------------------- >> -< Magnetic Fields are Pretty Short Range >- Ummm... just so we don't trample physics too much: magnetic fields obey the same inverse square rule that electric fields do. ================================================================================ Note 395.22 Bad Case of Screen Jitters Solved 22 of 32 EISNER::HARVEY "Jack Harvey" 5 lines 20-JUL-1989 20:15 -------------------------------------------------------------------------------- Right. If you like, electric fields are pretty short range, too. There has been some nonsense about people destroying tape libraries by walking past them with a magnet in their pocket. I just didn't want to leave an impression that the possibility of a field was sure trouble. ================================================================================ Note 395.23 Bad Case of Screen Jitters Solved 23 of 32 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 13 lines 21-JUL-1989 10:44 -< Mag tape is really hard to erase >- -------------------------------------------------------------------------------- >> There has been some nonsense about people destroying tape >> libraries by walking past them with a magnet in their pocket. About 10 or 12 years ago, Datamation published the results of a study on this. In brief, any magnet that was small enough to fit easily in your pocket had to be about a half-inch from the tape to destroy any data. If you ever take apart an RK07 disk pack, you will find a ring magnet about five inches in diameter (used to hold the lid on). I used to have great fun by walking up to a co-worker's terminal with the magnet concealed in a book. It would make the characters swim all over the screen. ================================================================================ Note 395.24 Bad Case of Screen Jitters Solved 24 of 32 EISNER::KENNEDY "Terry Kennedy" 4 lines 21-JUL-1989 20:22 -< Fun, but *DON'T* try this on a color CRT!!! >- -------------------------------------------------------------------------------- > I used to have great fun by walking up to > a co-worker's terminal with the magnet concealed in a book. > It would make the characters swim all over the screen. ================================================================================ Note 395.25 Bad Case of Screen Jitters Solved 25 of 32 EISNER::HARVEY "Jack Harvey" 16 lines 21-JUL-1989 22:06 -< Note of the Year >- -------------------------------------------------------------------------------- I hereby nominate Stuart Renes for the "Most Provocative Note of the Year in Hardware_Help". I can't recall having as much feedback from a problemee. (Gosh, Alan, what's a better word?) Stuart faces a serious problem (from his viewpoint, and ours by friendship) that we all can understand and empathize with: "Why is this hardware misbehaving in this part of my house?" It's a situation that almost demands on-site experts and expensive consultation, if a solution must be obtained. Yet, here we are, hundreds, if not thousands of miles away, delighting in speculating out loud about Stuart's problem. Communication has reached yet another level. [Ahem, solve his problem first. Then exult.] ================================================================================ Note 395.26 Bad Case of Screen Jitters Solved 26 of 32 EISNER::RICE "Gary Rice" 14 lines 22-JUL-1989 08:12 -------------------------------------------------------------------------------- > ...the lid on). I used to have great fun by walking up to > a co-worker's terminal with the magnet concealed in a book. > It would make the characters swim all over the screen. A co-worker did the same thing to my VR-241 (color) monitor. Afterward, all I could get in the way of color was "Rainbow"(s). Eventually, I called DEC to come in and de-gauss the unit. They said that it couldn't be done. My only recourse was to have it replaced. Turned out to be a rather expensive "joke". Gary ================================================================================ Note 395.27 Bad Case of Screen Jitters Solved 27 of 32 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 11 lines 22-JUL-1989 13:51 -< The joke should have been on DEC. >- -------------------------------------------------------------------------------- >Eventually, I called DEC to come in and de-gauss the unit. They said that >it couldn't be done. My only recourse was to have it replaced. > >Turned out to be a rather expensive "joke". Rubbish! I've used a standard TV repairman's degaussing coil (note, not the cheapies that are sold for 10 bucks or so, this item will cost around $50 these days) with good results. Just keep it away from the rear of the unit (there are some magnets around the electron guns that you DON'T want to degauss!). ================================================================================ Note 395.28 Bad Case of Screen Jitters Solved 28 of 32 EISNER::LEDERMAN "Bart Z. Lederman" 10 lines 23-JUL-1989 12:54 -< Tape erasers also work well. >- -------------------------------------------------------------------------------- If you have any bulk tape erasers, especially a good hand-held unit, it will also do a good job of demagnitizing color CRTs, and the metal cabinets and shields. Even a small tape head demagnitizer (one which runs on 110V AC, not the little battery units that pop into cassette recorders) will degaus a shadow mask if held close to the front of the screen. Gradually pull the unit far away from the unit before turning the power off, otherwise the collapsing field will re-magnitize the shadow mask. (I used to service TV sets, and still do it for myself and friends). ================================================================================ Note 395.29 Bad Case of Screen Jitters Solved 29 of 32 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 9 lines 24-JUL-1989 11:18 -< Degaussers built in now? >- -------------------------------------------------------------------------------- >> If you have any bulk tape erasers, especially a good hand-held >> unit, it will also do a good job of demagnitizing color CRTs, >> and the metal cabinets and shields. I thought that most modern color CRT terminals (and TVs) had built-in degaussers that were activated every time you turned the power on. Or is this only in the high-priced models? I remember one terminal with a "degauss" button. It sure made pretty patterns when you pressed it! ================================================================================ Note 395.30 Bad Case of Screen Jitters Solved 30 of 32 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 9 lines 24-JUL-1989 13:45 -< Built-in degaussers exist but have their limits. >- -------------------------------------------------------------------------------- >I thought that most modern color CRT terminals (and TVs) had >built-in degaussers that were activated every time you turned >the power on. They do, but a non-movable degaussing coil can only do so much. A handheld one is a must! (My VR241 required degaussing after it was moved about one mile east of its previous location. Moving it north or south probably wouldn't have bothered it.) ================================================================================ Note 395.31 Bad Case of Screen Jitters Solved 31 of 32 EISNER::RENES "Stuart Renes, DFWLug Chair" 15 lines 1-AUG-1989 17:19 -< This is getting OLD. >- -------------------------------------------------------------------------------- Well, I left for Colorado for a weeks vacation (from the screen jitters AND work!) and have now returned to both! I am going to see what effect (if any) a simple compass has when moved around various rooms in the house. (This is getting ridiculous, isn't it?) OTOH, whenever I walk in the 'haunted room' and see my mega-bucks MACsystem and LaserWriter, I stop smiling. Stuart. (amateur GhostBuster) ================================================================================ Note 395.32 Bad Case of Screen Jitters Solved 32 of 32 EISNER::PROVOST 4 lines 4-AUG-1989 11:23 -< Mu-metal substitute >- -------------------------------------------------------------------------------- We use mu-metal to shield accelerator. At one time mu-metal was unavailable, so we used stove-pipe. Worked fine. Tom  ================================================================================ Note 396.0 Benchmarks Rosetta Stone - VAX vs 680x0 & 80x86 2 replies EISNER::HASSINGER "Bob Hassinger" 87 lines 30-JUN-1989 15:18 -------------------------------------------------------------------------------- There is a a lot of benchmark information available to compare various VAXs. It is easy for us to say with some confidence how a microVAX II compares with a 6310. We also can get a great deal of benchmark information on PC class machines like 8xx86 and 68xxx based systems. It is fairly easy to compare an 80286 PC with an 80386 PC and it is even not too hard to make a fairly good comparison with a 68000 based Mac and so on. However, I for one have had very little luck finding anything that gives a basis for relating these two worlds of benchmark information to each other. This week I attended a meeting where some reasonably usable data in this area was presented. The following is an excerpt of some of what was presented. I have selected a subset of results and restated the original run timings as performance ratios normalized to the microVAX II (bigger numbers mean faster) for this note. (None of this is my work, I am just passing it on to give a starting point) The tests reported where: 1) "Osieve", apparently an "optimized" version of a test published this month in Amazing Computing and supposedly related to one Byte magazine uses. It tests integer operations and is written in C. 2) whets - Apparently based on the well known FORTRAN Whetstone test converted to C. Normally this tests mostly a mixture of floating point operations. The 68xxx tests where on various versions of the Amiga, compiled with Lattice C V5.02 and using the optimizations and libraries appropriate to the various configurations. The VAX runs where just done with standard CC compiles. The PC class runs where done after quite some difficulty getting them to compile under Microsoft C V5.1 Osieve whets microVAX II 1.0 1.0 VAXstation 3100 (PVAX) 2.6 2.5 VAXstation 3520 (Firefox) 2.6 3.1 VAX 83xx 1.4 1.13 VAX 8650 3.25 1.78 VAXstation 3600 3.25 2.1 VAX 8800 6.5 6.25 DECstation 3100 6.5 68000 7Mhz (Standard Amiga) 0.48 0.097 68000/881 14Mhz (CMI) 0.56 0.43 68020/881 14Mhz (Amiga 2500) 2.6 0.53 68030/882 25Mhz 4Mbx32bit mem (GVP) 4.3 2.1 Compac 286 0.42 0.066 WEISS 386 8Mhz 0.42 0.070 WEISS 386 16Mhz 0.85 0.157 IBM PS/2 Model 80 0.85 This was presented at a meeting of the Worcester Amiga Computer Users Group so of course the greatest interest was in the various Amiga configurations, how they compared with each other and VAXs, and an actual side by side comparison of a stock 7Mhz Amiga 2000 with the 68030/882 machine. Based just on raw CPU clock rate you would expect about a 3.58 ratio between these two for non-floating point work. The actual ratio of 9.0 in the Osieve test and an observed ratios of about 4.3 in actual runs of integer based image rendering during the meeting where explained by the 32 bit memory and related features on the GVP board which is a good bit faster and more efficient than the standard Amiga 16 bit memory. It is interesting to note that a full blown Amiga 2000 system with the GVP 68030/882, 4Mb of 32 bit memory plus 1Mb of standard memory for the display goes for somewhere in the $5K range and compares in these tests with something around a VAXstation 3600. Even a standard Amiga at a third to a half of that price compares favorably with 286 and even some 386 systems and does not look too bad against a microVAX II on non-floating point work. Overall, I was surprised at how well PC class machines seem to stack up to VAXs in these tests. Except for floating point, even the low end machines in the test are within range of low end VAXs. Full blown PC class configurations compared with VAXs a lot better than I expected. Now I want to see a test comparing the 68030/882 against a VAX with FORTRAN based floating point work. I might be able to sell the idea of getting top end, $5K Amiga to replace our 750 for number crunching... ================================================================================ Note 396.1 Benchmarks Rosetta Stone - VAX vs 680x0 & 80x86 1 of 2 EISNER::KENNEDY "Terry Kennedy" 11 lines 30-JUN-1989 18:33 -< Benchmarks require listings >- -------------------------------------------------------------------------------- It is always a good idea when these things are presented to see if a listing of the benchmarks was made available with the results. I would be interested in seeing it (but not here of course 8-{), especially considering the comment that the benchmarks were ported to the 80x86 family with a great deal of difficulty. I have not found it to be such big problem with the 80x86 C compilers, and feel that the results for those machine may have been misrepresented because of bad benchmark ports. I may be all wet on that, but that is one of the reasons the list- ings should be supplied. ================================================================================ Note 396.2 Benchmarks Rosetta Stone - VAX vs 680x0 & 80x86 2 of 2 EISNER::HASSINGER "Bob Hassinger" 5 lines 3-JUL-1989 11:47 -< Not given. This was just a start - more work in order. >- -------------------------------------------------------------------------------- They where not made available. The base code for the sieve was published as I noted. This was not presented as a rigorous set of benchmarks, only a starting point. Apparently the person doing the tests is none too favorably disposed towards the 86 family and that part seems to have been somewhat of an aside to the VAX and 68 tests.  ================================================================================ Note 397.0 MORE VERSATEC-6210 WOES. 4 replies EISNER::NIXON 12 lines 30-JUN-1989 15:50 -------------------------------------------------------------------------------- WE HAVE A DEC 6210 ON WHICH WE WOULD LIKE TO PUT A VERSATEC V-80 PRINTER, ON A SERIAL TERMINAL PORT, (DEC HAS NOT ALLOWED VERSATEC ON THE BI BUS AS OF YET). WHEN WE ATTEMPT TO PRINT TO THE VERSATEC WE SEEM TO LOOSE MOST OF THE DATA SENT TO THE VERSATEC. WE HAVE TRIED TO PLACE A LINE BUFFER BETWEEN THE PORT AND THE VERSATEC, BUT WE WERE UNABLE TO GET THE LINE BUFFER TO COMMUNICATE WITH THE VERSATEC. WOULD APRECIATE ANY INFORMATION IF ANYONE HAS HAD THE SAME PROBLEM AND FOUND A SOLUTION...... HELP...HELP.....HELP. ================================================================================ Note 397.1 MORE VERSATEC-6210 WOES. 1 of 4 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 22 lines 30-JUN-1989 20:53 -< I think Versatec uses DTR, not XON/XOFF >- -------------------------------------------------------------------------------- >> WE HAVE A DEC 6210 ON WHICH WE WOULD LIKE TO PUT A VERSATEC >> V-80 PRINTER, ON A SERIAL TERMINAL PORT, (DEC HAS NOT ALLOWED VERSATEC >> ON THE BI BUS AS OF YET). WHEN WE ATTEMPT TO PRINT TO THE VERSATEC WE >> SEEM TO LOOSE MOST OF THE DATA SENT TO THE VERSATEC. The last time I looked, Versatec used the DTR line for flow control, not XON/XOF like DEC uses. This is probably why you are losing much of your data. >> WE HAVE TRIED TO PLACE A LINE BUFFER BETWEEN THE PORT AND THE >> VERSATEC, BUT WE WERE UNABLE TO GET THE LINE BUFFER TO COMMUNICATE >> WITH THE VERSATEC. This is the easiest solution to this sort of problem. Just make sure the DEC end synchronizes with XON/OFF and the Versatec end uses DTR. Is this how you have it set up? Also, if you let us know what kind of terminal you are using to access DECUServe, we might be able to help you with your stuck shift key. (It is not polite to SHOUT at people here.) ================================================================================ Note 397.2 MORE VERSATEC-6210 WOES. 2 of 4 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 7 lines 1-JUL-1989 04:33 -< Use hardware flow control (dep. on mux type) >- -------------------------------------------------------------------------------- What kind of terminal mux are you using? If it's one that supports "full modem control", it supports the CTS line. You use a null modem (crosswire) cable between the printer and the mux, since they're both DTEs, correct? Try running the printer's DTR to the VAX's CTS. Then -- assuming that the mux is one of those that supports full modem control -- when the printer drops DTR to say "whoa! I've got enough to work on for a while!" the VAX will stop sending until the printer raises DTR again. ================================================================================ Note 397.3 MORE VERSATEC-6210 WOES. 3 of 4 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 17 lines 6-JUL-1989 17:42 -< By the numbers >- -------------------------------------------------------------------------------- How to cross DTR to CTS VAX Printer 1 - 1 Shield 2 - 3 Transmit 3 - 2 Receive 4 Request To Send 5 - 20 Clear To Send 6 Data Set Ready 7 - 7 Ground 8 Carrier 20 Data Terminal Ready 22 Ring If you tell VMS to ignore modem controls($SET TERM/NOMODEM) it won't get upset about RING, CARRIER, DSR missing and yet it will still "throttle" based on CTS/DTR going high or low. ================================================================================ Note 397.4 MORE VERSATEC-6210 WOES. 4 of 4 EISNER::HARVEY "Jack Harvey" 5 lines 6-JUL-1989 22:36 -< Beware of Flow Control Converters >- -------------------------------------------------------------------------------- On the other hand, if your MUX *doesn't* have this capability, beware of the device you try to use to convert the XON/XOFF. There are some cheapies in the PC stores that don't do it right. Try Black Box. They understand the problem.  ================================================================================ Note 398.0 IBM 3480 Tape Cartridge Compatibility 1 reply EISNER::SMITH_PA "Paul Smith" 6 lines 30-JUN-1989 16:58 -------------------------------------------------------------------------------- Does anybody know of a third party Tape Cartridge system compatible with the IBM 3480 that will attach to VAX? As I understand, it the DEC TA90 does this but requires an HSC, wis a significant extra expense when you only have a single MicroVAX system on which you need the tape drive. ================================================================================ Note 398.1 IBM 3480 Tape Cartridge Compatibility 1 of 1 EISNER::KILGALLEN "Larry Kilgallen" 5 lines 30-JUN-1989 20:23 -< "extra expense" is understating it! >- -------------------------------------------------------------------------------- > As I understand, it the DEC TA90 does this but requires an HSC, wis a > significant extra expense when you only have a single MicroVAX system on which > you need the tape drive. You have to buy a BigVAX to connect the MicroVAX to the HSC!  ================================================================================ Note 399.0 DS200/DL & CXA16 bugs 1 reply EISNER::MCCORMICK_B "Virtual Bob" 66 lines 3-JUL-1989 20:31 -------------------------------------------------------------------------------- I have searched through ALL TITLES and didn't see anything like this there, nor do I remember something like this being posted, so here goes... I've noticed some communications line problems in the last year or so, and have not been able to explain it. Two different configurations exhibit the problems: 1) MicroVAX 3XXX with CXA16 cards using DECconnect wiring system. Problem is noted by what appears to be frequent login attempts, which in itself causes system degradation. 2) DECserver 200/DL and DECserver 5X0 with CXA16 line cards. Degradation of the server occuring, with SHOW SERVER STAT showing upwards of 100% CPU utilization, and SHOW PORT X COU showing lotsa framing (and sometimes overrun) errors. I finally had some time to attempt debugging one of these problems. My setup was a DECconnect office com cabinet (H9646-EA) with two (new) DECserver 200/DL's, and 4 twisted pair wiring all to DECconnect spec (all parts from DEC). The H9646-EA was plugged into an isolated ground outlet on its own circuit. All wiring was checked out as OK ... no shorts, no excessive resistance (longest run under 200 feet), etc. All terminals (DEC VT320's) and printers (DEC LA210's) work fine, with no noted data errors. But: when the power to the terminals is shut off, the FRAMING ERRORS really start cranking! As noted above, I suspected a wiring problem (how can you screw up only four wires) but proved it was OK... Then I found that as soon as the terminal was turned on, the framing errors would stop (using MONITOR PORT X COU in the server). Even worse, I found that if you turned off the terminal, the errors would start, but then if you UNPLUGGED the terminal they would stop! Finally I got a very long extension cord (three pronger) and plugged the terminal into the power distribution outlets in the cabinet. No framing errors, ever! Here's a summary: TTY State Desk power SER power --------- --------- --------- Terminal ON No errors No errors Terminal OFF Errors No errors Terminal Unplugged No errors No errors All I can think of is there's something funny with ground inbalance between the power sources... but I didn't have the test equipment to go that far... I called DEC Atlanta, and they passed me on to Colorado Springs. I received a call back in no time at all... with some information: There is an ECO/FCO for CXA16's for this problem. It may have been released on 16-Jun-1989. The part number is EQ-01545-01. I was told to work with the local field service office regarding DECserver 200's ... I guess that means they know about the problem, but that there is no formal fix as of yet. In the mean time, you can ensure that your server or VAX is not clobbered by this "noise" by leaving your terminal on (per the recommendation by CSC when I spoke with them) or unplug it (at least by my experience). Has anyone else run into this situation, and can you share what you've found or done? ================================================================================ Note 399.1 DS200/DL & CXA16 bugs 1 of 1 EISNER::TANNENBAUM "Saul Tannenbaum" 7 lines 3-JUL-1989 21:46 -< I've seen framing error from unpowered terminals, too >- -------------------------------------------------------------------------------- >But: when the power to the terminals is shut off, the FRAMING ERRORS >really start cranking! I've seen the same sort of thing on DECServer 200's and the Ethernet terminal server. What I've done about is: nothing. At least now I know I wasn't imagining the problem...  ================================================================================ Note 400.0 What level of disk errors are normal? 9 replies EISNER::TANNENBAUM "Saul Tannenbaum" 35 lines 3-JUL-1989 22:03 -------------------------------------------------------------------------------- I have a relatively small disk subsystem: 1 RA82 4 RA81 2 RA60 connected to an HSC-50. There seem to be constant problems with the disks. The problems vary. Errors like SDI timeouts, Lost R/W ready's, Clock dropouts from online disks, etc. None of the problems seem media related. Often, we see disks go into "instantaneous" mount verification, with no errors indicated at all. Over the last 6-9 months there has been a consistent pattern. Errors occur. When it's clear that it wasn't a transient problem, we call field service. They tinker around, swap a board or two. Sometimes they fix the problem the first time, sometimes, it takes two tries. For a week or two, we have no errors. Then, the something else crops up. It seems to me that my whole disk subsystem has a problem, as all the drives have been involved at one time or another. DEC insists that that isn't the case, or more exactly, fixes each problem and says "There, that'll take care of it." I have a few questions for the assembled wisdom here: 1) Are these level of problems common? I'm talking in the range of 1-10 disk errors per day, on reasonably active disks. What do other people experience in terms of error rates? In other words, is this normal and the only problem here in my mind? 2) Assuming this is a real problem, how does one begin to address it? We've insisted on the obvious things - that cables be checked to see if they are loose, we've made sure boards are seated tightly, etc. We've been told that everything is up to rev. Any comments, questions, hints, etc. would be most appreciated. ================================================================================ Note 400.1 What level of disk errors are normal? 1 of 9 EISNER::KOZAM 10 lines 3-JUL-1989 22:46 -< 10/day seems excessive >- -------------------------------------------------------------------------------- It sounds like you're having an unusual number of problems. Two questions: 1. Do the problems affect all disks, or just certain ones? 2. Do the errors occur only a times of heavy use? If all disks are affected, I'd suspect an HSC problem, or possibly a cable problem (HSC to VAX). From the wide range of problems you've had, and their apparent resistance to simple board replacement cures, you might want to check out the HSC power supply. Power supply problems can often be "random", since ailing power supplies may not be able to handle small power fluctuations, while good supplies keep on ticking. ================================================================================ Note 400.2 What level of disk errors are normal? 2 of 9 EISNER::LEDERMAN "Bart Z. Lederman" 7 lines 4-JUL-1989 12:05 -< Also check line power. >- -------------------------------------------------------------------------------- In addition to checking the HSC power supply, I'd suggest you rent or buy a power line monitor. It's quite possible you are getting some kind of transient noise or dropouts which is causing problems with the boards. You want the kind of monitor which really monitors power, and can print out a record of what kind of faults (over or under voltage, missing cycles, spikes, off-frequency, or noise) and when it happens. ================================================================================ Note 400.3 What level of disk errors are normal? 3 of 9 EISNER::BRUCKER "Cyberspace Jockey" 14 lines 4-JUL-1989 17:24 -< more q's >- -------------------------------------------------------------------------------- what version of the hsc software are you running? We had timeout related problems back in v350 or maybe earlier. How many sdi cards (inc tape) do you have and how many power supplies? DEC recommends that you get a power supply upgrade if you have more than 3 sdi cards in the hsc50. they used to say 4, but have changed their song somewhere along the way. We also log these types of errors, but on the oder of 1-2 a week for the whole disk farm. We are running HSC50 with 3 disk and 1 tape channel. Connected drives are 6 ra81s, 1 ra82 and 2 ra90s with 3 tu78s. we see i/o rates of 10-20/sec/disk for at least 4 disks at any given time. -Ken ================================================================================ Note 400.4 What level of disk errors are normal? 4 of 9 EISNER::KENNEDY "Terry Kennedy" 8 lines 4-JUL-1989 17:27 -< FCO or bad unit select plug? >- -------------------------------------------------------------------------------- RA81's (at least) have several FCO's for grounding. The symptoms are what you describe (timeouts, soft ECC, dropping ready). As these do *NOT* involve boards, but are just little straps, the board changes would not have cured them. Also, I have found that the unit select plugs and sockets get dirty and the slightest vibration (seks, anyone?) will cause the drive to drop ready and go into mount verification. ================================================================================ Note 400.5 What level of disk errors are normal? 5 of 9 EISNER::TANNENBAUM "Saul Tannenbaum" 31 lines 4-JUL-1989 19:53 -< Some reponses >- -------------------------------------------------------------------------------- Thanks to everyone for their comments and questions. This all reinforces my basic assumption that there's something systemic wrong here. Some answers: All disk or certain ones? Usually, one or two at a time, but all our disks have been involved at one time or another. That it's never all disks at once has had me puzzled. It does suddenly occur to me that I should check whether they are all on the same disk channel... At times of heavy use? No. Sometimes they happen at 4 AM when there's nothing going on at all. The HSC power supply suggestion is a good one, as is the notion to check line power. We've had some problems that have seemed power related but when checked turned out to be something else. DEC had one of their Dranitz monitors on site for weeks a few years ago and found our power to be absolutely clean (it's a new building, designed for biomedical research), but things can change over the years. HSC Configuration: 2 disk channels, 1 tape channel, running HSC Version 370. We have 390, but I've resisted the temptation to install it, on theory that if this is ever to be diagnosed, we shouldn't change the environment unless we have to. I have no idea how many power supplies we have in the HSC, but my guess is 1. The RA81 ground strap sounds like something that was done, but I'll check. And I'll make sure the unit plugs are tight - we do swap them occassionally. ================================================================================ Note 400.6 What level of disk errors are normal? 6 of 9 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 5 lines 5-JUL-1989 09:14 -< One way to get rid of drive timouts... >- -------------------------------------------------------------------------------- FWIW, we used to have consistent drive timeout errors, always on our RA60 drives. Our FS person set the drive timeout parameter on the HSC to be 24. No more drive timeouts... Ed ================================================================================ Note 400.7 What level of disk errors are normal? 7 of 9 EISNER::LANG 5 lines 8-JUL-1989 00:41 -< Ditto .6 >- -------------------------------------------------------------------------------- We also "cured" our RA60 timeout problems by increasing the drive timeout parameter on the HSC to 24. This problem showed up for us when we upgraded to HSC version 370 about a year ago, and we have not had any RA60 timeouts since we changed the parameter. ================================================================================ Note 400.8 What level of disk errors are normal? 8 of 9 EISNER::TANNENBAUM "Saul Tannenbaum" 399 lines 18-JUL-1989 13:09 -< An update and some more information >- -------------------------------------------------------------------------------- An update and some more information.... DEC seems to be of two minds with respect to these disk problems. The disk specialists assert that there aren't any systemic problems here, while our Field Service Unit Manager seems less sanguine. I've spent some time with VAXSim and Spear producing a more comprehensive error summary [more comprehensive than DEC was able to do]. A 400 line report follows. Press Keypad whatever to skip it.... Once again, I appreciate any insight into these problems whatsoever.  File Environment SPEAR Version 3.1(2446) Input file: SYS$COMMON:[SYSERR]ERRCOMMON.SYS;1 Created: 17-JUL-1989 18:46:08.1 Output file: SYS$COMMON:[SYSERR]SUMMAR.RPT Selection Criteria: DISK, Date of first entry processed: 3-JAN 10:15:02 Date of last entry processed: 14-JUL 10:04:51 Number of entries processed: 150. Number of inconsistencies detected in error file: 0. Entry Occurance Counts: 150. Logged Message ...(100.)  *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* * * * DSA Disk Transfer Error Entries * * and SDI Error Entries * * Summarized by Device, then Error Type. * * * * Summarized by Event/Status Code. * * (Event/Status is in parentheses * * and is printed in hexadecimal.) * * * *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* Device: HSC000$DJA0 (RA60 S/N:D83) HSC000$DJA1 7. ( 6B) Position Errors (mis-seek) 1. ( EB) Drive detected error 1. ( 188) Five Symbol ECC Errors 5. ( 1AB) Receiver Ready Collision Errors Device: HSC000$DJA0 (RA60 S/N:5B) HSC000$DJA1 16. ( EB) Drive detected error 4. ( 1AB) Receiver Ready Collision Errors Device: HSC000$DUA2 (RA81 S/N:16611) 3. ( 2B) SDI Drive Command Timeout Errors 1. ( 8B) Lost Read/Write Ready During or Between Transfers Errors 1. ( EB) Drive detected error Device: HSC000$DUA2 (RA81 S/N:93E1) HSC000$DUA5 5. ( 2B) SDI Drive Command Timeout Errors 2. ( 6B) Position Errors (mis-seek) 19. ( 8B) Lost Read/Write Ready During or Between Transfers Errors 6. ( CB) Lost Receiver Ready for Transfer Errors 1. ( E8) Uncorrectable ECC Errors 59. ( EB) Drive detected error 3. ( 16B) Drive Failed Initialization Errors Device: HSC000$DUA4 (RA81 S/N:2A718) 2. ( 2B) SDI Drive Command Timeout Errors 2. ( 6B) Position Errors (mis-seek) 3. ( 10B) Controller Detected Pulse or State Parity Errors 4. ( 16B) Drive Failed Initialization Errors Device: HSC000$DUA2 (RA81 S/N:0) HSC000$DUA6 2. ( EB) Drive detected error Device: HSC000$DUA6 (RA81 S/N:2595) 1. ( 2B) SDI Drive Command Timeout Errors  *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* * * * The following Summarizes all DSA Disk Drive LED Code * * and Front Panel errors. These codes are a result of * * Drive Detected Errors (Event/Status = 'EB') during * * an SDI error (Format = '3'). * * * * LED Code and Front Panel codes are in hexadecimal. * * Error Count is decimal. * * * *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* Device: HSC000$DJA0 (RA60 S/N:D83) HSC000$DJA1 LED Front Count Code Panel LED Code Translation ----- ---- ----- -------------------- 1. 2D 00 Bad samples during linear mode Device: HSC000$DJA0 (RA60 S/N:5B) HSC000$DJA1 LED Front Count Code Panel LED Code Translation ----- ---- ----- -------------------- 4. 13 00 Lid unlocked when run command is issued 12. 30 00 Spin-up required too short a time Device: HSC000$DUA2 (RA81 S/N:16611) LED Front Count Code Panel LED Code Translation ----- ---- ----- -------------------- 1. 39 1A Write and Off Track Device: HSC000$DUA2 (RA81 S/N:93E1) HSC000$DUA5 LED Front Count Code Panel LED Code Translation ----- ---- ----- -------------------- 2. 00 00 Undefined LED code (HARDCORE) 8. F1 13 Slave load timeout 3. F8 13 Slave seek timeout 30. 39 1A Write and Off Track 15. F1 1A Slave load timeout 1. F1 1C Slave load timeout Device: HSC000$DUA2 (RA81 S/N:0) HSC000$DUA6 LED Front Count Code Panel LED Code Translation ----- ---- ----- -------------------- 1. 07 1C Level 2 message frame sequencing error 1. F1 1C Slave load timeout  *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* * * * This report Summarizes all DSA Disk Drive Read Data Errors * * by Drive and Volume Serial Number. All values are decimal. * * * * If 'REP' is printed under BBR status, this indicates that * * the block was replaced. An asterisk ('*') next to this * * indicates that the block was replace with the 'Force Error' * * modifier. If 'OK' is printed, then Bad Block Replacement * * was requested, but the block was tested to be OK. * * If there is nothing printed, then retry was successful and * * no replacement occured. * * * *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* Device: HSC000$DJA0 (RA60 S/N:D83) HSC000$DJA1 Sector Block Phys. From Error BBR Volume S/N LBN Type Cyl. Head Sect Index Count Status Error Type ---------- ------- ---- ---- ---- ---- ----- ----- ------ --------------------- 1615711845 346005 LBN 1374 1 9 41 6 Position (mis-seek) 1615711845 49902 LBN 196 3 6 6 1 Position (mis-seek) 1615711845 296960 LBN 1178 3 20 9 1 OK Five Symbol ECC Device: HSC000$DUA2 (RA81 S/N:16611) Sector Block Phys. From Error BBR Volume S/N LBN Type Cyl. Head Sect Index Count Status Error Type ---------- ------- ---- ---- ---- ---- ----- ----- ------ --------------------- 73993 547698 LBN 767 1 9 23 1 Lost R/W Ready Device: HSC000$DUA2 (RA81 S/N:93E1) HSC000$DUA5 Sector Block Phys. From Error BBR Volume S/N LBN Type Cyl. Head Sect Index Count Status Error Type ---------- ------- ---- ---- ---- ---- ----- ----- ------ --------------------- 2591 729285 LBN 1021 5 36 2 1 Position (mis-seek) 73993 869657 LBN 1218 0 5 5 1 Lost R/W Ready 73993 870366 LBN 1219 0 0 0 1 Lost R/W Ready 73993 870374 LBN 1219 0 8 8 1 Lost R/W Ready 73993 871080 LBN 1220 0 0 0 1 Lost R/W Ready 73993 871086 LBN 1220 0 6 6 1 Lost R/W Ready 73993 88626 LBN 124 1 39 1 1 Lost R/W Ready 73993 90027 LBN 126 1 12 26 5 Lost R/W Ready 73993 90028 LBN 126 1 13 27 5 Lost R/W Ready 73993 90032 LBN 126 1 17 31 1 Lost R/W Ready 73993 90039 LBN 126 1 24 38 2 Lost R/W Ready 73993 80352 LBN 112 7 27 21 1 REP Uncorrectable ECC 184923 33157 LBN 46 6 7 39 1 Position (mis-seek) Device: HSC000$DUA4 (RA81 S/N:2A718) Sector Block Phys. From Error BBR Volume S/N LBN Type Cyl. Head Sect Index Count Status Error Type ---------- ------- ---- ---- ---- ---- ----- ----- ------ --------------------- 133951 720536 LBN 1009 2 8 36 1 Position (mis-seek) 133951 792780 LBN 1110 4 36 40 1 Position (mis-seek)  *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* * * * The following Summarizes RA Series Drive Media related * * errors by head. * * * *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* Device: HSC000$DJA0 (RA60 S/N:D83) HSC000$DJA1 Read Error Count: 8. Media: Volume S/N (decimal): 1615711845 Head 0 1 2 3 4 5 ----------------------------- Count 6. 2. Device: HSC000$DUA2 (RA81 S/N:16611) Read Error Count: 1. Media: Volume S/N (decimal): 73993 Head 0 1 2 3 4 5 6 7 8 9 10 11 12 13 --------------------------------------------------------------------- Count 1. Device: HSC000$DUA2 (RA81 S/N:93E1) HSC000$DUA5 Read Error Count: 1. Media: Volume S/N (decimal): 2591 Head 0 1 2 3 4 5 6 7 8 9 10 11 12 13 --------------------------------------------------------------------- Count 1. Device: HSC000$DUA2 (RA81 S/N:93E1) HSC000$DUA5 Read Error Count: 20. Media: Volume S/N (decimal): 73993 Head 0 1 2 3 4 5 6 7 8 9 10 11 12 13 --------------------------------------------------------------------- Count 5. 14. 1. Device: HSC000$DUA2 (RA81 S/N:93E1) HSC000$DUA5 Read Error Count: 1. Media: Volume S/N (decimal): 184923 Head 0 1 2 3 4 5 6 7 8 9 10 11 12 13 --------------------------------------------------------------------- Count 1. Device: HSC000$DUA4 (RA81 S/N:2A718) Read Error Count: 2. Media: Volume S/N (decimal): 133951 Head 0 1 2 3 4 5 6 7 8 9 10 11 12 13 --------------------------------------------------------------------- t 1. 1.  Overall Error Distribution Main-|Disk|Tape|Unit|Comm|Net-|Soft| CI | UDA|Totals frame| | |rec | |work|ware| | -----+----+----+----+----+----+----+----+----+------- Totals |150.| | | | | | | | 150.  [the following is edited for brevity...] Error distribution 03-Jan-89 Totals | 1.| 05-Jan-89 Totals | 3.| 06-Jan-89 Totals | 1.| 08-Jan-89 Totals | 1.| 11-Jan-89 Totals | 1.| 17-Jan-89 Totals | 4.| 20-Jan-89 Totals | 8.| 25-Jan-89 Totals | 1.| 26-Jan-89 Totals | 1.| 27-Jan-89 Totals | 1.| 02-Feb-89 Totals | 1.| 03-Feb-89 Totals | 6.| 05-Mar-89 Totals | 7.| 15-Mar-89 Totals | 3.| 21-Mar-89 Totals | 1.| 06-Apr-89 Totals | 1.| 25-Apr-89 Totals | 2.| 27-Apr-89 Totals | 2.| 03-May-89 Totals | 1.| 04-May-89 Totals | 1.| 08-May-89 Totals | 2.| 11-May-89 Totals | 2.| 12-May-89 Totals | 3.| 17-May-89 Totals | 2.| 11-Jun-89 Totals | 1.| 12-Jun-89 Totals | 2.| 13-Jun-89 Totals | 2.| 14-Jun-89 Totals | 6.| 15-Jun-89 Totals | 2.| 17-Jun-89 Totals | 2.| 18-Jun-89 Totals | 24.| 19-Jun-89 Totals | 4.| 01-Jul-89 Totals | 1.| 02-Jul-89 Totals | 5.| 08-Jul-89 Totals | 4.| 10-Jul-89 Totals | 34.| 11-Jul-89 Totals | 1.| ================================================================================ Note 400.9 What level of disk errors are normal? 9 of 9 EISNER::KENNEDY "Terry Kennedy" 22 lines 19-JUL-1989 00:30 -< 6 items to do >- -------------------------------------------------------------------------------- > Once again, I appreciate any insight into these problems whatsoever. Some thoughts: 1) RA60/5B - Test and/or replace cover latch/pack-in-place switches. 2) RA81/93E1 - Replace front panel *and* unit select plug. Also replace the main processor card - it seems to be falling asleep more than it should. 3) Others seeom ok. 4) Check for damaged/bent SDI cable pins, especially where the cables attach to the drive logic and where they pass through the static fil- ters in the bottom of the drive cabinet(s) and HSC cabinet(s). 5) Verify sufficient power supply capacity in HSC. Scope voltages in HSC at channel card backplane position. Look for dips (un-spikes?) in all voltages. 6) Last-chance option - install power line monitor (available thru FS support) on AC power line. RUn for at least 1 week.  ================================================================================ Note 401.0 Optical WORM; Perceptics Laserware; Reactions? 1 reply EISNER::TIHOR "Stephen Tihor;SITE SIG,VAX SIG" 15 lines 5-JUL-1989 19:47 -------------------------------------------------------------------------------- It looks like we will be buying a WORM optical disk system in the near future. We want a disk emulation mode, clusterable, ideally HSC hosted and DEC supported. Our operations manager has proposed a system using the Perceptics driver. According to thier manual this is not cluster sharable. The DEC device does not offer multi-user disk style access, just "magnetic tapes" with improved seek times. (a) is anyone using LaserWare or any of its parts and if so what were your experiences (or reactions if you chose not to use it?) (b) what other optical solutions have people found usable? [FYI we are aleary getting a DEC CD-ROM system] ================================================================================ Note 401.1 Optical WORM; Perceptics Laserware; Reactions? 1 of 1 EISNER::LANG 27 lines 8-JUL-1989 00:52 -< We Use Perceptics >- -------------------------------------------------------------------------------- We have been using Laserware atour site for over three years. The software is reliable and relatively bug free. Once I finally got my users using the disks, they have been in constant use. Our hardware consists of a TD systems UHA-0 scsi controller (on a DWbua) in a VAX 8350. The drives are Optimem 1000's. We have three drives currently. As for access across a network (as opposed to hanging off an HSC,) we are using DFS and it seems to work quite well. The only major problems that we have seen seem to be due mostly to media defects. For instance, recently, the night operator at our site had trouble writing a large file to one optical disk cartidge. We pulled the disk itself from the cartridge, and found some particles that looked like tiny pieces of packing foam! We cleaned them off, and the system started working better. All in all, we have been very pleased with Perceptics. The support people are VERY helpful when we think we might have a problem. We have solved a number of problems with them including hardware and software. If anyone else is interested in this topic, I could post a more detailed note of how our site uses the opticals. rml  ================================================================================ Note 402.0 Versatec V80 DTR problem 2 replies EISNER::NIXON 6 lines 6-JUL-1989 10:20 -------------------------------------------------------------------------------- We have tried using all the possible setups, on decserver 200 ports, and dhv11's and as a slave on a vt340. They all produce the same results. Do you have to have a special cable to enable DTR ? And do you know how the board on the V80 is set up ? ================================================================================ Note 402.1 Versatec V80 DTR problem 1 of 2 EISNER::NIXON 12 lines 6-JUL-1989 12:38 -< no more shouting... >- -------------------------------------------------------------------------------- < Note 402.0 by EISNER::NIXON > -< Versatec V80 DTR problem >- We have tried using all the possible setups, on decserver 200 ports, and dhv11's and as a slave on a vt340. They all produce the same results. Do you have to have a special cable to enable DTR ? And do you know how the board on the V80 is set up ? sorry about the shouting first time user.... ================================================================================ Note 402.2 Versatec V80 DTR problem 2 of 2 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 1 line 6-JUL-1989 16:17 -< This belongs under 397.* >- --------------------------------------------------------------------------------  ================================================================================ Note 403.0 Dual-ported DSSI (RF30, RF71) disks 17 replies EISNER::DRIESMAN "Debbie Driesman, CSC" 7 lines 6-JUL-1989 12:18 -------------------------------------------------------------------------------- How do dual-ported DSSI disks work? Can both processors read and write the disk at the same time? If not, how is access transferred between the two processors? Can (should) the system disk be dual-ported? We are planning a production system where one processor would be an on-line system and the other a backup system, able to take over if the on-line system failed. The dual-ported disk would contain the application data files. ================================================================================ Note 403.1 Dual-ported DSSI (RF30, RF71) disks 1 of 17 EISNER::KILGALLEN "Larry Kilgallen" 16 lines 6-JUL-1989 14:43 -< Separate access paths to the disk >- -------------------------------------------------------------------------------- > How do dual-ported DSSI disks work? Can both processors read and write > the disk at the same time? If not, how is access transferred between the > two processors? Can (should) the system disk be dual-ported? Both processors communicate directly with the disk, although there is only one head actuator and data path, so they don't read and write at *exactly* the same time, but about 3 orders of magnitude closer than SDI disks. DEC recommends dual-porting a shared system disk if you envision powering down one processor while the other keeps running. Separate system disks would also allow this. Note in order to dual-port a shared system disk, at least one of the processors must *not* be a MicroVAX-II due to the boot ROM only being able to go to root 0. (Also, of course 3300/3400 processors cannot be mixed with other types of processors in a dual-port situation). ================================================================================ Note 403.2 Dual-ported DSSI (RF30, RF71) disks 2 of 17 EISNER::KILGALLEN "Larry Kilgallen" 2 lines 6-JUL-1989 14:48 -< Powering down half with a shared system disk >- -------------------------------------------------------------------------------- The recommendation presumes that the shared disk is in a cabinet separate from either processor. ================================================================================ Note 403.3 Dual-ported DSSI (RF30, RF71) disks 3 of 17 EISNER::TIHOR "Stephen Tihor;SITE SIG,VAX SIG" 6 lines 6-JUL-1989 20:50 -< CI:DSSI cable::HSC+Disk:DSSI disk >- -------------------------------------------------------------------------------- When trying to think about DSSI disks you get good intuitions if you think of DSSI as an inexpensive CI and every DSSI disk a microHSC with one controller and one drive. This would imply that DSSI disks can't do controller based shadowing (only host based). I believe this is correct has anyone tried it? ================================================================================ Note 403.4 Dual-ported DSSI (RF30, RF71) disks 4 of 17 EISNER::MCCORMICK_B "Virtual Bob" 5 lines 6-JUL-1989 20:56 -< DSSI shadow maybe by 3rd party >- -------------------------------------------------------------------------------- From a friend, I believe you can shadow on DSSI today with a third party product from a company called Bear (I think in Holywood CA). I no nothing further (and hence am not affiliated with or endorsing such a product) ... ================================================================================ Note 403.5 Dual-ported DSSI (RF30, RF71) disks 5 of 17 EISNER::TIHOR "Stephen Tihor;SITE SIG,VAX SIG" 7 lines 6-JUL-1989 21:33 -< ..maybe..do you remember if it does shared disks? >- -------------------------------------------------------------------------------- Yes, the bear product would work since its a HOST based shadower. But you would have to MSCP serve the shadowed disk to the other node or use two virtual disks and disk switch between a straight virtual disk and a network virtual disk. I better pull out my tape of the bear talk on their virtual disk product sheaf to find out if they have a deamon to do the failover. ================================================================================ Note 403.6 Dual-ported DSSI (RF30, RF71) disks 6 of 17 EISNER::BATES "Ken Bates" 12 lines 11-JUL-1989 20:23 -------------------------------------------------------------------------------- re: .1 I don't understand the "3 orders of magnitude" difference between SDI and DSSI that you refer to. As far as I know, there is no difference, other than that caused by the disk communication mechanism, and that is miniscule (I wrote the code for both the HSC and the DSSI controllers, so I may be prejudiced). re: .3: The current DSSI products do not contain shadowing code ala HSC style shadowing (or ala anything, for that matter). - Ken ================================================================================ Note 403.7 Dual-ported DSSI (RF30, RF71) disks 7 of 17 EISNER::KILGALLEN "Larry Kilgallen" 17 lines 15-JUL-1989 21:49 -< 3 orders of magnitude switch time >- -------------------------------------------------------------------------------- > Note 403.6 > EISNER::BATES "Ken Bates" > > I don't understand the "3 orders of magnitude" difference between SDI and DSSI > that you refer to. As far as I know, there is no difference, other than that > caused by the disk communication mechanism, and that is miniscule (I wrote the > code for both the HSC and the DSSI controllers, so I may be prejudiced). The topic was whether two processors can write to the disk at "the same time". With KDA50s driving SDI disks, the dual porting is "static", with a switching time of about one second (and the way VMS uses that is only in the case of failures). With DSSI disks, the dual porting is "dynamic", as with Massbus disks. Thus reads from two different processors can be interleaved, so I estimate the "switch" time as 1 millisecond (wild guess). Thus, the difference in the degree to which separate processors approximate writing to the disk "at the same time" is 3 orders of magnitude. ================================================================================ Note 403.8 Dual-ported DSSI (RF30, RF71) disks 8 of 17 EISNER::BATES "Ken Bates" 19 lines 16-JUL-1989 21:18 -------------------------------------------------------------------------------- I think I must still be missing something. The difference is that the KDA (not SDI) is a single host processor. Only one processor at a time can have a specific disk drive online. If a host on port A sends an SDI disconnect command and the host on port B sents an ONLINE command, SDI protocol dictates that the ONLINE must take a certain period of time. This time delay is long enough to make dynamic sitching impractical. DSSI, on the other hand, does not really have the concept of dual porting. Instead, it (the ISE) supports multiple hosts. As a result, both hosts can initiate an SCA connection to the disk simultaneously. Because of this, both hosts can have the disk online, and both hosts can issue I/O commands to the disk in parallel. Very similar to the HSC in this respect. I think that instead of "dynamic switching", a more proper term would be "dual host access" (or some other such thing). Disclaimer: This is from the standpoint of the I/O subsystem; I'm not sure how the VMS driver actually implements all the potential functionality. - Ken ================================================================================ Note 403.9 Dual-ported DSSI (RF30, RF71) disks 9 of 17 EISNER::KILGALLEN "Larry Kilgallen" 17 lines 16-JUL-1989 22:46 -------------------------------------------------------------------------------- > DSSI, on the other hand, does not really have the concept of dual porting. Both KFQSA's can be online to the disk at the same time, and send their requests directly, rather than over the Ethernet to the other node for processing. > Instead, it (the ISE) supports multiple hosts. As a result, both hosts can > initiate an SCA connection to the disk simultaneously. Because of this, both > hosts can have the disk online, and both hosts can issue I/O commands to the > disk in parallel. Very similar to the HSC in this respect. I think that There is a certain similarity to the HSC, but that is a device not available for the class of machines under discussion. (The HSC also introduces a single point of performance contention for multiple spindles). For KDA50 controllers (the only MicroVAX controller before the KFQSA with any dual porting) the switch-over time is analogous to the switch-over of a single SDI disk between two HSCs. ================================================================================ Note 403.10 Dual-ported DSSI (RF30, RF71) disks 10 of 17 EISNER::BATES "Ken Bates" 37 lines 21-JUL-1989 20:36 -------------------------------------------------------------------------------- re: .-1 >> DSSI, on the other hand, does not really have the concept of dual porting. > > Both KFQSA's can be online to the disk at the same time, and send > their requests directly, rather than over the Ethernet to the other > node for processing. Sorry, this is most emphatically *not* dual porting. It's dual connections via a single port to different hosts. The controller in the host, be it a KFQSA or EDA640 is viewed by the ISE as a host. Both go through one port on the disk. It's similar to n hosts accessing a single disk on an HSC. Still only on one port. Since the ISE's do not connect to an external controller, there's no need for dual porting. > There is a certain similarity to the HSC, but that is a device not > available for the class of machines under discussion. (The HSC > also introduces a single point of performance contention for multiple > spindles). For KDA50 controllers (the only MicroVAX controller > before the KFQSA with any dual porting) the switch-over time is > analogous to the switch-over of a single SDI disk between two HSCs. Once again, controllers are not dual ported, disks are. The KDA50 is not a dual ported controller. A disk which is dual ported may be connected to both an HSC and a KDA (or KDB or UDA), and only one of the controller may access it at a time. Switchover time (offline on one controller and online to another) is rigidly specified by DSA. The value is such that dynamic dual porting is not feasible. As for HSC performance contention with multiple disks, I would be very interested in the configuration that caused it. To date, we have seen only one or two cases where an HSC was saturated or was a bottleneck. In general, it may appear to be the problem, but analysis usually finds that the problem lies elsewhere. If you have a case, please let me know what you found so that we can look into it. - Ken ================================================================================ Note 403.11 Dual-ported DSSI (RF30, RF71) disks 11 of 17 EISNER::KILGALLEN "Larry Kilgallen" 20 lines 21-JUL-1989 22:49 -< We obviously differ on the use of language. >- -------------------------------------------------------------------------------- >Sorry, this is most emphatically *not* dual porting. It's dual connections via >a single port to different hosts. The controller in the host, be it a KFQSA or >EDA640 is viewed by the ISE as a host. Both go through one port on the disk. Either VAX can access the disk without having its data go through the other VAX. The disk talks to either VAX directly. Whether the electronics built into the disk is a small number of parts to talk MASSbus, or a larger number of parts to talk DSSI (and named ISE to indicate the presence of firmware), either VAX can talk directly to the disk (via associated control circuitry regardless of its name or complexity) without involving the other VAX. That, to me, is dual porting. The exact details of how many cables go into the disk and whether they are jumpered inside is of little consequence. The point is that you don't pay a performance penalty of going through the other VAX. And the point of the original response was that the two VAXen do not write to the disk "simultaneously", but it is about three orders of magnitude more simultaneous that the corresponding KDA50 hardware connections (which is why one routes disk requests through the other VAX). ================================================================================ Note 403.12 Dual-ported DSSI (RF30, RF71) disks 12 of 17 EISNER::BATES "Ken Bates" 30 lines 23-JUL-1989 10:41 -------------------------------------------------------------------------------- Having participated in the development of DSA, my use of "ports" and "dual porting" was based on DSA. This, I suppose, was the difference. The differences lie in the use of the terms "dual ports", "multiple hosts", and "multiple connections". RA disks are "dual ported". DSSI disks are "single ported". The KDA, KDB, and UDA are "single host, single connection", the HSC is "multiple host, multiple connection", and the DSSI ISE's are "multiple host, multiple connection". The KFQSA and EDA640 adapters are analagous to the CI780. The confusion appears to be that the controller and disk are separate and distinct on RA disks, while they are integrated in DSSI (hence the term "Integrated Storage Element"). As for "three orders of magnitude" difference between a KDA and DSSI, you should have field service come in and check things out, since you may have a hardware problem. The additional time for a QIO operation on an MSCP served disk is far less than 10 ms for a 780 class machine, and proportionally shorter for more powerful machines. (If I recall correctly, it's only abouyt 5 ms, but I'm not positive about this). Since a typical data access time for a disk on VMS is about 25 ms, this would increase to about 35 ms (maximum) on a 780. DSSI is also about 25 ms, so the difference would be about 40%. If you have a 75 ms access time via the KDA (three orders of magnitude increase) with the same type of access pattern, then something is wrong. The only thing I could postulate could be the cause of this is if you were transferring data with byte counts in the hundreds of KB per QIO, which will slow things down through the MSCP server. - Ken ================================================================================ Note 403.13 Dual-ported DSSI (RF30, RF71) disks 13 of 17 EISNER::BATES "Ken Bates" 57 lines 23-JUL-1989 16:28 -< Oops... >- -------------------------------------------------------------------------------- re: .-1 Well, that's what happens when I try to answer a note first thing in the morning without having had enough coffee. Disregarding my failure to multiply properly to get three orders of magnitude and the discussion about DSA terminology, I thought about it a bit and finally realized the point of the discussion. Probably my large systems background is what took me so long to finally get the point. Based on the subsystem design, internals, and measurements: If you access a disk from one host, there is no appreciable difference between an RA disk on a KDA and a DSSI disk, given similar disks. If you really measure carefully, then a KDA will be faster than DSSI, but you won't see this difference in normal circumstances. If you want to access a disk from two hosts simultaneously, there is no practical difference between a KDA and a DSSI disk, since the MSCP server time on the VAX is negligible. If you're running benchmarks and using measuring equipment, or accessing the disk with extremely high byte counts, then you will be able to see a small difference (~10 ms per access, plus data transfer time). This difference is due to the host which has the KDA having to MSCP serve it for other hosts, causing not only a small delay, but also stealing valuable CPU cycles. If you want to access a disk from one host exclusively and then switch over to exclusive access from the other host, then there will be no difference between DSSI and DSA, since the MSCP protocol dictates that there will be a slight time delay between the hosts having access, due to the MSCP requirements. This delay is noticible, and effectively prevents dynamic switching for real time environments. If you want both hosts to be able to access the disk with no delay whatsoever, then DSSI is the only way. Since both hosts can have the disk online at the same time without inserting the additional layer of the MSCP server, DSSI will outperform the KDA. In effect, this could be considered analagous to a CI cluster, with the HSC and disk together acting like the RF disk, and the CI adapter acting like the DSSI port on the host. Dual porting is only effective if the disk is connected to two controllers (such as an RA disk connected between two HSCs). The disk can only be online to one port (controller) at a time, and if one controller fails, it will fail over to the other. Since in DSSI the disk and controller are integrated, then failure of the controller will also render the disk unreachable. Failure of the host in a two host system will cause the ISE to disconnect from the failing host and remain online to the other host. Sorry for prolonging this discussion unnecessarily; I forgot that DSSI disks connected to uVAXes, and things such as a KDA would be inappropriate for high performance use in certain areas. Larry is correct, if you want the absolute top performance on a dual-host, non-CI system, then DSSI will offer superior performance. - Ken ================================================================================ Note 403.14 Dual-ported DSSI (RF30, RF71) disks 14 of 17 EISNER::KILGALLEN "Larry Kilgallen" 30 lines 24-JUL-1989 13:57 -< Ken, you and I may get a DECUServe disk space charge >- -------------------------------------------------------------------------------- > Dual porting is only effective if the disk is connected to two controllers > (such as an RA disk connected between two HSCs). The disk can only be online > to one port (controller) at a time, and if one controller fails, it will fail > over to the other. Back when SDI was announced, this was described as "static dual porting", and the normal (Massbus) dual porting was described as "dynamic dual porting" (whereas formerly it had been described as simply "dual porting"). My point in discussion with Ken is that from a VMS perspective the behaviour if DSSI is much more similar to a Massbus than to a KDA. That is why I look at the DSSI as being "dual-ported", even though the wires from both CPUs to the RF71 may be connected to each other inside the RF71. > Larry is correct, if you want the absolute top performance on a dual-host, > non-CI system, then DSSI will offer superior performance. Well thanks, but that isn't what I meant to say, even though it is what I believed. My original intent was to answer the question as to whether the hosts write to the disk "simultaneously". My answer was "no, there is only one actuator", but also I wanted to indicate that there was a significant difference in the nature of the "direct" connection, such that if you outlaw going through another VAX and were instead to "switch" the hardware, it would take about one second to go from one KDA to another. (Not that VMS does this, but this is the hardware conference ;-) ). Well, I am sure the two of us have succeeded in boring the rest of you to tears, including Debbie (the author of .0) who has probably gone off to buy an HP 3000 by now ;-). ================================================================================ Note 403.15 Dual-ported DSSI (RF30, RF71) disks 15 of 17 EISNER::HARVEY "Jack Harvey" 6 lines 24-JUL-1989 14:57 -------------------------------------------------------------------------------- > -< Ken, you and I may get a DECUServe disk space charge >- Over my dead body. ================================================================================ Note 403.16 Dual-ported DSSI (RF30, RF71) disks 16 of 17 EISNER::MAYHEW "Bill Mayhew" 6 lines 24-JUL-1989 16:41 -< Silence := approval! >- -------------------------------------------------------------------------------- > Well, I am sure the two of us have succeeded in boring the rest > of you to tears... Not me. I'm following this with considerable interest even though I don't have anything to add to it. ================================================================================ Note 403.17 Dual-ported DSSI (RF30, RF71) disks 17 of 17 EISNER::DRIESMAN "Debbie Driesman, CSC" 15 lines 25-JUL-1989 10:00 -< going with microVAX 3300/3400s >- -------------------------------------------------------------------------------- > Well, I am sure the two of us have succeeded in boring the rest > of you to tears, including Debbie (the author of .0) who has probably > > gone off to buy an HP 3000 by now ;-). No way! We are going to use microVAX 3300/3400 systems. The development systems will be on 3400s (we need the extra slots for another project will will also be supporting) and the production systems will be 3300s. Since I posted the original note, people have being trying to decide whether they really need shared disk access, and if they want to cluster the machines or just network them. Since I've never worked with a cluster, I've been doing a lot of research the last few weeks (a lot of it right here on DECUServe). Thanks for all of the information. And I certainly haven't been bored with the discussion.  ================================================================================ Note 404.0 Need Postscript Laser Printer Suggestions 4 replies EISNER::ROBERTS_D "don roberts #241" 3 lines 7-JUL-1989 11:07 -------------------------------------------------------------------------------- We need a postscript printer for our vaxen. Any comments or experience with the current offerings on the market? Has this been discussed somewhere? ================================================================================ Note 404.1 Need Postscript Laser Printer Suggestions 1 of 4 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 10 lines 7-JUL-1989 12:27 -< Try DEC LN03r? >- -------------------------------------------------------------------------------- We have a DEC LN03r ("ScriptPrinter") for ANSI, TEK 4014, REGIS, and PostScript output. See topics : DESKTOP PUBLISHING 27 HARDWARE_HELP 64 LATEST_RELEASE_INFORMATION 81 VMS 532 We're still waiting at this point for the version 2 software, so I can't give you good feedback on how well that works. ================================================================================ Note 404.2 Need Postscript Laser Printer Suggestions 2 of 4 EISNER::OSUDAR "John Osudar" 19 lines 7-JUL-1989 17:17 -< LN03R, T1590+PSI, Dataproducts >- -------------------------------------------------------------------------------- *** DISCLAIMER *** The opinions expressed in this message are my own. I do not speak for my employers, nor for any other person or organization. Nor do I have any vested interest in the products mentioned. ================== We're using DEC LN03R's ("ScriptPrinter") and are reasonably happy with them. We also have a Talaris T1590 with their PSI option (a PostScript emulation that is *not* from Adobe) and have had some problems with it -- most notably, lack of speed (it's a 15 page/min. printer but comes nowhere near that doing PostScript), faintness of PostScript fonts (especially Courier), and occasional "glitches" (e.g. mis-positioned characters, errors reported that aren't reported by the LN03R's.) While I wouldn't buy a T1590 for its PostScript support, it was reasonable to add the support onto a T1590 we already had. The third PostScript printer we've had is a Dataproducts 12 page/min. printer that we're currently evaluating. It uses Adobe's PostScript and seems to work reasonably well -- although I've only been testing it for a few days so far, and without a major print load. ================================================================================ Note 404.3 Need Postscript Laser Printer Suggestions 3 of 4 EISNER::MERRIMAN "George Merriman - CCA/NY" 2 lines 7-JUL-1989 22:41 -< other topics >- -------------------------------------------------------------------------------- See Topic 216 in this conference for more information. I think there are some topics in DESKTOP_PUBLISHING. ================================================================================ Note 404.4 Need Postscript Laser Printer Suggestions 4 of 4 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 23 lines 10-JUL-1989 12:21 -< Camintonn/AST TurboLaser-PS/Plus-3 >- -------------------------------------------------------------------------------- >> -< Need Postscript Laser Printer Suggestions >- I have an Camintonn/AST TurboLaser-PS/Plus-3 and like it very much. It uses the same 8-ppm Ricoh engine that DEC uses for the LN03 series and it prints very good large black areas. The "PS/Plus-3" in the name means that it provides Adobe Postscript plus three other printer emulations: o DEC LN03 o Diablo 630 ECS (Extended Character Set) o HP Laserjet I especially like it because I can switch back and forth between LN03 and Postscript with just an escape sequence. No need for DEC's software! You can even mix the two on the same page. I have found and reported a few minor errors in the LN03 emulation and, if their past performance is any guide, expect them to be corrected rapidly. I haven't used the Diablo or HP emulations yet, so I can't speak for them.  ================================================================================ Note 405.0 Sharing a Xerox 9700 Printer 7 replies EISNER::SHANNON "Terry C. Shannon, erstwhile Matco" 9 lines 7-JUL-1989 16:58 -------------------------------------------------------------------------------- One of my clients has a DEC-IBM shop which is graced with four Xerox 9700 printers. Two of these expensive beasts are connected to a VAX host, the other two are connected to an IBM mainframe. The client would like to decommission one printer, dedicate one printer to each host and install a switch that would allow the remaining printer to be shared by the DEC and IBM systems. Is this a feasible, practical scheme? If so, where could one find a switch that would make printer sharing possible? ================================================================================ Note 405.1 Sharing a Xerox 9700 Printer 1 of 7 EISNER::BURNET "John Burnet" 13 lines 7-JUL-1989 21:45 -< Try Black Box for device-sharing switches >- -------------------------------------------------------------------------------- > Is this a feasible, practical scheme? If so, where could one find a switch > that would make printer sharing possible? Have you looked at Black Box? They have a good selection of both manual and computer-actuated switches in both serial (RS-232) and parallel (Centronics, Dataproducts) flavors. By "switch" I mean not only the usual kind, but also port-sharing devices that go by other names. I don't have BB's address or number handy, but if you can't find the number somewhere in HARDWARE_HELP, you can get it from Pittsburgh directory assistance. How are the printers currently attached to the hosts? ================================================================================ Note 405.2 Sharing a Xerox 9700 Printer 2 of 7 EISNER::KENNEDY "Terry Kennedy" 14 lines 8-JUL-1989 02:51 -< Xerox laser saga >- -------------------------------------------------------------------------------- > Is this a feasible, practical scheme? If so, where could one find a switch > that would make printer sharing possible? Is the 9700 the IBM channel-attached one with the built-in 9-track tape drive? In a former job I was charged with designing an interface between a Xerox laser and a Data General system. Xerox kept saying "you *want* to make tapes for it" and my company kept saying "no, we *require* a direct attachment". I wound up implementing the whole S/370 channel interface for a DG to drive the beast. Of course, the customer then decided they didn't want the printer... If this is the case and Xerox doesn't have some sort of DEC interface, then I'd agree: you *want* to make tapes for it. ================================================================================ Note 405.3 Sharing a Xerox 9700 Printer 3 of 7 EISNER::GLEASON "CyberPunk" 11 lines 9-JUL-1989 03:11 -< Tape Ape >- -------------------------------------------------------------------------------- > If this is the case and Xerox doesn't have some sort of DEC interface, >then I'd agree: you *want* to make tapes for it. A while back, we looked at the available interfaces for our VAX and 9700 - concluded that it was cheaper to pay an operator to move tapes across the room. We wrote a PSM symbiont that takes care of writing the tapes, and the users just think it's another printer. Haven't priced interfaces in a couple of years, though. ================================================================================ Note 405.4 Sharing a Xerox 9700 Printer 4 of 7 EISNER::HARVEY "Jack Harvey" 16 lines 9-JUL-1989 11:36 -< Put All Lasers on One System >- -------------------------------------------------------------------------------- A few years ago, I was helping a DEC/IBM shop that had two 5000 lpm laser printers on the IBM system. The VAX had a 300 lpm Dataproducts. Some genius had the idea of establishing a link between the VAX and the 3090, so VAX users could get to the lasers. I did it with DEC's 3780 emulator package and wrote a special print symbiont to drive it. Worked great. The VAX had a print queue for each of the printers on the IBM system. We extended the scheme to include the 2200 lpm impact printers as well, and decommissioned the old Dataproducts, which was getting very weary. VAX users went wild with printouts. Then a bean counter discovered that the growing monthly fee for the lasers was based on page count... ================================================================================ Note 405.5 Sharing a Xerox 9700 Printer 5 of 7 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 7 lines 11-JUL-1989 09:55 -< Tape, simple and cheap >- -------------------------------------------------------------------------------- Print out from our VAXen at Ford on our Xerox 9700 are done by transfering the file by tape. I heard a rumor that the 9700 actually used a PDP-11 to read the tapes and drive the beast. Gee, didn't DEC and Xerox work together on a original Ethernet spec? ================================================================================ Note 405.6 Sharing a Xerox 9700 Printer 6 of 7 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 12 lines 11-JUL-1989 10:58 -< I don't know how simple the PDP-11 connection was >- -------------------------------------------------------------------------------- >> I heard a rumor that the 9700 actually used a PDP-11 to read >> the tapes and drive the beast. Yup, an 11/34, plus an LSI-11 to drive the (optional) 50-bin output sorter. The font disk was based on the CDC 9762 (similar to DEC RM03), modified for parallel-head I/O. I don't know a lot about the beast, but got to work with one while working at Xerox on a message switching system for the feds. As I recall, we managed to drive it directly from a PDP-11/44, but I don't know the details. If you still want to pursue it, I can call my old boss and ask him. ================================================================================ Note 405.7 Sharing a Xerox 9700 Printer 7 of 7 EISNER::HAMPARIAN "Don Hamparian - IDI" 6 lines 11-JUL-1989 14:41 -< Tape works Fine >- -------------------------------------------------------------------------------- At my previous place of employ we had a Xerox 8700 with a tape drive that took tapes from IBM/VM, VMS, CDC NOS, CDC NOS/VE and CDC NOS/BE. Worked great except for the operator overhead. We wrote our own print symbiont on VMS to do the job.  ================================================================================ Note 406.0 TA90 tape specs 6 replies EISNER::ESBENSEN 8 lines 9-JUL-1989 16:21 -------------------------------------------------------------------------------- Does anyone out there have the tape length and density specs for the TA90? I have heard that each cartridge can hold 270MB, but need to also know the tape density. Thanks, Dan E. ================================================================================ Note 406.1 TA90 tape specs 1 of 6 EISNER::HARVEY "Jack Harvey" 6 lines 9-JUL-1989 20:33 -< I'm curious too, about why it's important >- -------------------------------------------------------------------------------- Nothing but a TA90 or IBM's clone will read it.  Just kidding, gang. I know that IBM manufacturers the TA90. ================================================================================ Note 406.2 TA90 tape specs 2 of 6 EISNER::MAYHEW "Bill Mayhew" 1 line 9-JUL-1989 23:22 -< Apr-Jun '89 Systems/Options book sez: 38,000 bpi >- -------------------------------------------------------------------------------- ================================================================================ Note 406.3 TA90 tape specs 3 of 6 EISNER::ESBENSEN 4 lines 10-JUL-1989 00:07 -< thanks >- -------------------------------------------------------------------------------- Thanks for the quick response...38,000 BPI...densities sure are getting up there!! Dan E. ================================================================================ Note 406.4 TA90 tape specs 4 of 6 EISNER::MADISON "Matt Madison" 6 lines 11-JUL-1989 19:45 -< Only 270 MB? >- -------------------------------------------------------------------------------- > Does anyone out there have the tape length and density specs > for the TA90? I have heard that each cartridge can hold 270MB, > but need to also know the tape density. I think its the TK52 (aka TK70) that holds 270 MB. I think the TA90 holds substantially more than that. ================================================================================ Note 406.5 TA90 tape specs 5 of 6 EISNER::MAYHEW "Bill Mayhew" 5 lines 11-JUL-1989 21:42 -< Actually neither of 'em is 270Mb... >- -------------------------------------------------------------------------------- TA90: IBM-manufactured, IBM 3480-"compatible"... 200Mb/cartridge, 38,000 bpi, $Big Buck$ (>$100K subsystem). TK70: DEC-manufactured(?), reads but doesn't write TK50 cartridges, 296Mb/cartridge, 10,000 bpi, medium bucks (~$7K/subsystem). ================================================================================ Note 406.6 TA90 tape specs 6 of 6 EISNER::GLEASON "CyberPunk" 7 lines 12-JUL-1989 00:12 -< It's an assault tape drive >- -------------------------------------------------------------------------------- > TA90: IBM-manufactured, IBM 3480-"compatible"... 200Mb/cartridge, > 38,000 bpi, $Big Buck$ (>$100K subsystem). Yeah, not much per cartridge - but, it's clip fed, full auto. You can shove the fire selector switch over and cut loose ;-)  ================================================================================ Note 407.0 STI RA 8 Multiple Spindle Disks 2 replies EISNER::CHADWICK "Keith Chadwick, Fermilab" 26 lines 10-JUL-1989 11:43 -------------------------------------------------------------------------------- I am interested in hearing from installations which have had experience with the GIGA-RIG [tm] series of drives sold by System Technology Associates, (Models RA 8-550, RA 8-1100, RA 8-2100, RA 8-600, RA 8-1200). The vendor claims that these drives are 100% SDI compatible, and just "plug and play". The ventor has admitted that these drives are based on Exsys Storage Systems Inc.'s ESDI to SDI converter board (which was recently reviewed in Digital News). The vendor further claims to offer "rotational synchronization of spindles as a no cost factory installed option available on request at time of order". We are considering these (and other) drives for future connections to a large CI cluster via HSC as they offer a price per MB that is roughly 50% below the RA-90. We are especially interested in problems encountered with this implementation. I have read HARDWARE_HELP notes 295.*, so in the interests of conserving bandwidth, please do not rehash those notes in this thread. If you do not feel confortable with documenting your experiences with this hardware in this 'public' forum, please send mail. Any such private communications will be 'sanitized' (i.e. your name and affiliation will be removed) prior to their further release (to my management). I have no affilation with System Technology Associates, save as a *POSSIBLE* future customer. -Keith. ================================================================================ Note 407.1 STI RA 8 Multiple Spindle Disks 1 of 2 EISNER::HAMPARIAN "Don Hamparian - IDI" 3 lines 11-JUL-1989 14:43 -< Don't forget FS and Monthly FS Charges >- -------------------------------------------------------------------------------- I would be *very* curious about field service support for this product; in general that issue is a non-issue at purchase time and a suprise at the time somthing breaks. ================================================================================ Note 407.2 STI RA 8 Multiple Spindle Disks 2 of 2 EISNER::RICE "Gary Rice" 10 lines 12-JUL-1989 12:28 -------------------------------------------------------------------------------- One other question you need to ask: Who is going to do your maintenance? We asked DEC to do it on a potential MTI buy and they responded with "Yes we will. The price will be 4 times the cost of our RA90s." It didn't take too many years of maintenance to totally wipe out the up front savings. Gary  ================================================================================ Note 408.0 Emulex disks on UC04 and VMS V5.0 installation No replies EISNER::TILLMAN "Brian Tillman, Smiths Industries." 14 lines 11-JUL-1989 19:12 -------------------------------------------------------------------------------- I have a MicroVAX II on which I've been running VMS V4.5. This MicroVAX's disks are Emulex PDM/170 removable Winchesters driven by a UC04 interface. Under V4.5, the SHOW DEVICE command identified the disks as RA60s. I don't know if this is correct; I don't know what the disks are supposed to mimic. At any rate, that doesn't matter. What does matter is that I've upgraded to VMS V5.0 (in the process of going to V5.1-1). I built the system disk from scratch (i.e., an install of VMS, not an "upgrade" or "update") and in the process, VMS told me it couldn't recognize the device type and that it was assuming the characteristics of an RK07 no less (Huh? That's not even an MSCP disk!). VMS V5.0 installed, as did V5.0-1. The system boots, runs fine, but the SHOW DEVICE command now says "type not yet identified" in the field where the device type (formerly "RA60") goes. Anyone see this before?  ================================================================================ Note 409.0 IEEE to Daytronic A/D Advice Request No replies EISNER::MERUSI 11 lines 12-JUL-1989 08:50 -------------------------------------------------------------------------------- I wasn't sure this would be the proper place to make this request ... anyway .... I am looking for feedback from anybody having experience with using IEEE communicating with the Daytronic System 10K D/A ... We are using the National Instruments IEEE GPIB board on a MicroVAX II. We are experiencing some problems with the Daytronic unit itself, primarily in that it periodically times-out because it doesn't think it got a good command. The thing will work for about 10-20 seconds flawlessly, then go into one of these time-out conditions. It does recover from the time-out, but in another 10-20 seconds, it will happen again. Appreciate hearing from anyone who might have had experience with this sort of thing. Thanx.  ================================================================================ Note 410.0 HP LaserJet II interfacing 4 replies EISNER::MAYHEW "Bill Mayhew" 42 lines 12-JUL-1989 23:07 -------------------------------------------------------------------------------- We have rented an HP LaserJet II short-term to handle a printing job that our LN03 just sort of stares at (7.25x10.5" paper, to be precise). Now comes the fun part, making it work ;-} ... It came with only the User's Guide which contains precious little interfacing details. I can put it into serial mode and the right baud rate and set "Robust XON=ON" (? whatever that means) but it still seems not to flow-control properly on a standard XON/XOFF terminal port. Ideally I'd like to hook this thing to the printer port of a VT320. We did that and ended up having to slow it down to 1200 baud, since it wasn't throttling the data flow properly.* My guess is that it is doing hardware flow control which may not be getting through to the VT320, or which the 320 may be ignoring. I say "may not be" because I couldn't get a DECconnect cable with either of the 2 known DEConnect-to-male-DB25 adapters to work at all (the printer constantly appeared "not ready" to the VT320). I made it work by plugging in an old home-brew male-to-male adapter to a _female_ DECconnect-DB25 adapter, with the other end of the male-to-male going to the printer. I believe this particular male-to-male is wired to permanently assert the "I'm a printer, and I'm alive" line to the VT. So, questions are: 1. Is there a way to persuade the HP to use real XON/XOFF? 2. If not, how can I custom-wire a cable to take the HP's hardware flow control and feed it to the VT320 to make it happy? 3. I have a VT241 kicking around too that could be used in place of the VT320... would that help? 4. Assuming (as yet another configuration) a full 25-pin cable from the LJ II to a DZV11 or DHV11 port, does anyone know if RSX-11M-Plus/MicroRSX will correctly handle the hardware flow control? (This assumes the DZV/DHV can handle it too.) Thanks many times over... *1200 baud works OK until you exhaust the 100-page paper tray, at which point you lose data while you change paper . ================================================================================ Note 410.1 HP LaserJet II interfacing 1 of 4 EISNER::HAHN "DECUServe Vice Chair" 14 lines 13-JUL-1989 08:26 -< Customer assist number >- -------------------------------------------------------------------------------- The standard cable connection to a PC serial port is as follows (from the HP Laser Jet Manual) Also there is a Customer Assist number 208/323-2551 they answer questions and ask vey few of you. HP Laser Jet II Computer (male 25 pin) (Female Connector) Chassis Ground 1- - - - - - - - - - -1 Chassis Ground RD 3- - - - - - - - - - -2 TD TD 2- - - - - - - - - - -3 RD Signal Ground 7- - - - - - - - - - -7 Signal Ground DTR 20- - - - - - - - \ - -5 Clear to Send \ -6 Data Set Ready ================================================================================ Note 410.2 HP LaserJet II interfacing 2 of 4 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 33 lines 13-JUL-1989 16:28 -< Get out your side cutters and soldering iron ! >- -------------------------------------------------------------------------------- According to the manuals this **SHOULD** work. The DEConnect printer port is configured as follows (pg 222 of Installing and Using the VT330/VT340 Video Terminal): 1 DTR 2 TXD 3 Gnd (TXD) 4 Gnd (RXD) 5 RXD 6 DSR Of course the question is, on the printer port is the printer the "Data Set" of the "Data Terminal" (Some one must know ?) On the LaserJet II the pins are as follows (pg 3-81 Getting Started with LaserJet series II) 1 Ghassis Gnd 2 TXD ("to computer") 3 RXD ("from computer") 7 Gnd 20 DTR **DO NOT CONNECT** The reason I say not to connect DTR is that the HP book says "The LaserJet II printer always sends **BOTH** XON/XOFF and DTR." VAXen (and I **ASSUME** VTxxx don't like you to play with DTR ! The trick would be to get a DEConnect cable and snip off one end. Hook DTR (pin 1) to DSR (pin 6); connect all Grounds together (what the heck?); you will probably need to cross TXD to RXD and vice-versa. Bottom line: HP LaserJet II **DO** do XON/XOFF flow control properly. ================================================================================ Note 410.3 HP LaserJet II interfacing 3 of 4 EISNER::MAYHEW "Bill Mayhew" 17 lines 13-JUL-1989 22:42 -< Ready to do battle... >- -------------------------------------------------------------------------------- Thanks for the info, Jack. It looks like that's enough for me to try to cobble something together. We have all the stuff needed to make "custom" DECconnect cables. Also, another DECUServe member is sending me a copy of his LJ II manuals while he's away on vacation (Hooray for DECUServe!!!) so it looks like I should be in good shape! Does it say anything in there about the bits/parity configuration the LJ II likes? (assuming I won't get the loaned manuals for a couple days) I can't find an obvious way to set it. And, yes, I'm pretty certain DECconnect pin 6 is the DSR or "Ready" signal from the printer. I remember I was told once that one could use the printer port on a VT2xx with a non-XON/XOFF printer if one wired that line to the right hardware flow-control pin on the printer, but I can't find anything in the VT3xx docs that confirms or denies this for VT3xxen. ================================================================================ Note 410.4 HP LaserJet II interfacing 4 of 4 EISNER::HYDE "From the laboratory of Dr. Jekyll" 4 lines 14-JUL-1989 12:04 -< should be able to set it for normal xon/xoff >- -------------------------------------------------------------------------------- The "robust xon-xoff" is HP's combination of xon/xoff and and DTR flow control. The HP gear that I've helped folks play with usually has a way to set normal xon/xoff wo that you can have it act like you expect.  ================================================================================ Note 411.0 Modems on DEC423 ports 10 replies EISNER::MAYHEW "Bill Mayhew" 18 lines 17-JUL-1989 18:29 -------------------------------------------------------------------------------- From the SOAPBOX conference: < Note 217.21 by EISNER::MCMICHAEL "Chuck McMichael" > > The VAXstation 2000 also has a > 6-pin connector, but the hardware knows how to compensate and > provide the correct behavior. It took repeated calls before anyone > inside Digital could be found who knew where the difference lay. Eh? My VS2000 has a 25-pin connector and a 9-pin connector (the latter designated as the printer port, the former as the comm port). A MicroVAX 2000 has 3 DEC423 MMJ 6-pin ports, but also has a 25-pin connector which is designated for the modem. So I don't understand the hardware you're describing here, but if you've found a way to make a modem work on a 6-pin DECconnect port, I'm sure interested in hearing about it! ================================================================================ Note 411.1 Modems on DEC423 ports 1 of 10 EISNER::BRYANT "Geoff Bryant" 13 lines 21-JUL-1989 11:11 -< Seems like a common problem >- -------------------------------------------------------------------------------- I hope someone can answer this question (I saw the thread this morning and when asked later I said the answer was out there on DECUServe! - oh well, just not yet!). I just got a call from someone who is trying to hook up a modem to the 6 pin connector on the VS3100. He gets the same result as SOAPBOX 217.*: - /NOMODEM does allow things to work, but you loose having the modem hang up the line - If /MODEM, the device driver is toggling DTR on/off every 30 seconds. So, how do you hook up a modem to a VS3100? ================================================================================ Note 411.2 Modems on DEC423 ports 2 of 10 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 19 lines 24-JUL-1989 08:55 -< Buy a DECserver 200 >- -------------------------------------------------------------------------------- The bottom line is "You can't" ! At least not in the sense of it working like a traditional VAX with dial in modems. The problem as noted is that none of the **STANDARD** ports on the VS3100 have modem control. This is required by VMS to know when the phone has been hung up so as to kill the process ( or at least disconnect it). The only solution is what is listed in the title. Buy a DECserver 200 ! Make certain to get the version that supports modem control (DSRVB-AA). Reading between the lines, the new microVAX 3100 can use the microVAX 2000 DSH32 multi-function communication controller. This board has 1 sync port (with modem control) and 8 async ports (only **ONE** of which has modem control). The question is can you squeeze a DSH32 inside a VS3100 (or inside a VS2000 for that matter). ================================================================================ Note 411.3 Modems on DEC423 ports 3 of 10 EISNER::MAYHEW "Bill Mayhew" 13 lines 24-JUL-1989 13:14 -< Yes it fits, but will it work? >- -------------------------------------------------------------------------------- > The question is can you squeeze a DSH32 inside a VS3100 (or inside > a VS2000 for that matter). The answer is "yes"... although on the 2000, it requires a "different" expansion adapter... BA40A-BA, I believe, vs. the original BA40A-AA; the way I keep 'em straight is the price; the new one lists at ~$600, while the original is ~$1250. But I'm not sure this is an answer because I "was" certain that the async ports on the DSH32 were all *non* modem-control-equipped. Guess I'll have to go back and re-learn... ================================================================================ Note 411.4 Modems on DEC423 ports 4 of 10 EISNER::MCMICHAEL "Chuck McMichael" 32 lines 24-JUL-1989 17:31 -< Still looking for my notes >- -------------------------------------------------------------------------------- > someone who is trying to hook up a modem to the 6 pin connector > on the VS3100. He gets the same result as SOAPBOX 217.*: > > - /NOMODEM does allow things to work, but you loose having the > modem hang up the line > > - If /MODEM, the device driver is toggling DTR on/off every > 30 seconds. > > So, how do you hook up a modem to a VS3100? I believe that after 4 minutes the modem will hang up. (Can you say "Inflated phone bills?" I thought you could.) See note 358.* for more discussion: <<< EISNER::DUA0:[NOTES$LIBRARY]HARDWARE_HELP.NOTE;1 >> ================================================================================ Note 358.3 VAXstation 3100, Scholar Modem, 'Nuff Said 3 of 3 EISNER::MCMICHAEL 8 lines 15-MAY-1989 15:38 -< VAXstation 2000 vs. VAXstation 3100 >- -------------------------------------------------------------------------------- The latest word I have from Field Service is that the VAXstation 2000 data line has a control register with bits for DSR, DTR, & CD. When DSR is turned on, DTR & CD are forced. The VAXstation 3100 was engineered with the data line control register containing a bit for DSR, but not DTR nor CD. Thus the problem. Other customers have complained, and DEC Engineering is reportedly arguing internally whether to modify the hardware, or come up with a software fix. ================================================================================ Note 411.5 Modems on DEC423 ports 5 of 10 EISNER::MCMICHAEL "Chuck McMichael" 16 lines 25-JUL-1989 17:38 -< Here's my best shot >- -------------------------------------------------------------------------------- I spoke with the fellow who sets up modems in our company, and he gave me the following diagram: __________ _______ __ _______ | | |----------| | | VAXstation| | |----------| | MODEM | 3100 |--------------------------| |----------|> | | | DECconnect Office Cable |_______| ^ -- ------- ----------- (6-pin) H8571-C | ^ | | Pins 2,3,7 Strap straight 6 & 20 through together The VAXstation 3100 still has to have the port set /NOMODEM ================================================================================ Note 411.6 Modems on DEC423 ports 6 of 10 EISNER::MAYHEW "Bill Mayhew" 9 lines 25-JUL-1989 17:53 -< DSH32 doesn't have modem control on async ports >- -------------------------------------------------------------------------------- > I "was" certain that > the async ports on the DSH32 were all *non* modem-control-equipped. I went back and checked a couple sources and, based on the published specs I can find and short of actually having one and trying it, I was correct: the only modem-controlled port on the DSH32 is the *sync* port; the 8 async ports are plain-old DEC423, with no modem control. Thus the DSH32 is not a solution to this problem. ================================================================================ Note 411.7 Modems on DEC423 ports 7 of 10 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 30 lines 26-JUL-1989 11:55 -< New controller ! >- -------------------------------------------------------------------------------- From the E-Store description of the microVAX 3100 ... TABLE OF PACKAGED STANDARD SYSTEMS Entry-level Medium Large System System System ----------- ------- ------ Communications 3 DEC423 3 DEC423 11 DEC423 asynchronous asynchronous asynchronous serial lines serial lines serial lines 1 RS-232 1 RS-232 1 RS-232 asynchronous serial asynchronous asynchronous line with serial line serial line modem control with modem control with modem control 1 RS-232 synchronous communications controller with full modem control I am **ASSUMING** that the MV/VS3100 uses the **NEW** DSH32-B* controller as oppsed to the MV/VS2000 DSH32-A* controller. Ther are no details available on the E-Store for tis device. ================================================================================ Note 411.8 Modems on DEC423 ports 8 of 10 EISNER::MAYHEW "Bill Mayhew" 3 lines 26-JUL-1989 12:18 -< Not a DSH32 issue >- -------------------------------------------------------------------------------- But note that that "modem control" async port exists on all three configs cited, including the entry-level one, which contains NO DSH32, only the on-board ports. ================================================================================ Note 411.9 Modems on DEC423 ports 9 of 10 EISNER::MCMICHAEL "Chuck McMichael" 16 lines 26-JUL-1989 17:28 -< clarification >- -------------------------------------------------------------------------------- Clarification: The modem I'm using is a Scholar Plus and the homebrew connector is strapping pins 6 & 20 of the modem together. __________ _______ __ _______ | | |----------| |scholar| VAXstation| | |----------| |plus | 3100 |--------------------------| |----------| <|modem | | DECconnect Office Cable |_______| ^ -- ------- ----------- (6-pin) H8571-C | ^ | | Pins 2,3,7 Strap straight 6 & 20 through together The VAXstation 3100 still has to have the port set /NOMODEM ================================================================================ Note 411.10 Modems on DEC423 ports 10 of 10 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 7 lines 7-AUG-1989 09:47 -< DSH32 and VS3100 don't mix >- -------------------------------------------------------------------------------- >>The question is can you squeeze a DSH32 inside a VS3100 ... >The answer is "yes"... I saw a copy of Sales Update the other day that indicated that the only way to hook a VS3100 or DS3100 to a WAN was with a some other box (ie. another VAX or a MicroServer).  ================================================================================ Note 412.0 "RS422 synchronous video"? 7 replies EISNER::HASSINGER "Bob Hassinger" 8 lines 17-JUL-1989 23:53 -------------------------------------------------------------------------------- The press release on the new LN03 Image Printer that was posted in INDUSTRY_NEWS refers to RS422 synchronous video. I don't think I recognize this reference. Can anyone give some details? I have always regarded "video" as referring to very distinctly analog signal formats. This reference seems to be talking about a format that must be digital. ================================================================================ Note 412.1 "RS422 synchronous video"? 1 of 7 EISNER::BURNET "John Burnet" 13 lines 18-JUL-1989 02:23 -< RS422 is a different beast entirely >- -------------------------------------------------------------------------------- > The press release on the new LN03 Image Printer that was posted in > INDUSTRY_NEWS refers to RS422 synchronous video. I don't think I > recognize this reference. Can anyone give some details? The EIA standard number must have been garbled somewhere along the line, since RS422 is an asynchronous, bit-serial, differential-signaling standard for digital communication. Loosely speaking, the degenerate case of RS422 is RS232 (single transmitter, single receiver, single-ended signaling rather than differential). The Mac's serial I/O, as an example, adheres to the RS422 standard. What the correct sync-video standard number *is*, I don't know. Does "5 MB" refer to megabytes/second, megabits/second, or what? ================================================================================ Note 412.2 "RS422 synchronous video"? 2 of 7 EISNER::LEDERMAN "Bart Z. Lederman" 6 lines 18-JUL-1989 09:35 -< One video standard. >- -------------------------------------------------------------------------------- > What the correct sync-video standard number *is*, I don't know. Does "5 MB" Usually, RS-170. This is the 75 ohm baseband video, such as you get on the video cables on VCRs and cameras, and the analog connection between Rainbow and PRO PCs and their video monitors (also VT-240s and the ouput from VT-125s, more or less). ================================================================================ Note 412.3 "RS422 synchronous video"? 3 of 7 EISNER::HASSINGER "Bob Hassinger" 31 lines 18-JUL-1989 10:42 -< See what I mean? >- -------------------------------------------------------------------------------- >The EIA standard number must have been garbled somewhere along the line, >since RS422 is an asynchronous, bit-serial, differential-signaling >standard for digital communication. Loosely speaking, the degenerate >case of RS422 is RS232 (single transmitter, single receiver, single-ended >signaling rather than differential). Yup, that is what I thought too. Note however the press release explicitly says "RS-422" three times. >Does "5 MB" refer to megabytes/second, megabits/second, or what? The release says "5MB" twice. Of course in my idea of "video", you would expect this to be expressed in MHz - and this would be a fairly typical bandwidth for conventional television video. >What the correct sync-video standard number *is*, I don't know. I am familiar with references to the RS-170 mentioned. It is common in television video. I have never seen the spec to know what it actually spells out but based on the way it is used, I suspect it covers a lot more than what was mentioned. I would expect it to spell out many of the myriad details of a television signal including things like timing, sync pulse, interlace and signal levels for example. I am not clear how far RS-170 goes into the NTSC standard used for color television encoding. When I think of the term "video" I think of two type of signal. Obviously the most common use now is for television pictures. The earlier use is for radar. Both are structured analog signals that have nothing to do with digital data, synchronous or otherwise. That is why I asked. ================================================================================ Note 412.4 "RS422 synchronous video"? 4 of 7 EISNER::RICE "Gary Rice" 8 lines 20-JUL-1989 10:55 -------------------------------------------------------------------------------- > Yup, that is what I thought too. Note however the press release > explicitly says "RS-422" three times. It is quite common for press releases to be written by the PR Department where the most technical thing they do is turn on their PC each morning. Gary ================================================================================ Note 412.5 "RS422 synchronous video"? 5 of 7 EISNER::HASSINGER "Bob Hassinger" 1 line 20-JUL-1989 13:07 -< No matter, I am still interested in the connection... >- -------------------------------------------------------------------------------- ================================================================================ Note 412.6 "RS422 synchronous video"? 6 of 7 EISNER::MAYHEW "Bill Mayhew" 23 lines 20-JUL-1989 15:09 -< Another puzzle piece appears >- -------------------------------------------------------------------------------- Potentially relevant quote from an article in the 7/89 VARBUSINESS which appeared today... "Think of it as a trial separation. Ricoh Corp. [maker of the LN03 engine] is taking the controller out of its printers. The West Caldwell, NJ printer vendor will just manufacture printer engines. Making and selling the controllers will be left to controller and board vendors... "Ricoh is supplying controller vendors with specifications on its external *video* interface so they can design hardware to fit the Ricoh printer engines, according to Dell Glover, marketing director of the peripherals division. He says more than a dozen printer controller vendors are working on controller cards for Ricoh's external video interface." [Emphasis mine. 3 vendors of the more-than-a-dozen were listed, did not include DEC.] (The thrust of the article was about the business opportunity for VARs in dealing with such "open-architecture" printers, allowing the printer to be configured with different types of controller boards for different purposes, etc.) My suspicion is that DEC is one of the more-than-a-dozen, but this is just speculation on my part. ================================================================================ Note 412.7 "RS422 synchronous video"? 7 of 7 EISNER::MERRIMAN "George Merriman - CCA/NY" 7 lines 21-JUL-1989 15:58 -< RS422 both synchronous and async >- -------------------------------------------------------------------------------- > The EIA standard number must have been garbled somewhere along the > line, since RS422 is an asynchronous, bit-serial, differential- > ^^^^^^^^^^^^ For the record, RS422 and RS232 both cover synchronous and asynchronous transmission.  ================================================================================ Note 413.0 CD reading woes 7 replies EISNER::GEIER "Jim Geier, General Instrument" 11 lines 18-JUL-1989 15:36 -------------------------------------------------------------------------------- At the 1989 DECUS Symposium in Atlanta, I got a CD called "VMS ON LINE" from the Software Bookstore. When I place it into an RRD50, the red light comes on indicating an error. We use the RRD50 successfully to read other CDs, so I do not think the RRD50 is suffering a hardware problem. Could the CD be in a different format than what the RRD50 expects? Is it likely that the CD is somehow defective? Do I just need to find out from whom I can get a replacement CD? Then, of course, my next question is from whom can I get a replacement? Just who is the "DECUS Software Bookstore?" (no address on the receipt) ================================================================================ Note 413.1 CD reading woes 1 of 7 EISNER::DELARISCH "Arnold S. De Larisch" 28 lines 18-JUL-1989 17:23 -< The "VMS ON LINE" CD give-away is BLANK! >- -------------------------------------------------------------------------------- >> < Note 413.0 by EISNER::GEIER "Jim Geier, General Instrument" > >> -< CD reading woes >- >> At the 1989 DECUS Symposium in Atlanta, I got a CD called >> "VMS ON LINE" from the Software Bookstore. When I place it into >> an RRD50, the red light comes on indicating an error. We use the >> RRD50 successfully to read other CDs, so I do not think the RRD50 >> is suffering a hardware problem. Nor do I think you have a CDROM reader hardware problem .... actually DECUS Software Bookstore has played a 'bit of a joke' on you ... the CD you have is BLANK ... yes BLANK ... it has NO INFORMATION ON IT! I think it was a stupid 'give away' ... I've got my hanging on the wall along with my RK07 disk platter, Paper tape, and PDP 11/34 boot ROM (Yes ... the one with all the Diodes on it). I used the jewel box as a replacment for a broken one in my Audio CD collection. >> Could the CD be in a different format than what the RRD50 expects? >> Is it likely that the CD is somehow defective? >> Do I just need to find out from whom I can get a replacement CD? >> Then, of course, my next question is from whom can I get a replacement? >> Just who is the "DECUS Software Bookstore?" (no address on the receipt) As for the "DECUS Software Bookstore" I believe that is is an offshoot of DEC Press and/or Educational Services 'Bookstore'. -Arnold ================================================================================ Note 413.2 CD reading woes 2 of 7 EISNER::GEIER "Jim Geier, General Instrument" 10 lines 18-JUL-1989 22:06 -< Having the real info solves the problem again >- -------------------------------------------------------------------------------- Then I think I'll take advantage of a person within Digital who told me that they would give me a CD of VMS Documentation if I could not get it replaced via DECUS. I think they owe us... we actually lived through the VMS 5.2 field test! Thank goodness that is over! (Do not take that comment to mean that 5.2 will be available to anyone at any time!) Rather then getting a red error light, I guess I expected some other indication that the CD was blank, or some file on the CD telling me what it really was, a blank promo. ================================================================================ Note 413.3 CD reading woes 3 of 7 EISNER::KENNEDY "Terry Kennedy" 14 lines 19-JUL-1989 00:40 -< *Almost* blank >- -------------------------------------------------------------------------------- > Rather then getting a red error light, I guess I expected some > other indication that the CD was blank, or some file on the CD telling > me what it really was, a blank promo. Well, it *does* have a directory (not in terms of VMS fiel structure, but in terms of CD formaat) which says there are 10 tracks and the play time is 47:22. However, that is all. The block numbers pointed to by the directory are non-existant. I asked how they were moving and the pserson in the booth told me that once she told people they were blank, they said "no thanks". I offered to take a bunch off her hands and she thanked me and gave 'em to me! Perhaps they were being graded by how many they unloaded? (Like Arnold, I have a lot of audio CD's w/ broken cases...) ================================================================================ Note 413.4 CD reading woes 4 of 7 EISNER::LEDERMAN "Bart Z. Lederman" 8 lines 19-JUL-1989 07:44 -< You went to the wrong booth. >- -------------------------------------------------------------------------------- Gee, I wish I knew they were giving them away, I could use a few blanks and boxes! Anyway, while it's probably not the same as what you were looking for, did you consider purchasing the DECUS Library CD-ROM? It has several VAX SIG and L&T SIG symposium tapes on it, plus several other popular packages. It's almost full: over 600 MB of software. It's still available from the Library. ================================================================================ Note 413.5 CD reading woes 5 of 7 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 11 lines 19-JUL-1989 10:50 -< Mine is also on the wall >- -------------------------------------------------------------------------------- >> I used the jewel box as a replacment for a broken one in my >> Audio CD collection. They also come in handy now that DEC no longer ships the VMS CDROM in a jewel box, but in the RRD40 carrier. I, too, was very disappointed in this joke(?) from DEC. Since it had to be run through the press to stamp the directory information on it, they could at least have put *something* on it. I would really like to have copies of the VAXstation demos they show at DECUS. ================================================================================ Note 413.6 CD reading woes 6 of 7 EISNER::HASSINGER "Bob Hassinger" 7 lines 19-JUL-1989 20:52 -< There actually was some logic ... >- -------------------------------------------------------------------------------- I was told at the Symposium by one of the people handing them out that they had wanted to have a smaller one but the supplier could not tool up in time so they had to go with the regular size and there was no way to put anything on them in the available time. I think the idea was the small one would have been more of a jewelry type thing that would not be mistaken for the real thing - to be used just to promote the idea of the new CD distributions. ================================================================================ Note 413.7 CD reading woes 7 of 7 EISNER::BONE "George Bone, MINSY" 8 lines 26-JUL-1989 12:08 -< Whadya mean, there's no santa claus?? >- -------------------------------------------------------------------------------- I didn't see them put this thingie into my bag, so later that evening when it slid out I was amazed, and anxious to see what was on it. Unfortunately we hadn't received our 3100 w/rrd40 yet, so I was STILL waiting anxiously up to today. Now you've told me. Why do I feel just a little bit irritated??  ================================================================================ Note 414.0 "3800/3820" printers? Equivalents for DEC systems? 9 replies EISNER::HASSINGER "Bob Hassinger" 9 lines 19-JUL-1989 10:47 -------------------------------------------------------------------------------- The following question was posted on our company's internal Notes system today: "Is there an equivalent printer to the 3800 or 3820 in the DEC world? Letter quality, fast, and very large volume." I don't recognize the reference to 3800/3820. I presume the basic context is the IBM world. Is anyone familiar with these printers? Any ideas on "equivalents" in the DEC world? ================================================================================ Note 414.1 "3800/3820" printers? Equivalents for DEC systems? 1 of 9 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 16 lines 19-JUL-1989 18:18 -< IBM & KodaK have printers >- -------------------------------------------------------------------------------- I have no idea about those numbers, but... Kodak has a BIG laser up near 100 ppm, and now it has Postscript. Dec has just announced a cooperative marketing thingie with them, so these should be getting more visible in the DEC world. Check recent IND news postings. They look like the same engine used for their BIG office copiers, and probably also accept the 'finisher' hardware such as staplers etc. that let you crank out finished manuals. The laser printer itself is about $100,000. At DEXPO, IBM was selling one of their fancy impact printers setup for the DEC marketplace. I assume that all they did was to tack on a Dataproducts interface. I think it was just under $30,000. ================================================================================ Note 414.2 "3800/3820" printers? Equivalents for DEC systems? 2 of 9 EISNER::COY "Dale E. Coy" 33 lines 19-JUL-1989 22:40 -< You called? >- -------------------------------------------------------------------------------- > "Is there an equivalent printer to the 3800 or 3820 in the > DEC world? Letter quality, fast, and very large volume." [so, the guy who WORKS for IBM shows up, and says...] I can get whatever info you need on these, and I'm sorta familiar with the characteristics. All of the 38xx species are laser-type (not sure of the exact technology), and support both built-in and down-loadable fonts, placement of things anywhere on the page, etc. Similar in general printing capabilities to the LN-03+ (I'm sure I could get arguments about that, but...) The 3812 is just a bit bigger than an LPS-40, overall. I *think* that it's speed is 12 pages per minute. At any rate, the slowest of the line. I've never seen a 3820, but I was *told* that it prints about 20 pages per minute. There's also a 3825, which prints about 55-or-so pages per minute. All of the above are sheet-feed page printers. You can sorta find DEC equivalents (in addition to several 3rd parties). Then, there's the 3800. Still a laser-type, but this hummer is continuous-feed because there's no practical way to sheet-feed that fast. Don't know the printing rate, but it's *FAST*. Eyeball estimate says maybe 120 pages per minute, but I could be way off. Obviously, it's also *BIG*. [as noted, I'm employed by IBM] ================================================================================ Note 414.3 "3800/3820" printers? Equivalents for DEC systems? 3 of 9 EISNER::KOZAM 11 lines 19-JUL-1989 23:11 -< About 120 ppm is right >- -------------------------------------------------------------------------------- > Then, there's the 3800. Still a laser-type, but this hummer is > continuous-feed because there's no practical way to sheet-feed that > fast. Don't know the printing rate, but it's *FAST*. Eyeball estimate > says maybe 120 pages per minute, but I could be way off. 120 ppm is about right. This amazing machine requires special tall boxes of paper, otherwise the operators would be changing supplies every few minutes. A nice feature is the built in vacuum cleaner - at the capacity of this beast, you get lots of paper dust pretty quickly. [I WAS employed by IBM] ================================================================================ Note 414.4 "3800/3820" printers? Equivalents for DEC systems? 4 of 9 EISNER::ERSKINE "Chris Erskine" 1 line 20-JUL-1989 06:55 -< I thought Xerox 9700 is 120 PPM also (cut sheet) >- -------------------------------------------------------------------------------- ================================================================================ Note 414.5 "3800/3820" printers? Equivalents for DEC systems? 5 of 9 EISNER::HASSINGER "Bob Hassinger" 14 lines 20-JUL-1989 12:54 -< Sounds like the right ballpark >- -------------------------------------------------------------------------------- Seems like the 3800 at 120 ppm is what they must be talking about. This morning there was a note suggesting a 2000 lpm band printer or a 40 ppm laser where the best available on DEC and they would not do the job. There was also a mention of Interlink working on a DQS sort of thing to print on IBM printers from DEC "transparently" and also a note saying DEC was rumored to be working on the same sort of thing. > -< I thought Xerox 9700 is 120 PPM also (cut sheet) >- What is the story on getting a 9700 going on DEC? Also, I seem to recall a presentation about one of these big laser printers a few years ago that talked about a roll feed (almost like a newspaper press I think). ================================================================================ Note 414.6 "3800/3820" printers? Equivalents for DEC systems? 6 of 9 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 1 line 20-JUL-1989 13:19 -< See 405. >- -------------------------------------------------------------------------------- >> What is the story on getting a 9700 going on DEC? ================================================================================ Note 414.7 "3800/3820" printers? Equivalents for DEC systems? 7 of 9 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 3 lines 20-JUL-1989 23:17 -< Know Bull? >- -------------------------------------------------------------------------------- Bull has a magnetographic printer that is supposed to do 90 ppm and work on VAXen. As I recall, its pretty cheap, maybe $60k. See the adds in the trade rags. ================================================================================ Note 414.8 "3800/3820" printers? Equivalents for DEC systems? 8 of 9 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 3 lines 24-JUL-1989 08:37 -< Re-read 414.1 >- -------------------------------------------------------------------------------- Check out the Kodak new laser printer ! It's fast (don't know how fast) and is part of the Digital printing architecture (or what ever DEC's buzz words are). ================================================================================ Note 414.9 "3800/3820" printers? Equivalents for DEC systems? 9 of 9 EISNER::COY "Dale E. Coy" 7 lines 26-JUL-1989 23:46 -< Good walking speed >- -------------------------------------------------------------------------------- 3800-6 printer is rated at 134 pages-per-minute 3800-3 is rated at 215 ppm Note: 215 ppm is close to 5 miles per hour. If anyone is sufficiently interested, I have more details available now.  ================================================================================ Note 415.0 DRQB3 Question ... 2 replies EISNER::MERUSI 5 lines 19-JUL-1989 14:27 -------------------------------------------------------------------------------- Looking for anyone having had experience with the DRQB3 parallel interface. We're trying to use more than one of these things on a VAX/GPX and apparently you cannot have more than one of these things in use on the same processor at the same time. Has anyone else confirmed this? Has anyone been successful in getting a double configuration to work? Thanx. ================================================================================ Note 415.1 DRQB3 Question ... 1 of 2 EISNER::DELARISCH "Arnold S. De Larisch" 15 lines 19-JUL-1989 16:35 -< VAXlab Software Library Supports Multiple DRQ3Bs >- -------------------------------------------------------------------------------- >> < Note 415.0 by EISNER::MERUSI > >> -< DRQB3 Question ... >- >> Looking for anyone having had experience with the DRQB3 parallel interface. >> We're trying to use more than one of these things on a VAX/GPX and apparently >> you cannot have more than one of these things in use on the same processor at >> the same time. Has anyone else confirmed this? Has anyone been successful in >> getting a double configuration to work? Thanx. Look into the VAXlab Software Library product. It seems to support more that one DRQ3B. I don't have the SPD (Software Product Description) in front of me so I won't say for sure. The DRQ3B is FAST and can HOG the bus. Use common sence when configuring these fast beasts! -Arnold ================================================================================ Note 415.2 DRQB3 Question ... 2 of 2 EISNER::HARVEY "Jack Harvey" 14 lines 19-JUL-1989 20:39 -< Check VAXlab in Systems and Options >- -------------------------------------------------------------------------------- SPD 28.30.02 for the VAXlab software library (formerly Labstar) mentions the DRQ3B as a supported device, but I can't find a limit on how many. However, VAX Systems and Options Catalog (Black Book) has a bunch of standard VAXlab configurations including the VAXlab software license. The ordering guide sections typically state: "A maximum of three each of the interfaces - DRV11-J, DRV11-WA and DRQ3B-AA - can be ordered for each system." *Three each* is pretty convincing, but Arnold's remark about bus hogging should be considered.  ================================================================================ Note 416.0 Micropolis RD53 problem 15 replies EISNER::SMITHJR "Charles T. Smith, Jr." 8 lines 20-JUL-1989 00:38 -------------------------------------------------------------------------------- I have access to several Micropolis drives that appear to be very close to the RD53. The drives look almost identical, except for a couple components on the motor control board, and there's one suffix difference in the part number. These drives will work with a PC/AT without mods, and came off a wang system. User microvax diags recognize them as a RD53, but they fail during format. Anyone got any ideas about how to get these drives to work as RD53's? ================================================================================ Note 416.1 Micropolis RD53 problem 1 of 15 EISNER::KENNEDY "Terry Kennedy" 4 lines 20-JUL-1989 03:53 -< Seen note 93.2 here >- -------------------------------------------------------------------------------- > ...RD53, but they fail during format. Anyone got any ideas about how to > get these drives to work as RD53's? Yes. See note 93.2 in this conference for the details. ================================================================================ Note 416.2 Micropolis RD53 problem 2 of 15 EISNER::SMITHJR "Charles T. Smith, Jr." 9 lines 21-JUL-1989 21:48 -< Well, we're making progress? >- -------------------------------------------------------------------------------- > Yes. See note 93.2 in this conference for the details. Installed the R7 jumper per the instructions. This did create a difference... the format ran, and ran, and ran. After a couple hours, I stopped it, started over, and still continued to run. This drive was recently rebuilt, and was working on a PC system. As stated above, it was originally from a WANG system. Any ideas? ================================================================================ Note 416.3 Micropolis RD53 problem 3 of 15 EISNER::KENNEDY "Terry Kennedy" 9 lines 22-JUL-1989 05:52 -< *Lots* of questions >- -------------------------------------------------------------------------------- > Installed the R7 jumper per the instructions. This did create a difference... > the format ran, and ran, and ran. After a couple hours, I stopped it, started > over, and still continued to run. What diagnostic? On an 11, it's ZRQC??. Does it ask any questions like "use exiting ..." or "continue even if ... is unaccessible"? These need to be an- swered NO and YES, respectively. Does the drive light come on? What kind of box is the drive in? What are the DS jumpers (drive select) set to? How many other drives? What controller? ================================================================================ Note 416.4 Micropolis RD53 problem 4 of 15 EISNER::SMITHJR "Charles T. Smith, Jr." 34 lines 22-JUL-1989 17:42 -< The config... >- -------------------------------------------------------------------------------- > What diagnostic? On an 11, it's ZRQC??. The formatter is the microvax II customer formatter. The formatter prints a warning about this procedure will wipe all existing data, and to write protect any drives not being formatted. It then runs a couple minutes, then promts for drive type, to which I answer RD53. Then, it goes out and beings formatting, printing a status message "Format continuing, xxx minutes." >Does it ask any questions like "use >exiting ..." or "continue even if ... is unaccessible"? These need to be an- >swered NO and YES, respectively. No... >Does the drive light come on? Yes. Steadily. >What kind of box is the drive in? BA123. >What are the DS jumpers (drive select) set to? DS3 is set. > How many other drives? Two "real" RD53's installed as DUA0 and DUA1. An RX50. >What controller? RQDX3. ================================================================================ Note 416.5 Micropolis RD53 problem 5 of 15 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 34 lines 22-JUL-1989 21:55 -< Another $0.02... >- -------------------------------------------------------------------------------- < Note 416.4 by EISNER::SMITHJR "Charles T. Smith, Jr." > >promts for drive type, to which I answer RD53. Then, it goes out and beings >formatting, printing a status message "Format continuing, xxx minutes." FWIW, one of the problems resolved by the newest version of the RQDX3 firmware was some type of looping format that never finished. I think it revolved around not doing BBR correctly, though I think it mentioned this in connection with RD54s. On the other hand, with the original firmware, my RQDX3 wouldn't format any type of drive, DEC or not, so it might be that a firmware upgrade is called for... >>What kind of box is the drive in? >BA123. >>What are the DS jumpers (drive select) set to? >DS3 is set. Shouldn't matter which DS jumper is in place in a BA123. The important thing is that there's a jumper installed. >> How many other drives? >Two "real" RD53's installed as DUA0 and DUA1. An RX50. Correct me if I'm mistaken, but doesn't the RQDX3 support a maximum of four units, and that an RX50 is considered to be two units? If this is so, you're at the limit of the RQDX3's capability right now. Why don't you try losing one of the "real" RD53s, and putting the one you want formatted in its place? Ed ================================================================================ Note 416.6 Micropolis RD53 problem 6 of 15 EISNER::KENNEDY "Terry Kennedy" 6 lines 22-JUL-1989 22:27 -< Correct, 4 units is the limit per RQDX3 >- -------------------------------------------------------------------------------- > Correct me if I'm mistaken, but doesn't the RQDX3 support a maximum > of four units, and that an RX50 is considered to be two units? If this is > so, you're at the limit of the RQDX3's capability right now. Why don't you > try losing one of the "real" RD53s, and putting the one you want formatted > in its place? ================================================================================ Note 416.7 Micropolis RD53 problem 7 of 15 EISNER::SMITHJR "Charles T. Smith, Jr." 10 lines 24-JUL-1989 20:36 -< Still no go >- -------------------------------------------------------------------------------- > > Correct me if I'm mistaken, but doesn't the RQDX3 support a maximum > > of four units, and that an RX50 is considered to be two units? If this is > > so, you're at the limit of the RQDX3's capability right now. Why don't you > > try losing one of the "real" RD53s, and putting the one you want formatted > > in its place? We tried unhooking the RX50; no change. We've found that in this config, drive 2 of the RX50 is not configured; 1 works normally. ================================================================================ Note 416.8 Micropolis RD53 problem 8 of 15 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 0 lines 24-JUL-1989 21:17 -< WHAT ARE YOU MOUNTING THESE IN? >- -------------------------------------------------------------------------------- ================================================================================ Note 416.9 Micropolis RD53 problem 9 of 15 EISNER::MAYHEW "Bill Mayhew" 1 line 24-JUL-1989 21:32 -< What's the Micropolis model number of the drive? >- -------------------------------------------------------------------------------- ================================================================================ Note 416.10 Micropolis RD53 problem 10 of 15 EISNER::SMITHJR "Charles T. Smith, Jr." 4 lines 26-JUL-1989 12:30 -------------------------------------------------------------------------------- > -< WHAT ARE YOU MOUNTING THESE IN? >- BA123 ================================================================================ Note 416.11 Micropolis RD53 problem 11 of 15 EISNER::SMITHJR "Charles T. Smith, Jr." 8 lines 26-JUL-1989 12:31 -< RD53W >- -------------------------------------------------------------------------------- > -< What's the Micropolis model number of the drive? >- 1325W (The Real RD53 is a 1325D) ================================================================================ Note 416.12 Micropolis RD53 problem 12 of 15 EISNER::SMITHJR "Charles T. Smith, Jr." 15 lines 1-AUG-1989 00:48 -< Further along... >- -------------------------------------------------------------------------------- The saga continues... Concerned that we might have accidently made contact with another line when we installed the jumper, we re did it, hooked it up to an RQDX3. The format blew off after ten minutes or so. Took the logic board off the drive, hooked board up to anote drive, got same results as above (blew off after 5 minutes). Next, we try a format using an RQDX2, and the sucker works...on both drives (using same interface board) At this point, does anyone have any idea of how to get thi disk up and unning on an RQDX3? ================================================================================ Note 416.13 Micropolis RD53 problem 13 of 15 EISNER::KENNEDY "Terry Kennedy" 4 lines 1-AUG-1989 03:27 -< Try this FCO >- -------------------------------------------------------------------------------- > At this point, does anyone have any idea of how to get thi disk up and > running on an RQDX3? Get FCO kit EQ-01532-02, which corrects formatter bugs in the RQDX3? ================================================================================ Note 416.14 Micropolis RD53 problem 14 of 15 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 18 lines 1-AUG-1989 15:53 -< I'd suspect RQDX firmware revs... >- -------------------------------------------------------------------------------- >< Note 416.13 by EISNER::KENNEDY "Terry Kennedy" > > -< Try this FCO >- > >> At this point, does anyone have any idea of how to get thi disk up and >> running on an RQDX3? > > Get FCO kit EQ-01532-02, which corrects formatter bugs in the RQDX3? Terry's got a point; my raggedy ol' RQDX3 wouldn't format *anything* until I got a new version of the firmware. I don't know if this is the FCO I have installed, but when I get into diags, it says my RQDX3 is at version 4. I'd be happy to try to format the drive on my system, but given the distances involved, it would probably be easier if you upgraded your RQDX3... Ed P.S. What versions are: A: your diags, and B: your RQDX3? ================================================================================ Note 416.15 Micropolis RD53 problem 15 of 15 EISNER::PROVOST 21 lines 4-AUG-1989 11:30 -< You're not alone >- -------------------------------------------------------------------------------- On the theory that misery loves company I will relate our experience with Micropolis drives. We had purchased a Micropolis RD53 equivalent from TRIMARCHI with skid plate, indicator panel, cables, etc. It worked fine. Eventually, during a powerfail which took out the RQXD3, the drive failed. We purchased a cheap replacement from a Micropolis distributer. After several weeks of moving jumpers, switching drives and logic boards, and attempting format, we sent it back. We purchased a replacement from TRIMARCHI for a little more money. Formatting was unnecessary. Drive worked fine. TRIMARCHI tech people were very helpful. They mentioned that they change two components on the board, and format the drive. I don't know if the difference is in the components or in the formatter or in the RQXD3 or equivalent. Tom  ================================================================================ Note 417.0 LN03 bleeding! 13 replies EISNER::KOLP 14 lines 22-JUL-1989 23:00 -------------------------------------------------------------------------------- I got a DEC Postscript printer (LN03R?) for desktop publishing and have experienced a problem with the print shadowing on some of the bold characters. Digital has attempted to fix the problem and the quality has gotten somewhat better but still does not produce 'camera ready' documents. Their next move is to replace the printer. Our Field Service tech has said, off the record, that this is a problem inherent in the LN03 printer and probably will not be resolved. I feel an argument coming up with DEC about getting a refund and using the refund for a 3rd party printer such as Apple. Has anyone experienced this problem with the print or am I a lucky one? What solution is there? Should I scrap the DEC solution for another vendor? Thanks in advance for any input. ================================================================================ Note 417.1 LN03 bleeding! 1 of 13 EISNER::BURNET "John Burnet" 41 lines 23-JUL-1989 23:31 -< LN03R shadowing (?); LN03R vs. other Postscript printers >- -------------------------------------------------------------------------------- > I got a DEC Postscript printer (LN03R?) for desktop publishing and have > experienced a problem with the print shadowing on some of the bold > characters. We use LN03s heavily (ten or twelve printers in six offices for several years, but no LN03Rs so far) and have never run into the problem you describe. It sounds like something that might be attributable to an aging print engine. (It's hard to see how it could have to do with the printer's Postscript capability.) Did you buy the LN03 new, or had it seen a lot of wear before you got it? > Their next move is to replace the printer. Sounds reasonable, iff the cause of the problem is indeed a print engine that's seen better days. If the symptom disappears after the replacement is installed, and if you have a service contract on the printer so that you can depend on DEC to fix the problem in a similar manner in the future, then you should be all set. :-) > Our Field Service tech has said, off the record, that this is a problem > inherent in the LN03 printer and probably will not be resolved. Did you get any details? That's the first I (and, I'm sure, many others) have heard of such a limitation of the LN03 -- which would actually seem to be a limitation of the Ricoh engine on which the LN03 is based. Does it affect only Postscript printers? If so, how? (It wouldn't have much to do directly with the engine itself in that case.) If not, why haven't more people seen the same symptoms on their vanilla LN03s that you have seen on your LN03R? > I feel an argument coming up with DEC about getting a refund and using the > refund for a 3rd party printer such as Apple. Whatever the merits of that general approach might be, you might want to get some comparison data re. LN03R vs. Apple Laserwriter vs. other printers before you do that. The Apple printer will probably cost you more to maintain, and its write-black Canon engine is not nearly as good for Postscript as are the write-white Ricoh engines used in the LN03R, AST Camintonn Turbolaser/PS (?), TI Omnilaser, and others. (The Omnilaser uses a different, faster Ricoh engine, but it's still write-white.) Also, DEC won't let you use its Scriptprint symbiont with the Apple (or any other third-party printer). ================================================================================ Note 417.2 LN03 bleeding! 2 of 13 EISNER::HASSINGER "Bob Hassinger" 26 lines 24-JUL-1989 00:19 -< I don't think it is an inherent problem... >- -------------------------------------------------------------------------------- > Also, DEC won't let you use its >Scriptprint symbiont with the Apple (or any other third-party printer). I don't have to run non-DEC printers from the Scriptprinter symbiont so I have not been tracking this closely but it is my impression that the early issues have been relaxed to the point where it is quite possible to do this, and indeed, to even get the software without first buying a DEC printer as once seemed to be the case. As to the shadowing, I have seen a lot of different problems with the LN03 and the LN03R but nothing inherent, just ware and repair issues that a DEC service contract has taken care of. Sometimes it has taken persistence to get a satisfactory resolution, to be sure. I use the LN03R to produce camera ready copy from both the VAX and an Amiga running applications like desk top publishing and I am pretty picky about print quality. What paper are you using? Sometimes that can be a factor. Is there something else you have not told us about your use or your expectations that would help explain your DEC guy's claim that this is an inherent problem? Note that at least around here the people DEC sends out to fix LN03 family machines are not necessarily the ultimate technical reference. That is, I rarely take their word for something like this without further investigation. ================================================================================ Note 417.3 LN03 bleeding! 3 of 13 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 13 lines 24-JUL-1989 11:06 -< Time or toner >- -------------------------------------------------------------------------------- When I first got my Camintonn Turbo-Laser (same Ricoh engine as the LN03), it also had a light "dusting" around the characters which I was unhappy with. It went away after a few thousand pages and now is nonexistant. I get the impression that there are several different toner formulations from Ricoh, plus assorted third-party ones. My second batch of toner (from Camintonn) had a different part number than the first, but I don't know what the exact difference is. (Hmmm, perhaps that was when the "dusting" went away. I simply can't remember.) I would like to know more about toner. ================================================================================ Note 417.4 LN03 bleeding! 4 of 13 EISNER::KOLP 15 lines 24-JUL-1989 19:48 -< LN03R follow-up >- -------------------------------------------------------------------------------- In response for more info: The Postscript printer is brand new with less than 100 copies. DEC already replaced the entire guts and the problem still is there. The paper that I'm using is the paper supplied with the printer (I like to test the printers with DEC's paper). As far as the problem being inherent, we have several dozen LN03's and I've noticed the hazing around the bold characters produced by DECpage. I've never yelled about it before because camera ready output was not required, but now it is. I am getting an entirely new printer tomorrow, courtesy of DEC. I'll keep you posted on results. Thanks for the feedback!!! P.S. I'll check with FS about the problem clearing up after several thousand copies although I would hate to have to wait around for that time. ================================================================================ Note 417.5 LN03 bleeding! 5 of 13 EISNER::MAYHEW "Bill Mayhew" 17 lines 24-JUL-1989 21:29 -< I think I see it too... >- -------------------------------------------------------------------------------- I have long had a faint displeasure with the quality of the LN03 output (my wife, who does more camera-ready stuff with it, is more displeased than I). There seems to be what I'd describe as "stray toner" scattered around in general, and it does seem to be worse around bolded characters, though I wouldn't describe it as "shadowing" -- we may just be using different semantics. We went through several LN03s at one point to fix a different problem (5 of 'em within a few weeks, I think) and they all exhibited this. They all (being supplied by Field Service) had varying numbers of pages on the page-counter; none were brand-new. Upon discussing this with one of the local "LN03 experts" that was working on our other problem, I was told that this was an inherent limitation of the Ricoh engine, and that it would never produce output that looked as good as, say, an HP LaserJet. I didn't pursue this much and, like Bob Hassinger, am not sure how much stock to put in this answer (I suspect Bob and I may have the same servicefolk, in fact). ================================================================================ Note 417.6 LN03 bleeding! 6 of 13 EISNER::KOLP 10 lines 28-JUL-1989 21:21 -< LN03R R.I.P. >- -------------------------------------------------------------------------------- A follow-up for those interested: Field Service has punted. They claim that they are not able to 'fix' the printer to have it produce (what I think) camera ready documents. I got a replacement with 100,000 copies on it with new guts and still the same problem. Result: I'm getting a refund and having to go with a 3rd party printer from Interleaf. Is this typical for Field Service? I've never seen them so passive about something. ================================================================================ Note 417.7 LN03 bleeding! 7 of 13 EISNER::KENNEDY "Terry Kennedy" 12 lines 29-JUL-1989 06:44 -< Maybe the rep thought it was fine >- -------------------------------------------------------------------------------- > Is this typical for Field Service? I've never seen them so passive > about something. Well, I'm not sure it's your service rep's fault. If you report that your computer can't add any more, and the rep says "PRINT 2+2" and gets "3.14159289" back, you can both agree that something needs fixing. Where there is a subjective opinion involved, things get more sticky. Did you always have the same rep working on it or did you see several? Anyway, I hope you have a better experience with your new unit (and definietly let us know what you got and how it looks). You may want to "test drive" one in a store to see some actual output. ================================================================================ Note 417.8 LN03 bleeding! 8 of 13 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 9 lines 29-JUL-1989 11:52 -< Same engine here, no problems. >- -------------------------------------------------------------------------------- We have not seen the bleeding problem on our LN03's (not LN03R's). We have printed tens of thousands of pages on each of three printers, and a great deal of that is TeX output. From this experience, I strongly doubt that the problem is endemic to the print engine. If, otoh, there is an analog ("video") interface form the controller to the print engine, the problem might be endemic to DEC's postscrip controller -- it could be nothing more than ringing! ================================================================================ Note 417.9 LN03 bleeding! 9 of 13 EISNER::KOLP 12 lines 30-JUL-1989 21:51 -< Post-mortem LN03R >- -------------------------------------------------------------------------------- I guess I used the wrong terminology with my explanation of what I was seeing. It is a shadowing of characters especially noticed around the large boldface print. Not only have I noticed it on my ex-LN03R printer, but every one of our LN03's (25+) experience this problem. It is not easily seen using the regular built-in fonts, though...but something like DECpage's boldface headings really show it. True, it is a subjective opinion and if it didn't bother me, I'd be much better off. I've tolerated it because the LN03's are used for letters and inter-office stuff. The LN03R was to be used to produce brochures, newsletters, etc. and the quality was not sufficient. I have started FS working on some of the LN03's to reduce the shadowing and it does help somewhat but never totally eliminates it. ================================================================================ Note 417.10 LN03 bleeding! 10 of 13 EISNER::KILGALLEN "Larry Kilgallen" 1 line 30-JUL-1989 23:06 -< Try testing on a write-black printer >- -------------------------------------------------------------------------------- e.g., LaserWriter or other Canon engine based printer. ================================================================================ Note 417.11 LN03 bleeding! 11 of 13 EISNER::DOHERTY "Bob Doherty" 18 lines 31-JUL-1989 10:30 -< Standing back, waiting for brickbats... >- -------------------------------------------------------------------------------- If it's any consolation, we've seen the same thing with LN03 output as compared to Apple LaserWriter IINT(a write-black engine). We've generated TeX output both for the LN03 and for the LW II, and the LW II output looked distinctly better, especially around large bold letters, and for letters with very fine serifs. There is no doubt, however, that graphics with large balck areas look much better from the LN03 than from the LW IINT. It's just that most of what we want from laser printers is text, NOT graphics. For text, it seems to us that the write-black engines are superior. Note that the engine used in the LW II is the same one used in the HP LaserJet II. I have also seen output from the Qume LED printer which seems to combine the best of both worlds (it is also marketed by Jasmine). It does not use the same imaging technology as the laser printers. ================================================================================ Note 417.12 LN03 bleeding! 12 of 13 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 6 lines 1-AUG-1989 00:16 -< Time for pasture for LN03? >- -------------------------------------------------------------------------------- Get the feeling it's time for DEC to get a new laserprinter in the product line? The LN03 has gotta be at least 4 years old. Don't get me wrong.. we've got a mess of LN03s that are used a lot and seem to do a good job. But.. we also had a mess of VT100s that were used a lot and did a good job... ================================================================================ Note 417.13 LN03 bleeding! 13 of 13 EISNER::ALDRICH "Jeff Aldrich" 5 lines 21-AUG-1989 17:30 -< Yet another dissatified LN03R customer. >- -------------------------------------------------------------------------------- We've had the same problem with seemingly random toner "splashed" around the letters... Hmmm... We are generating TeX and have tried it on a LW+ and it looks great... I think I vote for a new model.  ================================================================================ Note 418.0 RK06/RK07 service tips 3 replies EISNER::KENNEDY "Terry Kennedy" 2 lines 23-JUL-1989 02:12 -------------------------------------------------------------------------------- This topic will be used to post and discuss service tips & techniques for the DEC RK06/RK07 series of disk drives. ================================================================================ Note 418.1 RK06/RK07 service tips 1 of 3 EISNER::KENNEDY "Terry Kennedy" 32 lines 23-JUL-1989 02:24 -< RK07 - identifying proper read/write board (54-11060-YA) >- -------------------------------------------------------------------------------- One item which can cause bizarre problems in an RK07 drive is the incorrect read/write board. This is the board that surrounds the head opening. For the RK06, the correct part is 54-11060 (no suffix). For the RK07, the correct part is 54-11060-YA. Simple enough, right? Unfortunately, the -YA suffix is frequently not marked on the etch, but is instead rubber-stamped in a random location on the board. A sure-fire way to determine the board type is to examine closely the components directy above the U-shaped cutout for the heads. You should see the following: o-====-o <- resistor (no suffix) or jumper (YA suffix) o--XX--o <- ferrite bead o-====-o <- resistor o-====-o <- resistor o--XX--o <- ferrite bead o <- test point An RK07 with a non-YA board will appear to work, but will be very sensitive to media defects. A pack which works flawlessly in one drive might show 20 or 30 errors in the drive with the wrong board. Also, if the spindle is worn, these errors may appear as random recoverable errors (but always at the same places on the pack). Replacing the spindle with one with less non-repetitive runout will convert these errors into *hard* err- ors, as you now have less 'slop' on track center. Thus, you will stay over the low-signal area. ================================================================================ Note 418.2 RK06/RK07 service tips 2 of 3 EISNER::KENNEDY "Terry Kennedy" 9 lines 23-JUL-1989 02:30 -< RK06/RK07 - emergency head retract failure >- -------------------------------------------------------------------------------- The RK06 and RK07 use batteries to perform the emergency head retract function. If the battery pack has failed or weakened, the heads will not retract fully in a power-fail situation. Older RK06/7 drives have the batteries tied down underneath the card cage. Newer drives have the battery in a clip holder under the center col- umn of the drive. See note 267.2 in this conference for a source for these batteries. ================================================================================ Note 418.3 RK06/RK07 service tips 3 of 3 EISNER::PROVOST 11 lines 4-AUG-1989 11:22 -< MORE RK07 WOES >- -------------------------------------------------------------------------------- Card cage must be lifted to perform drive alignment. Card cage is hard to push squarely back into place. When card cage is left ajar, backplane pins scratch on back panel under vibration. This results in blinking fault lights during heavy system usage. We tape our back panel just in case. On older RK07's the boards are held in place by flexible metal straps. These can short out the 5V on the boards. We have also taped them. Newer RK07's have used a plastic mounting. Tom  ================================================================================ Note 419.0 MTI MA24 experiences 7 replies EISNER::NORTON "Bill Norton" 36 lines 26-JUL-1989 10:24 -------------------------------------------------------------------------------- Now that the MTI MA24 has been installed, here are some of my *personal* impressions. o Installation was easy, just connect cables and power. o The front panel just sits there showing "Unit 2", rather than anything informative. And the status lights don't tell you much either. There seems to be no indication when it's rewinding or searching. o The init/mount sequence is *incredibly* slow, and several times during the operation all other activity on the same HSC tape controller is stalled for some reason. There *may* be a problem between the MA24 and the HSC70. Last night the HSC complained about Microdiagnostic Completion Overdue on the tape requestor and rebooted. This was sometime after a batch job had tried to access the MA24 to do a backup. After the reboot, the HSC complained a lot: Cycles of "Went Available without request" followed by "Not asserting Available when it should be" repeated every 5 seconds for 15 minutes, eventually followed by: "Tape formatter connected to Requestor 2, Port 2. Has been declared inoperative intervention required" This morning an attempt to access the MA24 produced "Not asserting Available when it should be", at the same time as the front panel showed no circuit established, and the drive was ready. SHO TAPE on the HSC didn't show the MA24's units until the MA24's port select button was cycled off and on. So it looks like coming online after a HSC reboot isn't automatic, at least. It's a good thing we didn't disconnect the TA78 yet. MTI support has been phoned - results posted as received. ================================================================================ Note 419.1 MTI MA24 experiences 1 of 7 EISNER::SHANNON "Terry C. Shannon, erstwhile Matco" 7 lines 2-AUG-1989 21:02 -< Service with a Smile? >- -------------------------------------------------------------------------------- Do you have any comments on MTI's service and support? One of my cohorts is considering the acquisition of either the MTI 8mm or a forthcoming competitive product from SI. If all other things equal, service and support may be the deciding factor. ================================================================================ Note 419.2 MTI MA24 experiences 2 of 7 EISNER::NORTON "Bill Norton" 17 lines 3-AUG-1989 09:43 -< Service with wings >- -------------------------------------------------------------------------------- So far I've seen the serviceman twice. Once to install it, and once to look at it after the HSC problem in .0. Guy was competent. I get the impression there's two servicemen east of the Mississippi, one in WVa and one in Detroit. If this seems a little light it may be that I only heard about two, or only heard about two that install MA24's, or I'm just all wet. I don't know how they can make any money when they fly the guy in for each service call, on a $25K unit. When we talked to SI they seemed adamant about performing their own service. MTI seemed receptive to the idea having DEC service the unit if we wanted it that way. This flexibility was one factor that led us to select MTI. MTI also seemed to be nearer to actually delivering product. ================================================================================ Note 419.3 MTI MA24 experiences 3 of 7 EISNER::NORTON "Bill Norton" 11 lines 3-AUG-1989 09:51 -< Status update >- -------------------------------------------------------------------------------- Re the problem in .0: DEC says there's a V394 of HSC software that fixes HSC crashes, having something to do with a TA79. Since the MA24 identifies itself to the HSC as a TA79 this may help. Working on getting the new software. The issue of requiring manual intervention to come online to the HSC after an HSC reboot may cause us to return the unit, since the whole point of getting it was *unattended* operation. ================================================================================ Note 419.4 MTI MA24 experiences 4 of 7 EISNER::AZZOLI 10 lines 3-AUG-1989 18:47 -< MA224 and HSC V394 >- -------------------------------------------------------------------------------- > DEC says there's a V394 of HSC software that fixes HSC crashes, > having something to do with a TA79. I am running V394 with a recently installed MA224 and have not seen the problem with the drives coming online to the HSC. Although I did have to load a tape in each drive for my cluster nodes to CONFIGURE the drives. ================================================================================ Note 419.5 MTI MA24 experiences 5 of 7 EISNER::CHADWICK "Keith Chadwick" 38 lines 7-AUG-1989 12:47 -< Real TA79s confuse the issue >- -------------------------------------------------------------------------------- We just installed two MTI HSC 8mm tape drives on one of our clusters, and have noted the following problem: The MTI drives identify themselves to the HSC and VMS as being TA79 tape drives. Unfortunately, we also have twelve real TA79 tape drives. This will present problems when these drives are released to the 'general user community' (for the moment they are ACL'ed to SYSTEM), since the commands: ALLOCATE MU ALLOCATE/GENERIC=TA79 MOUNT MU Have a good chance of giving the user the wrong type of tape drive. So far we have come up with the following ideas: 1. Install them with large device numbers ($1$MUA100 & $1$MUA101 currently). For the reasons listed above, this is not a robust solution. 2. Change the controller from "A" to "B" ($1$MUB100 & $1$MUB101). We could then (attempt) to educate our users to either specify MUA or MUB depending on the type of tape drive that they desired. Unfortunately our tech's cannot find any jumper on the K.sti card to do this. 3. Change the allocation class of the HSC. Unfortunately we cannot change the allocation class for these tape drives, since the HSC to which they are connected provides service to half of our shadowed system disk, and the other HSC has four of our twelve TA79's. One could entertain purchasing an HSC40 to serve as a dedicated I/O controller for the MTI 8mm tape drives, but needless to say, this is viewed as a rather expensive workaround. Does anyone have any good ideas? -Keith. ================================================================================ Note 419.6 MTI MA24 experiences 6 of 7 EISNER::KILGALLEN "Larry Kilgallen" 1 line 7-AUG-1989 14:01 -< Disk and tape allocation classes are different >- -------------------------------------------------------------------------------- ================================================================================ Note 419.7 MTI MA24 experiences 7 of 7 EISNER::NORTON "Bill Norton" 8 lines 28-AUG-1989 09:57 -< New firmware from MTI >- -------------------------------------------------------------------------------- MTI has installed new firmware in the Exabyte transports and in their controller. Now the front panel can display things other vendors are trumpeting in their ads. It can now display error rate, tape remaining, and current operation for each drive. You have to manually cycle through the displays, which makes it less convenient than some brands that display it all at once, but it's a big improvement for just a ROM change.  ================================================================================ Note 420.0 microvax 2000/vaxstation 2000 memory from 3rd party 2 replies EISNER::BRUCKER "Cyberspace Jockey" 10 lines 26-JUL-1989 15:03 -------------------------------------------------------------------------------- Does anyone have any experience with 3rd party memory for the VAXstation 2000 platform? I have already seen the note in MICROVAX_SYSTEMS 70.* and the only reference was to an 8MB board available from Clearpoint. Are there any other vendors that have a memory product for the vs2000 and do you have any positive or negative feedback about the product or the vendor? Thanks in advance... Ken ================================================================================ Note 420.1 microvax 2000/vaxstation 2000 memory from 3rd party 1 of 2 EISNER::CHADWICK "Keith Chadwick, Fermilab" 21 lines 27-JUL-1989 14:32 -< Nemonix also sells VS2000 memory >- -------------------------------------------------------------------------------- > Does anyone have any experience with 3rd party memory for the > VAXstation 2000 platform? I have already seen the note in > MICROVAX_SYSTEMS 70.* and the only reference was to an 8MB board > available from Clearpoint. Are there any other vendors that have > a memory product for the vs2000 and do you have any positive or > negative feedback about the product or the vendor? Clearpoint also offers a 16MB board. Nemonix also offers 8MB and 16MB boards. I have had two of the Clearpoint 8MB boards installed in two color VS2000s for about a year without any problems. I have also had two of the Nemonix 8MB boards installed in two (different) color VS2000s for about six months, without any problems. The only tricky thing that I had to watch for was the ROM version of the VS2000. The original VS2000 ROM only supported a maximum of 6 MB, later versions of the ROM support up to 16MB. A phone call to the vendors technicial support should help you resolve this issue. -Keith. ================================================================================ Note 420.2 microvax 2000/vaxstation 2000 memory from 3rd party 2 of 2 EISNER::BERK 20 lines 23-AUG-1989 15:06 -< DATARAM also makes 2000 memory >- -------------------------------------------------------------------------------- > Does anyone have any experience with 3rd party memory for the > VAXstation 2000 platform? I have already seen the note in > MICROVAX_SYSTEMS 70.* and the only reference was to an 8MB board > available from Clearpoint. Are there any other vendors that have > a memory product for the vs2000 and do you have any positive or > negative feedback about the product or the vendor? We've got two MicroVAX 2000s with 16 MB of Clearpoint memory. It works perfectly. On one of them I did have to install some sort of floating daughter board, but it was included and easily done. I've recently come across the fact that DataRAM now also makes 2000 memory. One thing to watch out for is that you lose the 2MB on the motherboard if you install a 16MB daughter board (which means that you pay for 16MB to add 14MB). Nemonix and DataRAM have addressed this by supplying a 12MB daughter board. Also, there seems to be nothing to do with the daughter board being replaced; 4MB down the drain.  ================================================================================ Note 421.0 DS200<-->STRATUS hookup info wanted No replies EISNER::NORTON "Bill Norton" 12 lines 26-JUL-1989 17:18 -------------------------------------------------------------------------------- We've been blessed by the addition of a STRATUS system in IBM System/88 guise. My initial investigation shows the hardware(?) seems to want pins 5(CTS), 6(DSR), and 8(DCD) all high before it will talk to a terminal. Can anyone tell me a way to configure the s/88 ports so that we can use our 3-wire null-modem cables between it and a DS200 for reverse-LAT access by our VT220's? As for access by VT220's, does the s/88 editor work with them?  ================================================================================ Note 422.0 MicroVAX Enclosure Part Numbers 6 replies EISNER::ROBERTS_D "don roberts #241" 7 lines 27-JUL-1989 10:54 -------------------------------------------------------------------------------- Can someone tell me how to identify the various types of enclosures used for uVAX IIs? The documentation of what we have here and in Detroit is close to non-existant, and when we want to consider products that specify type of enclosure, we don't know what to say! For instance, here we have what looks to me like the "standard" uVAX box and an expansion box that says microPDP 11/73 on it. Which enclosures do we have? ================================================================================ Note 422.1 MicroVAX Enclosure Part Numbers 1 of 6 EISNER::MAYHEW "Bill Mayhew" 11 lines 27-JUL-1989 11:30 -< BA23s and BA123s >- -------------------------------------------------------------------------------- There are two basic enclosures, the BA23 (which is about 5"x19" x24" deep, and can either stand on one of the 5"x24" surfaces in a floor-mounted "tower" plastic enclosure, or can be rack-mounted horizontally), and the BA123, which is about the size of a two-drawer file cabinet and is not rack-mountable. The BA23 is also configured into rack-mount cabinets for larger MicroVAX IIs in which case the overall cabinet is an H9642, but the least common denominator is the BA23 or BA123. Also, the BA123 is on wheels, the BA23 isn't. I believe all 11/73s, or certainly most, were configured into BA23s. ================================================================================ Note 422.2 MicroVAX Enclosure Part Numbers 2 of 6 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 68 lines 27-JUL-1989 11:55 -< Here are the numbers and stats >- -------------------------------------------------------------------------------- < Note 422.0 by EISNER::ROBERTS_D "don roberts #241" > -< Can't tell the players without a number >- >> Can someone tell me how to identify the various types of >> enclosures used for uVAX IIs? BA23 Available in a couple different versions, depending on whether it is the main cabinet, expansion cabinet, disk expansion, etc. It is five inches high (wide) and contains an 8-slot backplane (3 Q22/CD, 5 Q22/Q22) and space for two full-height 5.25" disk/tape drives. Available in rack-mount, table-top, and pedestal configurations. It is sometimes seen in a double-high rack-mount configuration called a H9642-J. It has been used for both the MicroVAX and MicroPDP-11. A real b*t*h to work on, but great for small configurations. BA123 A caster-mounted office enclosure 13" wide, 24" high, and 28" deep. Disk/tape access is on the narrow (13") end, with write-protect switches, etc. behind a door at the lower left. Cable access is through a door at the opposite end of the cabinet. Board access is by removing the right-hand side panel (pop-off) and opening an inside cover. The backplane has four Q22/CD and 8 Q22/Q22 slots. There are five spaces for disk/drives. It has been used for both MicroVAX and MicroPDP-11 systems. In my opinion, it is one of the nicest products Digital has ever made, but cannot be rack mounted. BA213 Also known as the "skunk box", it is fully shielded and available in caster, rack-mount, and pole-mount versions. It is 21" wide, 27" tall, and 18" deep. All access is from the front (21" side). It has a 12-slot backplane, but all of them are Q22/CD, precluding their use with two dual-height cards. The backplane slots are on 0.95" centers instead of 0.5" and have special shielded covers/handles. There is space for up to three disk/tape drives above the backplane. This relatively-new enclosure has sevaral variations which I am not familiar with. As with all the others, it is used for both MicroVAX and MicroPDP-11 systems. BA214 This is a small (13" wide) version of the BA213 with a six-slot backplane and no space for disk/tape drives. It is intended for "embedded" applications to be booted via Ethernet. It is shorter than the BA213 because of no space for disk/tape drives. BA215 Essentially a BA214, but with space for two full-height 5.25" disk/tape drives above the backplane. Like the BA214, the BA215 is 13" wide, but the BA215 is taller, being the same height as the BA213 (27"). Since any of these enclosures can be used for either MicroVAX or MicroPDP-11 systems, it is not surprising that you don't have the right badges on your enclosures. Badges you might see include: MicroVAX II, MicroPDP-11, VAXstation II, VAXstation II/RC, VAXstation II/GPX, etc. I am trying to collect a complete set of them, so feel free to send me any you don't want. Naturally, DEC makes them in different sizes for the different boxes, just to frustrate the collector. :-) ================================================================================ Note 422.3 MicroVAX Enclosure Part Numbers 3 of 6 EISNER::MAYHEW "Bill Mayhew" 3 lines 27-JUL-1989 13:48 -< Very thorough answer, Alan, but... >- -------------------------------------------------------------------------------- I've never seen a MicroVAX *II*, or any MicroPDP other than an 11/83, in a BA200-series box. Given the module-handle issues, do they really exist? ================================================================================ Note 422.4 MicroVAX Enclosure Part Numbers 4 of 6 EISNER::KILGALLEN "Larry Kilgallen" 12 lines 27-JUL-1989 14:00 -< Wall-mount BA213? >- -------------------------------------------------------------------------------- > Note 422.2 > EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" > BA213 > > Also known as the "skunk box", it is fully shielded and > available in caster, rack-mount, and pole-mount versions. I did not know about pole-mount, but I thought wall-mount was possible. Does one have to order a special model or special accessory to get wall-mounting, or does it come automatically with normal BA213s? (I have been told that the new Tempest BA213 will *not* wall-mount.) ================================================================================ Note 422.5 MicroVAX Enclosure Part Numbers 5 of 6 EISNER::HARVEY "Jack Harvey" 16 lines 27-JUL-1989 19:16 -< BA213 has Rackmountable Insides >- -------------------------------------------------------------------------------- > Does one have to order a special model or special accessory to get > wall-mounting, or does it come automatically with normal BA213s? When you pop the plastic front panel off a BA213 and peer around a bit, you discover that what's inside is a deep aluminum "dish" with flanges for normal 19" rack mounting. The flanges are bolted to the rest of the plastic cabinet, which is playing the rack. My understanding of the "wall-mount option" is that there is a standard electrical equipment wall mount enclosure that will house a 19" rack mountable object of these dimensions. As for the pole-mount option, I presume it also has brackets for chimney-mount :-) ================================================================================ Note 422.6 MicroVAX Enclosure Part Numbers 6 of 6 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 13 lines 29-JUL-1989 00:09 -< Not a Polish joke >- -------------------------------------------------------------------------------- >> As for the pole-mount option, I presume it also has brackets >> for chimney-mount :-) While I have not seen the pole-mount up close, I believe it is the same sort of mount used in large factories to mount electronic "stuff" up off the floor and out of the way. It may be that DEC doesn't sell this mount, but expects you to buy it from your "usual source". >> I believe you are right; the -II has not been sold in the BA2xx. (Now watch someone make a lier out of me. :-) )  ================================================================================ Note 423.0 m-VAX failure ?24 UNIXT 4 replies EISNER::HAHN "DECUServe Vice Chair" 23 lines 27-JUL-1989 19:25 -------------------------------------------------------------------------------- Our microvax II rebooted itself twice today with **** FATAL BUG CHECK, VERSION = V4.5 MACHINECHK, Machine check while in kernel mode CURRENT PROCESS - NULL followed with values for R0..PSL and KERNEL/INTERUPT STACK I went through a SHUTDOWN because I thought this may have been a temperature problem (~80F) in the room. When the temp was back down below 75 I hit the start button; and The micro-vax II, upon reboot it went through the 7..6..5..4..3 then came up with ?24 UNIXT, what does that mean, and where do I find these ?error? messages? On second restart it went through everything and rebooted. Can some one explain what I should look for ? ================================================================================ Note 423.1 m-VAX failure ?24 UNIXT 1 of 4 EISNER::RICE "Gary Rice" 7 lines 28-JUL-1989 10:27 -------------------------------------------------------------------------------- Chances are 95% that you are having a hardware failure. Most likely, a bad memory board. Check your error log for info surrounding the time that the crash occured. Gary ================================================================================ Note 423.2 m-VAX failure ?24 UNIXT 2 of 4 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 9 lines 29-JUL-1989 00:30 -< Not much help >- -------------------------------------------------------------------------------- >> The micro-vax II, upon reboot it went through the >> 7..6..5..4..3 then came up with ?24 UNIXT, what does that mean, >> and where do I find these ?error? messages? I think the message was really ?24 UNXINT, which indicates that the CPU got an unexpected interrupt or exception during the diagnostics. Test 3 (which was running) is the interrupt test. Unfortunately, there is no documentation that gives any guidance in what to look for. ================================================================================ Note 423.3 m-VAX failure ?24 UNIXT 3 of 4 EISNER::GRATTAN "Michael Grattan (Who #316)" 8 lines 29-JUL-1989 12:27 -< It's in the manual. >- -------------------------------------------------------------------------------- > I think the message was really ?24 UNXINT, which indicates that the CPU > got an unexpected interrupt or exception during the diagnostics. Test > 3 (which was running) is the interrupt test. Unfortunately, there is no > documentation that gives any guidance in what to look for. Your're right on. The only place I have ever found this stuff is in the MicroVAX II 630QB Technical Manual (order #AZ-FE09A-TN), and there it just tells you what it is, not what to do about it. ================================================================================ Note 423.4 m-VAX failure ?24 UNIXT 4 of 4 EISNER::SAUNDERS "John Saunders, Pacer Software" 2 lines 9-AUG-1989 06:37 -< Unexpected CPU Death? >- -------------------------------------------------------------------------------- We got this several times from the MVII board in our LPS40 after power hits. FS just replaced the board...  ================================================================================ Note 424.0 Symplex Datamizer: sync mux/compressor/expander 1 reply EISNER::BURNET "John Burnet" 12 lines 27-JUL-1989 22:50 -------------------------------------------------------------------------------- > <<< EISNER::DUA0:[NOTES$LIBRARY]DEC_NETWORKING.NOTE;1 >>> > Note 287.13 Talking to Detroit: DECnet, bridging, statmuxing or what? 13 of 17 > >> If you want to still use one leased line, then check out the "Datamizer" >> line from Symplex. Supposedly it is a mux optimized for passing sync data. > > A great idea! But it isn't just a mux. It is a smart-ass compressor, and DDCMP > with normal Decnet traffic is quite compressible. These can easily run 4 > 9.6 Decnet sync streams over a 9.6 link! Think of them as SYNC STATMUXES. Sounds too good to pass up! Can anyone supply an address and/or phone number for Symplex? Thanks. ================================================================================ Note 424.1 Symplex Datamizer: sync mux/compressor/expander 1 of 1 EISNER::PRIGOT "Jonathan M. Prigot" 6 lines 31-JUL-1989 16:24 -< Symplex address >- -------------------------------------------------------------------------------- Symplex Communications Corp. 5 Research Drive Ann Arbor, MI 48103 (313)995-1555 Note: I have no connection with Symplex. I don't even own one.  ================================================================================ Note 425.0 Problem replacing DHT32 with DSH32 on MV2000 3 replies EISNER::LEDERMAN "Bart Z. Lederman" 20 lines 31-JUL-1989 08:26 -------------------------------------------------------------------------------- We have a remote site which is currently equipped with a MV2000 with the following: an extra memory board (I think Cadminton), an extra disk drive (Trimarchi), and the DEC DHT32-AA 8 asynchronous line interface. We need to add P.S.I. to this machine so it will network to our other machines, so after consulting with DEC we bought the new DSH32-A board, which has on it 8 asynchronous and 1 synchronous line. The idea is to replace the DHT32 with the DSH32. DEC field service came in the other day and said there is no room for it: they can't fit in the RS-232 connector or something. Does anyone have any first-hand experience with the new DSH32 board and the MV2000? Will it fit in where the DHT board used to be? If not, what options do we have? The only DEC literature I've seen implies that both boards are the same size, and DEC told us it would work, but now we're not sure. Unfortunately this site is in an area of the country where it is several day's travel for one of our 'experts' to get out and see for ourselves; so we need the advice from someone who has actually seen one of these combinations. ================================================================================ Note 425.1 Problem replacing DHT32 with DSH32 on MV2000 1 of 3 EISNER::MAYHEW "Bill Mayhew" 18 lines 31-JUL-1989 12:33 -< You need a BA40A-BA >- -------------------------------------------------------------------------------- I have not actually seen one, *but* the information accompanying the announcement of the DSH32 stated that it is NOT compatible with the BA40A-AA expansion adapter (the 1" high metal box that sits on the bottom of an MV2000 -- it ships as a standard part of MV2000s (as of a few months after announcement -- original MV2000s did not have it as standard) and is available as an option on the VS2000, and is required to support the TK50 or external disk or DHT32). Instead the DSH32 requires the BA40A-BA, which is alleged to be upwards compatible with the -AA you probably have. The good news is that the -BA sells for $600 list, less than half the price of the -AA. The better news is that I might be interested in buying your -AA if the price is right. (Send MAIL.) The bad news is that whoever sold you the DSH32 didn't bother to read the announcement text for the product, or the Jul-Sep '89 Systems and Options Catalogue, page 3.39, right above the subsection labelled "Extended Cables" -- not quite where it belongs, but at least on the right page. ================================================================================ Note 425.2 Problem replacing DHT32 with DSH32 on MV2000 2 of 3 EISNER::LEDERMAN "Bart Z. Lederman" 11 lines 31-JUL-1989 13:30 -< Thanks, our people found another solution. >- -------------------------------------------------------------------------------- I got some feedback from the site. The field service person apparently knew about the BA40A-BA being needed (that's the number they gave us), and obviously the DEC office which sold it to us didn't. But the site needed the unit installed now. I don't know all of the details, but I'm told they just cut a hole in the side of the MV2000 and used some old boxes that came with the Trimarchi drive, and field service installed it. Since this is a government site, selling surplus costs more than it's worth (at least in this price range). Sorry. ================================================================================ Note 425.3 Problem replacing DHT32 with DSH32 on MV2000 3 of 3 EISNER::LEDERMAN "Bart Z. Lederman" 17 lines 8-AUG-1989 13:05 -< DEC says: Don't Install DSH32-AA in MicroVAX 2000 >- -------------------------------------------------------------------------------- Our DEC Sales Representative called us today with a warning not to install a DSH32-AA in a MicroVAX 2000. He says this information come directly from manufacturing, and that Field Service and Atlanta Support Center don't know about it yet. (They might by now, we called Atlanta and had them talk to our Rep.). He says engineering knows about it and that DEC is looking for solutions. The problem may be MV2000 specific, and would not affect the 3100 series. I would view this information with caution. But if you have ordered a DSH32 for a MV2000 I would talk with your representative, and have him/her check up the line for information. And have Field Service check also, before doing the installation. I do know that the system we already installed the DSH32 in crashes when we try to use the synchronous port.  ================================================================================ Note 426.0 STRATUS systems in DEC shops? 1 reply EISNER::NORTON "Bill Norton" 11 lines 31-JUL-1989 17:19 -------------------------------------------------------------------------------- Does anyone out there have a STRATUS or IBM SYSTEM/88? I'm looking for a terminal definition file that will allow us to use our VT220's as standard STRATUS terminals. I think this is something equivalent to a termcap file that specifies what the terminal uses for control keys, etc. I'm also looking for Kermit for STRATUS. Any pointers would be greatly appreciated. ================================================================================ Note 426.1 STRATUS systems in DEC shops? 1 of 1 EISNER::DOHERTY "Bob Doherty" 13 lines 1-AUG-1989 09:44 -< Try GNU Emacs distribution... >- -------------------------------------------------------------------------------- > I'm looking for a terminal definition file that will allow us to > use our VT220's as standard STRATUS terminals. GNU Emacs has many termcap entries, among which, are some for VT2xx terminals. The latest GNU emacs distribution is on the combined L&T and VAXSIG 1989A tape. It's in one of the VAX89An save sets as a compressed tar archive, and after uncompressing and untarring the termcap stuff is in the [.etc] directory. A bonus is a termcap manual which will tell you twice as much as you ever wanted to know about termcap.  ================================================================================ Note 427.0 HSC5X-DA - How Fast 5 replies EISNER::JENSEN_M 12 lines 2-AUG-1989 12:00 -------------------------------------------------------------------------------- Anyone (Ken?) out there know if the new HSC5X-DA controller boards are limited to the same max data rate as their predecessor - 22.5(approx) Mbit ? DEC is quoting Q290 delivery (OCT-DEC 89) but distributors are citing 30 days or less. Anybody got any of these boards yet ? Thanks in advance Mike ================================================================================ Note 427.1 HSC5X-DA - How Fast 1 of 5 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 21 lines 2-AUG-1989 14:29 -< HSC5X-DA mind dump... >- -------------------------------------------------------------------------------- >< Note 427.0 by EISNER::JENSEN_M > > -< HSC5X-DA - How Fast >- > >DEC is quoting Q290 delivery (OCT-DEC 89) but distributors are citing 30 >days or less. Anybody got any of these boards yet ? We got two with our TA90 (which requires them). Perhaps the delivery time DEC is quoting is for orders that don't specifically require the HSC5X-DA. As far as speed is concerned, it's hard to say. From the little I've read on the HSC5X-DA, it is essentially like the older dedicated disk and tape cards, the main difference being that instead of having the microcode for the card burned into a prom, it's kept on the HSC's load device, and loaded by the HSC on reboot. (Side thought - what would load one of these guys connected to a UDA? Anybody know?) There is an HSC command that allows you to specify which flavor of microcode you want to load (disk or tape). After the microcode is loaded, even the HSC can't distinguish an HSC5X-DA from its older bretheren... Ed ================================================================================ Note 427.2 HSC5X-DA - How Fast 2 of 5 EISNER::FULLER_S "Funky cold medina" 4 lines 2-AUG-1989 16:13 -------------------------------------------------------------------------------- >device, and loaded by the HSC on reboot. (Side thought - what would load >one of these guys connected to a UDA? Anybody know?) There is an HSC You don't put one of these guys on a UDA. ================================================================================ Note 427.3 HSC5X-DA - How Fast 3 of 5 EISNER::BATES "Ken Bates" 9 lines 3-AUG-1989 20:39 -< Same speed... >- -------------------------------------------------------------------------------- re: .0 The maximum data rate of the SDI is 22.5 MBits per second, so all the k.xx has to do is attain that rate. The new k.si boards are slightly faster in sending the SDI level 2 commands (Seek, Online, etc.) to the disk/tapes than the k.sxi boards, but you should only see the increase in speeds when running carefully designed benchmarks. - Ken ================================================================================ Note 427.4 HSC5X-DA - How Fast 4 of 5 EISNER::JENSEN_M 10 lines 3-AUG-1989 22:12 -< Thanks for the info >- -------------------------------------------------------------------------------- > -< Same speed... >- Thanks for the info. The interest was in two areas - 1) higher speed board might permit DEC to design/release higher data rate drives, 2) potential usability of 3rd party party 24 Mbit (3 Mbyte) drives. However, after recently listening to your (Ken) Atlanta talk on Disk subsystem performance, data transfer rate may not be all that important. The session was very informative, even without visual aids (cassette). Mike ================================================================================ Note 427.5 HSC5X-DA - How Fast 5 of 5 EISNER::ALDRICH "Jeff Aldrich" 5 lines 21-AUG-1989 17:19 -< We got 'em. >- -------------------------------------------------------------------------------- We did manage to get three of these cards from DEC when we ordered out HSC40. We did have to speak firmly to them when they first told us that they wanted to send us the others. But my boss was extremely clear that we would only accept -DA's and they complied.  ================================================================================ Note 428.0 MicroVAX 2000 ROM Tests 1 reply EISNER::KRULEWICZ 8 lines 2-AUG-1989 12:58 -------------------------------------------------------------------------------- Does anyone know how to access MicroVAX 2000 ROM based tests 60, 72, 73, 101, 102, and others? My machine sez: ?17 ILL CMD >>> ================================================================================ Note 428.1 MicroVAX 2000 ROM Tests 1 of 1 EISNER::JARZABEK 0 lines 7-AUG-1989 21:12 -< Just type T xx at the >>> >- --------------------------------------------------------------------------------  ================================================================================ Note 429.0 MicroVAX 3400 w/out DSSI Drives 1 reply EISNER::SKANRON "Closet DEChead" 3 lines 8-AUG-1989 10:14 -------------------------------------------------------------------------------- Can a MicroVAX 3400 bootstrap off a non-DSSI disk drive (i.e. CDC WREN VII with a QD33), or is it required to boot off of the DSSI??? ================================================================================ Note 429.1 MicroVAX 3400 w/out DSSI Drives 1 of 1 EISNER::MCCORMICK_B "Virtual Bob" 5 lines 12-AUG-1989 16:41 -< Should work FB >- -------------------------------------------------------------------------------- I've played with a bunch of KA640's (3300/3400) and I don't see anything that would stop it from booting off of something as you ask... as long as its a device that can be recognised by the boot code. At the >>> prompt type HELP, and set boot to the correct device name.  ================================================================================ Note 430.0 TELENET HARDWARE 1 reply EISNER::WRIGHT_S "Sim Wright" 8 lines 8-AUG-1989 12:20 -------------------------------------------------------------------------------- NETWORK HRDWR (TELENET) Need Telenet Hardware Immediately. Does anyone have any alternate sources other than TELENET. They are quoting 4 wks. for additional hardware. I need it tomorrow, will ship counter to counter if needed. Does anyone have any alternate sources to TELENET Hardware? ================================================================================ Note 430.1 TELENET HARDWARE 1 of 1 EISNER::STANGE 2 lines 14-AUG-1989 16:36 -< Need an old PAD? >- -------------------------------------------------------------------------------- What do you need? We have some old PAD box. Want me to dig it out? About 4 years old I'ld guess. Call me at (515)270-3587.  ================================================================================ Note 431.0 QBUS & UNIBUS spec handbooks needed 3 replies EISNER::IVLER "J.M. Ivler" 20 lines 9-AUG-1989 09:30 -------------------------------------------------------------------------------- Since a search of the DECUServe titles didn't help (and this conference is far too big for me to search note-by-note) I hope I am not being repetitive. My hardware team has searched, and not found, two "manuals" that they need. After coming to me on hands and knees (actually they came to the systems manager on hands and knees and he promised them the information... then he came and threatened to take away my color workstation and put me on a VT100) I agreed to post the request for the information here. There is a need for a "spec handbook" for both the QBUS and UNIBUS. If anyone knows where the hardware team can get copies of these handbooks I and my workstation would appreciate the help. Thanks. JMI ================================================================================ Note 431.1 QBUS & UNIBUS spec handbooks needed 1 of 3 EISNER::PROVOST 28 lines 9-AUG-1989 10:24 -< What are your needs? >- -------------------------------------------------------------------------------- re: < Note 431.0 by EISNER::IVLER "J.M. Ivler" > > -< QBUS & UNIBUS spec handbooks needed >- > There is a need for a "spec handbook" for both the QBUS and > UNIBUS. If anyone knows where the hardware team can get copies > of these handbooks I and my workstation would appreciate the > help. Could use more information on your needs... Best source is PDP11BUSHANDBOOK, EB 17525 20/79 070 14 55. Unfortunately it is out-of-print, and will probably stay so. Other good sources are also unavailable. You might try PDP-11 Architechture Handbook, EB-23657-18. It has technical LSI-11 bus specifications in appendix E and UNIBUS timing in appendix D. Can you give us a handle on what you need? PINS vs SIGNALS? TIMING? Instructions (e.g. DATOB)? Configuration rules (e.g. AC/DC Loading, device ordering)? Interface chips (e.g. 8881, 8838, 8640)? Are you building devices, or making them work? Tom ================================================================================ Note 431.2 QBUS & UNIBUS spec handbooks needed 2 of 3 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 16 lines 9-AUG-1989 11:46 -< Here are the bus specs >- -------------------------------------------------------------------------------- >> There is a need for a "spec handbook" for both the QBUS and >> UNIBUS. If anyone knows where the hardware team can get copies >> of these handbooks I and my workstation would appreciate the >> help. The Unibus section of the handbook Tom mentioned is a reprint of an old DEC manual "PDP-11 Unibus Design Description". It has no date, part number or (best of all) copyright notice. I would be happy to make a copy for you. Just send me MAIL with your address. The Q-Bus section is included as an appendix in every one of the microVAX CPU module technical manuals (EK-KA6xx-UG), and in many of the PDP-11 handbooks. You might also find useful information in the "Chipkit Users Manual" which comes with the DCK11-series of bus interface chips from DEC. ================================================================================ Note 431.3 QBUS & UNIBUS spec handbooks needed 3 of 3 EISNER::NAGY "Frank J. Nagy, VAX Guru" 3 lines 10-AUG-1989 07:26 -------------------------------------------------------------------------------- The QBus specifications can also be found in Appendix A of the VCB02 Video Subsystem Technical Manual which comes with the Owner's Guide/ Hardware Handbook shipped with a VAXStatio-II/GPX or VAXStation-3200.  ================================================================================ Note 432.0 Weird DEC coax ribbon cable 4 replies EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 15 lines 10-AUG-1989 12:17 -------------------------------------------------------------------------------- I have a 25' DEC cable used to connect the VT48 Display Processor to the VR48 Scope Assembly (all part of the VS60 Vector Display System). It consists of two flat (ribbon) cables, each consisting of nine coaxial cables (total of 18 coax cables). Can anyone tell me (1) where I can get about 50-100' of this cable, or (2) the impedance of the coax (so I can find it on my own). I simply want to extend my existing cable a bit. The DEC part numbers are 70-11746-25 for the complete assembly (with connectors) and 17-00001-01 for the raw cable. The cable has no markings of any kind on it. Needless to say, this is not in the price book. ================================================================================ Note 432.1 Weird DEC coax ribbon cable 1 of 4 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 5 lines 10-AUG-1989 17:03 -< Should be able to measure this... >- -------------------------------------------------------------------------------- If you have or can borrow a TDR (or take the cable to one) you should be able to measure the impedance with sufficient accuracy to allow selection of some standard cable for extensions (and you'll need the TDR anyway to assure that all your cut pieces are the same electrical length...). ================================================================================ Note 432.2 Weird DEC coax ribbon cable 2 of 4 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 17 lines 11-AUG-1989 13:34 -< Another piece of test equipment I don't have yet >- -------------------------------------------------------------------------------- > If you have or can borrow a TDR (or take the cable to one) Unfortunately, I don't have one and don't know anyone with one. I don't think that precise length matching is necessary for this application. I am more concerned about distortion of the analog X-Y deflection signals from ringing, reflections, etc. With a stroke-writing graphics display, these errors are very obvious (and annoying). The existing cable is 25' long and I want to add another ~ 25'. Since the DB-37 connector DEC uses is an obvious impedance "bump", is it a good or bad idea to have the connector in the middle of the combined cable, or offset to one side or the other? Obviously, I would be best off by making a totally new cable of the desired length (and if the cable is cheap enough, I will). However, I expect the price to be multiple dollars per foot for all 18 pairs. ================================================================================ Note 432.3 Weird DEC coax ribbon cable 3 of 4 EISNER::HARVEY "Jack Harvey" 42 lines 12-AUG-1989 18:41 -< Coax Cable Impedance >- -------------------------------------------------------------------------------- Once again, Alan, nature is on your side. The impedance of coaxial transmission line can usually be determined by physical measurement. This is because the impedance varies as the *log* of the ratio of the diameters of the inner and outer conductors. The dielectric constant of the material between the inner and outer conductors (insulation) is the other variable, but since the *square root* of the dielectric constant is the factor in the equation, it doesn't have much to say about the impedance either. This is why we don't have one ohm or 1000 ohm coax. Nature prefers that mere man stick close to 65 ohms. So, using a micrometer, measure the diameter of the inner conductor and the diameter of the dielectic which the shield braid rests on (establishing the inner diameter). Calculate their ratio. Then examine the dielectric itself. If it's a solid plastic it's probably polyethene or teflon with a dielectric constant of 2.26 or 2.1. (The square roots of these are 1.5 and 1.45, so you see it doesn't matter much.) If the dielectic is a foam, consider it to be air, with a dielectric constant of 1.0 and a square root of 1.0. So we are talking about 1, 1.45 and 1.5 for the dielectic factor. The bottom line is this table. It gives the ratio of the diameters of the inner conductor and the dielectric (insulation) between the inner conductor and the braid. Impedance Solid Foam ohms Dielectric Dielectric 50 3.5 2.3 75 6.5 3.5 90 9.5 4.5 Physical examination should do the job. Measure the diameters, calculate the ratio, and examine the dielectric. Then pick the table entry that seems to fit best. ================================================================================ Note 432.4 Weird DEC coax ribbon cable 4 of 4 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 10 lines 12-AUG-1989 23:49 -< Just after I measured it... >- -------------------------------------------------------------------------------- > Physical examination should do the job. Measure the diameters, > calculate the ratio, and examine the dielectric. Then pick the > table entry that seems to fit best. And 90 it is! I also got a call from a guy with a copy of the original engineering specification who confirmed this. Now I just have to find a source for the stuff. Shouldn't be too hard, he says confidently... :-) Thanks for the help.  ================================================================================ Note 433.0 Plessey 'DL11-W' manual needed 6 replies EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 3 lines 10-AUG-1989 12:21 -------------------------------------------------------------------------------- Does anyone have a manual for the Plessey "DL11-W" unibus terminal interface? I have a print set, but the function of some of the jumpers and switches is still a bit obscure. ================================================================================ Note 433.1 Plessey 'DL11-W' manual needed 1 of 6 EISNER::KENNEDY "Terry Kennedy" 7 lines 10-AUG-1989 18:45 -< Wil DEC DL11-W book do? >- -------------------------------------------------------------------------------- > Does anyone have a manual for the Plessey "DL11-W" unibus > terminal interface? I have a print set, but the function > of some of the jumpers and switches is still a bit obscure. Is it 'just like' a DEC DL11-W? I have a book for those. I also have an extra DEC DL11-W if you'd rather go w/ real DEC... ================================================================================ Note 433.2 Plessey 'DL11-W' manual needed 2 of 6 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 19 lines 11-AUG-1989 13:53 -< Need jumper and switch settings >- -------------------------------------------------------------------------------- > Is it 'just like' a DEC DL11-W? I have a book for those. I don't think so. > I also have an extra DEC DL11-W if you'd rather go w/ real > DEC... Since this is for a super-cheap home project, I am trying to spend zero money. So far the total cost is under $10 (but don't ask about time). Everything has come from somebody's junk pile. "If you don't come get it, it goes in the dumpster." I am trying to get an old (donated) VS60 running on an old (donated in pieces) 11/34 so I can play Lunar Lander. If I can't get the 11/34 running, I will see if I can get the pieces of an 11/60 (also donated) to fit in a BA11-K and work. If that fails, I will try an 11/23 through an Able Qniverter. (Yes, I do have a lot of "junk".) ================================================================================ Note 433.3 Plessey 'DL11-W' manual needed 3 of 6 EISNER::OSUDAR "John Osudar" 2 lines 11-AUG-1989 16:02 -< I may have it >- -------------------------------------------------------------------------------- I'll check our files -- we used to have a lot of old Plessey junk before we "decontaminated" most of our PDP-11's :-) ================================================================================ Note 433.4 Plessey 'DL11-W' manual needed 4 of 6 EISNER::KENNEDY "Terry Kennedy" 8 lines 11-AUG-1989 21:14 -< No, it's free >- -------------------------------------------------------------------------------- |> I also have an extra DEC DL11-W if you'd rather go w/ real |> DEC... > Since this is for a super-cheap home project, I am trying to > spend zero money. Nope, you misunderstood. I could either *give* you the manual or *give* you the DEC DL11-W... (or both, actually) ================================================================================ Note 433.5 Plessey 'DL11-W' manual needed 5 of 6 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 10 lines 12-AUG-1989 11:50 -< Wow, such a deal! >- -------------------------------------------------------------------------------- >> Nope, you misunderstood. I could either *give* you the manual >> or *give* you the DEC DL11-W... (or both, actually) How can I pass up an offer like that? I accept with many thanks. See my MAIL message for possible trade items you might want. Now, if I get the Plessey manual (and get it working), I will have something to pass on to the next person who needs hardware help. ================================================================================ Note 433.6 Plessey 'DL11-W' manual needed 6 of 6 EISNER::OSUDAR "John Osudar" 2 lines 14-AUG-1989 13:33 -< no luck >- -------------------------------------------------------------------------------- I struck out on the manual. Seems we threw away all of that junk a while ago... :-)  ================================================================================ Note 434.0 DECNET <--> 9600 Baud Modems 7 replies EISNER::KRULEWICZ 10 lines 11-AUG-1989 11:14 -------------------------------------------------------------------------------- Is anyone running DECNET over a 9600 Baud dialup modem line? We trying to select a modem model but no one seems to know if DECNET will like the 9600 baud, the error correction (if present), and the data compression (if present). Modem model numbers and sucess/problems would be greatly appreciated. Thanks. ================================================================================ Note 434.1 DECNET <--> 9600 Baud Modems 1 of 7 EISNER::HYDE "From the laboratory of Dr. Jekyll" 28 lines 11-AUG-1989 13:33 -< don't bother >- -------------------------------------------------------------------------------- Asynchronous DDCMP over a 9600 baud line is unsupported if the modems do any error correction, compression etc. You may be able to get the circuit up, but it will be unreliable, especially under load. To quote from the latest DECnet-VAX SPD: Communications Options DECnet-VAX uses ..(assorted stuff removed).. Asynchronous controllers use DDCMP, either when directly connected or connected via modems, to provide full duplex communications over point to point lines. (NOTE: Error correcting and data suppression modems are not supported. Refer to HARDWARE REQUIREMENTS for a short description of controllers.) (more stuff removed) DDCMP operation is not supported in cases where an asynchronous physical communication line is emulated by lower level protocols or communication subsystems. Examples of this include X.29 virtual terminals, asynchronous connections as emulated by terminal servers, and connections via data switches. (Me again) Error correcting, data suppression(compression) and independent flow control done between the modems tends to stomp all over the DDCMP protocol which does its own error correcting and flow control etc. ================================================================================ Note 434.2 DECNET <--> 9600 Baud Modems 2 of 7 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 27 lines 11-AUG-1989 19:39 -< use SYNC on digital links if possible >- -------------------------------------------------------------------------------- >Is anyone running DECNET over a 9600 Baud dialup modem line? Did you mean ASYNC? Most of the newer modems actually are SYNC underneath, and one can run them in that mode. No compression, no error correction (other than the trellis encoding), but this IS the best mode to use IF you have SYNC ports to use. If you are faced with buying a SYNC port, consider low end Ethernet remote bridges instead (assuming you have E-net) and link them with your sync modems. If you have any serious amount of traffic, and it is fairly long haul, also consider that switched 56kb now costs little more than a normal phone call, and you put your modem $s into DSU/CSU type hardware instead. Same is true for leased lines. Tarrifs are not all in place in all jusisdictions for everything you may want to do, but if you are doing any sort of even short term future planning, digital service is where you will want to be. The problem with the carriers is just the same as with DEC. Their sales people don't know what they can/should sell, and especially if it was just recently offered. And then there is ISDN... ================================================================================ Note 434.3 DECNET <--> 9600 Baud Modems 3 of 7 EISNER::DOHERTY "Bob Doherty" 7 lines 11-AUG-1989 19:51 -< V.32 modems should work... >- -------------------------------------------------------------------------------- You should be able to use V.32 modems at 9600 baud, since they don't do any data compression (if they're really running in V.32 mode) and ar really full duplex and not ping pong or such like. V.32 modems are becoming widely available at incresingly reasonable prices (<$1000 ea and falling). ================================================================================ Note 434.4 DECNET <--> 9600 Baud Modems 4 of 7 EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 3 lines 13-AUG-1989 18:04 -< Microcoms can do it >- -------------------------------------------------------------------------------- We run asynch DECnet through Microcom AX9624Cs with no problems. They are reliable. One must, however, make the modem as transparent as possible. We turn off any sort of flow control. ================================================================================ Note 434.5 DECNET <--> 9600 Baud Modems 5 of 7 EISNER::LEDERMAN "Bart Z. Lederman" 8 lines 14-AUG-1989 08:12 -< It's not very good, but it will work. >- -------------------------------------------------------------------------------- I ran an async link over 'modem type' connections: though it wasn't dial-up, it looked like one. I agree that you must not have any kind of flow control (with the possible exception of hardware RTS/CTS type signals), no compression, etc. Also, SET HOST terminal connections tend to be rather miserable over such links. But, it does work, and it can be a good way to transfer files, mail, etc., especially if you can't get something better. In our case, we didn't have synchronous interfaces available. ================================================================================ Note 434.6 DECNET <--> 9600 Baud Modems 6 of 7 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 6 lines 14-AUG-1989 11:35 -< ASYNC warning >- -------------------------------------------------------------------------------- > I agree that you must not have any kind of flow control (with the > possible exception of hardware RTS/CTS type signals), no compression, It has ben stated else wher but I'll repeat it. The **ASYNC** driver for DECnet does **NOT** support RTS/CTS. ================================================================================ Note 434.7 DECNET <--> 9600 Baud Modems 7 of 7 EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 3 lines 14-AUG-1989 12:06 -< No flow, but compression allowed >- -------------------------------------------------------------------------------- I can experientially verify that. The asynch DECnet links would not appear at all until I totally disabled all flow control. Compression, however, works just fine.  ================================================================================ Note 435.0 Equipment Cabinet Air Filters 4 replies EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 17 lines 13-AUG-1989 21:23 -------------------------------------------------------------------------------- This question started when the air filter on a BA11-K box disintegrated and shed particles all through the system. Is there a general solution to the problem of air filters and other foam plastic parts falling to pieces over the years? It seems to happen to every manufacturer, so I suspect that we will just have to live with it. Are there any plastics experts out there that can explain why some parts disintegrate and others turn into a sticky goo? Is it safe to replace the foam air filters with fiberglass air-conditioner type filters? I have seen stuff in a spray can that is used by motorcyclists (among others) to make the foam in their air filters more "sticky". Is it safe to use this stuff on computer air filters? ================================================================================ Note 435.1 Equipment Cabinet Air Filters 1 of 4 EISNER::KENNEDY "Terry Kennedy" 44 lines 14-AUG-1989 02:52 -< A brief answer; more information tomorrow >- -------------------------------------------------------------------------------- > Is there a general solution to the problem of air filters and > other foam plastic parts falling to pieces over the years? It > seems to happen to every manufacturer, so I suspect that we > will just have to live with it. If you have DEC equipment with *foam* filters, it is quite old (70's?). This is the stuff that's sort of greenish-brown. When they first started failing DEC switched over to a black material which is not really a foam. It's more like hardened spaghetti. If you look at it carefully, you will see that unlink a foam, which has round 'voids' in it, this has randomly sized/shaped openings. Of course, DEC is not entirely consistent - the cabinet doors on a 780 use the old foam, whereas the ones on RK07's use the new. Perhaps this was just a localized FS incident, though. Anyway, you should change foam to the new style if you can. > Are there any plastics experts out there that can explain > why some parts disintegrate and others turn into a sticky > goo? For the DEC ones, sticky is a step on the way to crumbly. It depends on the temperature and humidity of the environment. I am going to see a client tomorrow who is in the solvent business and I'll get more details. > Is it safe to replace the foam air filters with > fiberglass air-conditioner type filters? See above - I'd suggest the new-style DEC filters. I'll try to post the part number tomorrow night. > I have seen stuff in a spray can that is used by motorcyclists > (among others) to make the foam in their air filters more > "sticky". Is it safe to use this stuff on computer air > filters? You have a different application. You want to remove very coarse particles. The bikers don't want to allow even medium particles as they damage the engine. In the case of disk drives, there is already a fine filter inside, and for the other stuff, it really doesn't matter. ================================================================================ Note 435.2 Equipment Cabinet Air Filters 2 of 4 EISNER::LEDERMAN "Bart Z. Lederman" 13 lines 14-AUG-1989 08:19 -< O.K. by me. >- -------------------------------------------------------------------------------- We had lots and lots of older DEC equipment with the fans on the top of the cabinets and the plastic foam filters. Being in New York City, dirt is a problem. One of the first steps takes was to replace all of the foam filters with fiberglass filters, and to change them regularly. I don't know if it was a regular practice, but I did often see the techs spraying 'sticky stuff' (don't know the brand, but it was intended for this use) on the filters. Considering that this was old (PDP-9, PDP-15 and PDP-11/70) equipment which depends on lots of cooling air for it's survival, and that some of that equipment was up to 15 years old last time I looked, (and that's 24 hour 365 day continuous operation), I'd say that it's safe to use fiberglass filters. ================================================================================ Note 435.3 Equipment Cabinet Air Filters 3 of 4 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 40 lines 15-AUG-1989 02:19 -< fiber glass filters don't catch very much >- -------------------------------------------------------------------------------- One of the easiest things you can do is to keep the room air as clean as possible, and then the individual machine filters only catch the random new crud that hasn't been snagged by the better room filters. This is easy in a computer room with big airconditioners with hefty blowers and filter racks that can take the 2" or thicker HEPA pleated filters. These are so darn good that they are often used to clean cigarette smoke out of the air without the OZONE generated by the electrostatic filters, and they last a long time. If you have a lot of fresh large dust entering the space (do you brush your cats in the computer room?) you can often fit the relatively inefficient fiberglass filters in front simply to catch the large stuff before it gets to the HEPAs. The HEPAs actually slightly improve their filtering ability as the surface gets coated with crud catching crud. The pressure drop eventually gets high enough so that they need changing, but they will still be filtering well. The filters on our USED computer room A/C units are what came USED with the units, and we now have 1 1/2 years on them. Some one of these years we will have to replace them, but there is an automatic alarm that signals they are getting too dirty (pressure drop too high) that I am patiently waiting to hear go off first. Crud builds up much slower on the plastic filters on the fronts of things like 11/44s and Fugi Eagles now that we have REAL computer room airconditioning. When the foam filter gets too yukky to simply vacuum, take it to the kitchen sink, honk it with some spray cleaner then blast it from the rear through to front with the spray hose and then reinstall it when you think it is dry enough. Large sheets of foam media are readily available to custom cut new filters for any make room airconditioner. Most hardware stores have them, and refrigeration supply houses do too. It seems to be the same sort of stuff used for the computers, and is doubtlessly cheaper and easier to get. ================================================================================ Note 435.4 Equipment Cabinet Air Filters 4 of 4 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 23 lines 15-AUG-1989 11:41 -< Sometimes the obvious needs to be pointed out to me >- -------------------------------------------------------------------------------- > (do you brush your cats in the computer room?) What? And risk a catastrophic failure? > When the foam filter gets too yukky to simply vacuum, take it > to the kitchen sink, honk it with some spray cleaner then blast > it from the rear through to front with the spray hose and then > reinstall it when you think it is dry enough. That was my intent, but the filter wasn't quite up to it. The old(er) black filter simply crumbled to pieces when I touched it and the newer (light dirt-colored) ones tore apart. > Large sheets of foam media are readily available to custom cut > new filters for any make room airconditioner. Most hardware > stores have them, and refrigeration supply houses do too. It > seems to be the same sort of stuff used for the computers, and > is doubtlessly cheaper and easier to get. BINGO! The hardware stores I checked didn't have it, but I will try some A/C and refrigeration places. I thought that nobody but the computer industry used the thick foam. Dumb me. Thanks for the kick to my brain.  ================================================================================ Note 436.0 Need manuals for DR11-K & AA11-K 1 reply EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 8 lines 13-AUG-1989 21:28 -------------------------------------------------------------------------------- I am looking for documentation (manuals and/or print sets) for the DR11-K parallel interface and the AA11-K D/A interface. If you don't want to part with your originals, I promise to carefully copy them and return them to you ASAP. Send VMS MAIL to FRISBIE or call Alan Frisbie at (213) 256-2575 ================================================================================ Note 436.1 Need manuals for DR11-K & AA11-K 1 of 1 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 6 lines 22-AUG-1989 21:18 -< Fast service >- -------------------------------------------------------------------------------- > I am looking for documentation (manuals and/or print sets) > for the DR11-K parallel interface and the AA11-K D/A > interface. I now have the user's guide for both of these (thanks, Arnold!), but would still be interested in the print set.  ================================================================================ Note 437.0 EIKONIX digitizer driver needed 8 replies EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 7 lines 14-AUG-1989 12:29 -------------------------------------------------------------------------------- We have an EIKONIX digitizer on a VAX 11/785. The driver supplied by EIKONIX dates back to VMS 3. It will not run under VMS version 5; any attempt to access crashes the system. EIKONIX doesn't expect to put out an updated driver. We'd like to keep the machine running. Anybody know of a source for a VMS 5-compatible EIKONIX driver? ================================================================================ Note 437.1 EIKONIX digitizer driver needed 1 of 8 EISNER::COY "Dale E. Coy" 7 lines 14-AUG-1989 21:59 -< Rhetorical question >- -------------------------------------------------------------------------------- > EIKONIX doesn't expect to put out an updated driver. I hope somebody can help you. Just out of curiosity, given that EIKONIX isn't (apparently) going to update the driver... what did they say when you asked for the driver source code, so you could update it yourself? ================================================================================ Note 437.2 EIKONIX digitizer driver needed 2 of 8 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 13 lines 8-SEP-1989 14:51 -< A hearty handshake and a "Good luck, son". >- -------------------------------------------------------------------------------- > I hope somebody can help you. Just out of curiosity, given that > EIKONIX isn't (apparently) going to update the driver... > > what did they say when you asked for the driver source code, so > you could update it yourself? They're willing to sell us the (last) source code release. Doesn't do me a tremendous amount of good. I would become a very unpopular person were I to crash our system on a regular basis trying to debug a new driver. Users can be so picky sometimes... :-} I guess I'll just call them back occasionally and see if someone else has done it and will sell it. ================================================================================ Note 437.3 EIKONIX digitizer driver needed 3 of 8 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 37 lines 19-SEP-1989 16:32 -< Eureka! >- -------------------------------------------------------------------------------- > > I guess I'll just call them back occasionally and see if someone > else has done it and will sell it. GOTCHA! Following up a lead from those friendly folks at EIKONIX, I contacted Dr. Arthur W. Toga, at UCLA. He has a driver for an EIKONIX model 785 that will work under VMS 5.X, with SMP. For $900, you can get a 1600 or 6250 magtape with commented source, object modules, executables, and command procedures. He does NOT offer support for this. A check sent to him made out to the "Regents of the University of California (UC) wil get a tape made and sent out the next day. A Purchase Order is liable to take several weeks or months of processing. Dr. Toga allowed me to post his name in case anyone else is in need of this code. He is : Dr. Arthur W. Toga, Ph. D. Associate Professor of Neurology Assistant Chairman for Research Affairs Director, Laboratory of Neuro Imaging D & L Strauss Scholar for UCLA School of Medicine Department of Neurology Reed Neurological Research Center 710 Westwood Plaza UCLA School of Medicine Center for the Health Sciences 10833 Le Conte Ave Los Angeles CA 90024-1769 (213) 206-2101 (got all that?) Chalk up one for the happy ending boys. Now all we have to do is make his driver work with our programs... ================================================================================ Note 437.4 EIKONIX digitizer driver needed 4 of 8 EISNER::DAVIS "Steven P. Davis" 14 lines 20-SEP-1989 11:51 -< Is this for real? >- -------------------------------------------------------------------------------- In .-1, that is the lllooonnngggeeesssttt address I have ever seen. Here's the abbreviation: Dr. A. W. Toga, PhD Ass. Prof. Neur., Ass. Chrmn. Rsrch. Affrs. Dir. Lab. Neuro Imag., D & L Strauss Scholar UCLA Sch. Med. Dept. Neur., Reed Neur. Rsrch. Ctr. 710 Westwood Plaza UCLA Sch. Med., Ctr. Hlth. Scs. 10833 Le Conte Ave LA CA 90024 I'd like to see his business card. :^) ================================================================================ Note 437.5 EIKONIX digitizer driver needed 5 of 8 EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 0 lines 20-SEP-1989 14:36 -< My envelope's not big enough to hold all that address! >- -------------------------------------------------------------------------------- ================================================================================ Note 437.6 EIKONIX digitizer driver needed 6 of 8 EISNER::HAHN "DECUServe Vice Chair" 17 lines 20-SEP-1989 16:16 -< 3 lines will do >- -------------------------------------------------------------------------------- Having worked at UCLA you can get by with the following 3 lines the rest is fluff! > Dr. Arthur W. Toga, Ph. D. Associate Professor of Neurology Assistant Chairman for Research Affairs Director, Laboratory of Neuro Imaging D & L Strauss Scholar for UCLA School of Medicine Department of Neurology > Reed Neurological Research Center 710 Westwood Plaza UCLA School of Medicine Center for the Health Sciences 10833 Le Conte Ave > Los Angeles CA 90024-1769 (213) 206-2101 (got all that?) ================================================================================ Note 437.7 EIKONIX digitizer driver needed 7 of 8 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 0 lines 21-SEP-1989 00:57 -< Is that an address or a resume? :-} >- -------------------------------------------------------------------------------- ================================================================================ Note 437.8 EIKONIX digitizer driver needed 8 of 8 EISNER::DAVIS "Steven P. Davis" 3 lines 21-SEP-1989 13:47 -------------------------------------------------------------------------------- I can just imagine the guy on the other end of the phone saying, "Hang on, let me give you the address...".  ================================================================================ Note 438.0 DMF-32 ports 2-7 5 replies EISNER::NELSON_M 20 lines 15-AUG-1989 23:32 -------------------------------------------------------------------------------- Help! I have a DMF-32, purchased used and recently installed. All previous experience has been with look-alikes and a DZ-11. I am now having trouble moving some stuff from their previous TT: ports to the new TX: ports. The devices *WANT* to see DTR and after much trial and error, and a break-out box, I have concluded that DMF ports 2-7 don't have DTR high. All my TT: ports do. The TX: ports are configured identically to what the TT: ports were. Is this what is supposed to be happening or is something broken or not installed right on the DMF? I know only ports 0 and 1 on DMF's support full modem control but I didn't think DTR was part of full modem control. The I/O user's guide seems to say (for whatever that's worth) that DTR is supposed to be on all supported controllers. The DMF manual provides no help either - it seems to concentrate on what ports 0 and 1 do, ignoring the others. Can anybody enlighten me on the behavior of DMF ports 2-7? Right now, the DMF-32 seems more brain-damaged than the DZ-11! :-) Thanks. ================================================================================ Note 438.1 DMF-32 ports 2-7 1 of 5 EISNER::GRAUE "Geoff, Security Pacific Bank" 7 lines 16-AUG-1989 00:08 -< It's a feature, not a bug! >- -------------------------------------------------------------------------------- My experience with the DMF is that, indeed, DTR is not active on ports 2-7. Only pins 2, 3, and 7 are functional (xmit, recv, grnd). The DMF-32 was a "rush-job" product that cut a lot of corners (read "features") when it was developed. That product has a long history of unhappy users. Does anyone remember the fun with trying to get multi-drop DDCMP to work on the DMF? Say no more, say no more.... ================================================================================ Note 438.2 DMF-32 ports 2-7 2 of 5 EISNER::HARVEY "Jack Harvey" 11 lines 16-AUG-1989 03:17 -< DMF32 Security Problem >- -------------------------------------------------------------------------------- Most modems have a setup option to permit them to ignore DTR. Barring that, you may need a trick cable tying DSR back to DTR at the modem. Note, however, that even if you get it to work, VMS has no choice but to handle these ports on the DMF32 as local lines. That means that when a dial-in user breaks the connection without logging out, the user's process will remain, and the next person dialing in will get control of the first user's account... Kinda bad for security, if you worry about that sort of thing. ================================================================================ Note 438.3 DMF-32 ports 2-7 3 of 5 EISNER::KENNEDY "Terry Kennedy" 13 lines 16-AUG-1989 07:32 -< Consider 3rd-party DMF emulation whenmodem control is wanted >- -------------------------------------------------------------------------------- A number of 3rd-party companies make very nice DMF-32 lookalikes. Here we have 4 Able VMZ32/N boards (16 DMF ports/board, all with full modem control). They also do RS-42% and current loop as jumper options. The distribution panel packs 16 DB-25 connectors in the space of a DEC DMF-32 distribution panel. I bought 2 last week from a broker for a total of $300. That's under $10 per port! We also have 2 real DEC DMF-32's, which handle our Bitnet links (on the sync ports) and our LP25's (on the parallel ports). Can't complain about them either, although the early firmware was beastly, as was the VMS driver. Current revs of both seem to work flawlessly, though. ================================================================================ Note 438.4 DMF-32 ports 2-7 4 of 5 EISNER::FERGUSON "Linwood Ferguson, ARA Services" 22 lines 16-AUG-1989 07:51 -< Please say more (DMF's on multipoint) >- -------------------------------------------------------------------------------- > Does anyone remember the fun with trying to get > multi-drop DDCMP to work on the DMF? Say no more, say no more.... Please say more. We have several DMF's on very long line multipoint circuits. They work, they stay up, but they are incredibly, terribly slow. There is virtually no traffic on them, and even when completely idle it can take minutes to just get logged in set SET HOST (not a CPU problem; occurs also when control and trib CPU are both idle). We've had someone look at the line (we lack the tools) and their thought was that the DMF's were holding the line too long. We've played with all the parameters we can to no avail. I even have a circa-V3 internal Dec memo on tuning; again no real effect. I am reasonably sure that it is the DMF's fault. We replaced a bunch of the 750's that had them with 3500's and DMV11's and those run significantly faster (though those lines still have DMF's on them also). So say more... please... was there ever a solution? ================================================================================ Note 438.5 DMF-32 ports 2-7 5 of 5 EISNER::GRAUE "Geoff, Security Pacific Bank" 42 lines 16-AUG-1989 19:54 -< A lot goes into making performance work on MPT >- -------------------------------------------------------------------------------- As a matter of fact, I was the person who worked on the problem at ARA (and also at DELCO). The timing problem you are noticing has to do with the DECnet MPT tuning parameters. It has been a long time since I looked at this, but I recall that there are several parameters that work together to calculate timeout and line turn-around. I also recall that the type of modem and its parameters can also impact performance. The MPT algorithm is driven by change of line state and timeouts. Possibly your problem is that your modems are holding up carrier instead of toggling. In a point-to-point link, carrier is always high since it is a duplex dedicated link (no need to worry about when the other side is done sending). On a multipoint link, the master looks for the carrier to drop before it assumes it can move on to the the next slave to poll (the slaves always see carrier high on their side since the master has control - you have to be careful to configure the modems differently depending upon whether the cpu is a master or a slave). If the carrier does not drop, then a timer has to expire before the master tries again or moves on to poll someone else. This was the most common scenario I saw when we had trouble with performance over MPT DECnet. I do not have my notes anymore on what DECnet/NCP parameters you have to adjust to make MPT work more efficiently. I believe the defaults were adequate, however. They do depend upon the speed of the link and since DECnet cannot know how fast the physical link is, it cannot adjust timers to account for line speed. The reason this is important is if the link is, say, 2400 baud and the master is receiving a long block and the NCP receive timer is a little too short to allow that whole block to always make it before the timer expires, the master will abort the receive and force the slave to re-transmit. On the other hand, if the timer is too long and the modem does not drop carrier or the header gets trashed, the master will wait the timeout period before polling again. Very tricky business, this MPT stuff. Digital had a lot of trouble making this work to most customers' satisfaction. As I recall, they had planned to drop MPT in Phase V of DECnet, much to the chagrin of the minority who actually use this feature. If this is of concern to you, you may want to be sure to bring this up at the next DECUS DECnet Phase V session this fall.  ================================================================================ Note 439.0 Discussion of blank 8mm tape brands 4 replies EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 2 lines 16-AUG-1989 07:46 -------------------------------------------------------------------------------- This topic is for discussion of blank 8mm tapes. It has been moved here from the VMS conference, topic 638.*. ================================================================================ Note 439.1 Discussion of blank 8mm tape brands 1 of 4 EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 10 lines 16-AUG-1989 07:48 -< Which 8MM blanks are best? >- -------------------------------------------------------------------------------- EISNER::MEDSKER "Alan Medsker" 7 lines 11-AUG-1989 11:14 -------------------------------------------------------------------------------- Anyone have any recommendations on whi 8MM blank tapes I should use with my Exabyte (MTI) drive? We're buying Sony P6-120MPs now, from a mail-order audio/video/computer outlet. They are quite a bit cheaper than the ones you can get from the drive/controller manufacturers. What's the difference? Alan ================================================================================ Note 439.2 Discussion of blank 8mm tape brands 2 of 4 EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 18 lines 16-AUG-1989 07:49 -< We use SONY's ... >- -------------------------------------------------------------------------------- EISNER::DOHERTY "Bob Doherty" 15 lines 11-AUG-1989 20:04 -------------------------------------------------------------------------------- We have intermittent problems with our 8mm drive hanging up on certain tapes(spontaneously ejecting the tape,and/or going into mount verification, and/or software write locki the tape, etc. We were using SONY P6-120MP's and were convinced that we had run into a bad batch. We have since written a program to write 32 k blocks from start to end, rewind and read each block comparing it with what was written. Tapes which we were SURE were bad, i.e., gave one or more of the symptoms indicated above, at approximatelythe same place on the tape during backup, passed the tape verification with NO errors. At this point, I think we're being bit by a bug in the TUDRIVER, since our drive is recognized as a TU81+. I've run about 15 SONY tapes through our verifier and found no bad tapes. Caveat emptor. ================================================================================ Note 439.3 Discussion of blank 8mm tape brands 3 of 4 EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 12 lines 16-AUG-1989 07:50 -< Sony's are working fine for us >- -------------------------------------------------------------------------------- EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 9 lines 12-AUG-1989 09:58 -------------------------------------------------------------------------------- We also use the Sony P6-120MP tapes. We've had very few troubles with the Sony tapes (we've used a couple of hundred so far); in fact we had more problems (bad tapes) with the expensive vendor-certified tapes than with the Sony's we get in our stockroom. As I've noted, the Fermilab stockroom carries the Sony's which are being used by many, many people on site. One experiment to be run in this next accelerator run plans to take raw data on 8mm tapes - just a modest 10,000 tapes worth! ================================================================================ Note 439.4 Discussion of blank 8mm tape brands 4 of 4 EISNER::BERTRAM 16 lines 16-AUG-1989 20:55 -< we use sony tapes too >- -------------------------------------------------------------------------------- < Note 439.3 by EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" > -< Sony's are working fine for us >- EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 9 lines 12-AUG-1989 09:58 -------------------------------------------------------------------------------- We also use the Sony P6-120MP tapes. We've had very few troubles with the Sony tapes (we've used a couple of hundred so far); in fact we had more problems (bad tapes) with the expensive vendor-certified tapes than with the Sony's we get in our stockroom. As I've noted, the Fermilab stockroom carries the Sony's which are being used by many, many people on site. One experiment to be run in this next accelerator run plans to take raw data on 8mm tapes - just a modest 10,000 tapes worth!  ================================================================================ Note 440.0 Can't find thread on CPU differences 1 reply EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 9 lines 18-AUG-1989 11:36 -------------------------------------------------------------------------------- Software on DECUServe was a thread about the differences in complex arithmetic expersion (possibly using extende precision arithmetic) seen when running the same program on a 6000 series VAXen (with CVAX chip) versus "other" (ie 86xx, 88xx) VAXen but I can't find it. I would like to review this material as our research lab has just run into the same problem but I can't find it (even with SEARCH DECUSERVE_ALL_TITLES) !! ================================================================================ Note 440.1 Can't find thread on CPU differences 1 of 1 EISNER::TANNENBAUM "Saul Tannenbaum" 1 line 18-AUG-1989 11:49 -< Notes 360.*, this conference >- --------------------------------------------------------------------------------  ================================================================================ Note 441.0 ECMA 34 Cassette tape standard No replies EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 3 lines 21-AUG-1989 12:40 -------------------------------------------------------------------------------- Anyone know about a European data cassette standard called (I believe) ECMA 34? Does anyone make a Q-Bus interface for such a drive?  ================================================================================ Note 442.0 IVECS (Bridge Comm Inc.) install 2 replies EISNER::CLOSE 3 lines 21-AUG-1989 14:37 -------------------------------------------------------------------------------- Need assistance on properly configuring/installing an IVECS comm board by Bridge Comm on the UNIBUS. Have any of you done this before? Thanks. Ed Close (202)366-6214 ================================================================================ Note 442.1 IVECS (Bridge Comm Inc.) install 1 of 2 EISNER::GLEASON "CyberPunk" 11 lines 22-AUG-1989 00:10 -< Like a Bridge over troubled waters >- -------------------------------------------------------------------------------- > Need assistance on properly configuring/installing an IVECS > comm board by Bridge Comm on the UNIBUS. Have any of you > done this before? Thanks. Ed Close (202)366-6214 I've done this, dozens of times (though, I handled only the DEC side of things - another group programmed the parts upstream of the Ethernet transceiver). Are you looking for general info, or having a problem making it work? ================================================================================ Note 442.2 IVECS (Bridge Comm Inc.) install 2 of 2 EISNER::CLOSE 25 lines 22-AUG-1989 09:14 -< Ready to jump from the Bridge! >- -------------------------------------------------------------------------------- I'm not sure exactly what the problem "was". But it now works. I configured the board to: o a correct CSR address o emulate 4 DMF's o continuous self-test disabled o UNIBUS interrupt grant and request level = 5 o the correct SET TERM characteristics o the correct SYCONFIG params and it didn't work. The networking people could see the board through the transciever and SYSGEN configured, but when you connected to a port you couldn't generate a login. Then out of desperation, I set auto reboot to enabled (on the board) and reseeded (or is that reseated) all the jumpers and bingo...... it worked. A few questions: Is setting the IVECVS board for auto reboot on crash that important? Do you think it was the reseating of the jumpers that did it? We don't have a maintenance contract on these boards and have not had much luck (2 out of 2 boards went bad in the last 3 months). Do you recommend a maintenance contract? It took a long time to get the boards repaired and returned (in fact 1 came back still bad). Is this your experience? Thanks for your insight. Although the initial problem has been resolved, your ideas/experiences can still be of help. Eddie Close  ================================================================================ Note 443.0 Strange DEC modules X026 & X027 2 replies EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 12 lines 22-AUG-1989 21:29 -------------------------------------------------------------------------------- I picked up a bunch of interesting DEC "stuff" from a scrap dealer. The most interesting is a pair of hex Unibus boards with unusual module numbers: X026 and X027. They are definitely DEC boards, but I have never seen one with an "X" number before. X026 is labeled "IG OUTPUT BD" and has three small coax connectors labeled "Line", "Page", and "Video". X027 is labeled "IG INPUT". The circuit board part numbers are 50-12228 and 50-12229 respectively. Does anyone have any idea what these are? ================================================================================ Note 443.1 Strange DEC modules X026 & X027 1 of 2 EISNER::KENNEDY "Terry Kennedy" 18 lines 23-AUG-1989 01:49 -< "Rainbow" is an 11/34 option? >- -------------------------------------------------------------------------------- > Does anyone have any idea what these are? X026 is "RAINBOW OUTPUT HEX", used on 11/34 X027 is "RAINBOW INPUT HEX", also used on 11/34 The rest of the family is: X020 RAINBOW DATA PATH HEX X021 RAINBOW CD ROM CONTROL HEX X022 RAINBOW UNIBUS WINDOW SIMILAR TO DA11H X023 RAINBOW TAPE CONTROL HEX X025 RAINBOW,IG FONT MEMORY ARRAY,HEX X030 RAINBOW,CDC DISK CONTROL,HEX X032 IG FONT MEMORY ARRAY An odd assortment indeed. These are the *only* valid parts in the X family. I don't think "CD ROM COMTROL" has anything to do with the current meaning of "CD-ROM". Anybody else have any ideas? ================================================================================ Note 443.2 Strange DEC modules X026 & X027 2 of 2 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 0 lines 29-AUG-1989 00:17 -< Rainbow??? >- --------------------------------------------------------------------------------  ================================================================================ Note 444.0 8mm "Unattended" drives? 6 replies EISNER::DEMTY "Joe Demty" 3 lines 23-AUG-1989 15:37 -------------------------------------------------------------------------------- I'm getting a lot of brochures from vendors of 8mm 2.2-gig backup units. Does anyone REALLY use these? If so, can you really stack up a bunch of drives to run overnight "unattended"? Are they reliable? ================================================================================ Note 444.1 8mm "Unattended" drives? 1 of 6 EISNER::JARZABEK "Simone Jarzabek" 21 lines 23-AUG-1989 16:46 -< My experience with 8MM >- -------------------------------------------------------------------------------- This was discussed on Pageswapper, if I recall ... At my old site, I used an 8MM tape drive to do incrementals followed by fulls of 4 RA81 drives. This was done in a batch job which ran 3 to 3.5 hours and resubmitted itself each night. The operator only had to remember to load a tape before going home at night, the rest was automagic. The reason for both incrementals and fulls was that you could restore a file from incremental in about 10 minutes, but the access time was about 45 minutes per full saveset. The installation of the drive was a nightmare. The vendor's field service person incorrectly configured the controller board on the Unibus on my 785. (His first installation). The controller board was first, defective, and then, of the wrong rev level. The tape drive would hang. After the shakedown period, things worked fine. One caution : I could do Backups to the drive, but not Copys. Files copied to tape were never to be read back again, due to position lost errors. I was told this was a V4.7 TUDRIVER bug. Overall, whether or not to go with 8MM depends on how much risk you are willing to take in exchange for time and cost savings. Some of the risk can be minimized by doing backup more often. Simone ================================================================================ Note 444.2 8mm "Unattended" drives? 2 of 6 EISNER::DELARISCH "Arnold S. De Larisch" 27 lines 23-AUG-1989 19:20 -< Happy NEW 8MM Tape User! >- -------------------------------------------------------------------------------- >> < Note 444.0 by EISNER::DEMTY "Joe Demty" > >> -< 8mm "Unattended" drives? >- >> I'm getting a lot of brochures from vendors of 8mm 2.2-gig backup >> units. Does anyone REALLY use these? If so, can you really stack up a >> bunch of drives to run overnight "unattended"? Are they reliable? We just installed System Industry SI59 (with their CS1000 Q-Bus controller) 8 mm tape drive. It fits into the standard 5.25 inch slot on a BA23/BA123 box ( we are in a H9642 flavor cabinet). The controller looks like a standard TK50 (MSCP protocol) and requires NO DRIVER modifications. So far, I've been very happy with it. Its NOT fast on tape positioning, but is about the transfer rate of a TK50 (and holds 2GB!). Copy worked on my VMS V4.7A system (which is now VMS V5.1-1). We are using it for unattended backups late at night. You are best off either doing a single disk drive a night per tape drive OR not doing a verify pass (again, the tape positioning time is VERY slow). I understand that SI is now SHIPPING 8MM drives for HSC's. I like the technology. Exabyte has the exclusive on the 8MM drive mechnisms, which simplified getting it certified as an ANSI standard! I'm happy (but I'm a new customer to 8MM). Maybe Frank Nagy can comment more on the 8MM technology? -Arnold ================================================================================ Note 444.3 8mm "Unattended" drives? 3 of 6 EISNER::JARZABEK "Simone Jarzabek" 6 lines 23-AUG-1989 19:43 -< More experience ... >- -------------------------------------------------------------------------------- My drive (installed 10/88) was the same SI59 Exabyte deck as .-1 but on a Unibus. The COPY problems were also with 4.7A in a LAVc environment. My COPY problem was EASILY reproducible; the bug information was from SI Engineering However, the controller board used was 3rd party to SI; I was told that they were planning to manufacture their own. sj ================================================================================ Note 444.4 8mm "Unattended" drives? 4 of 6 EISNER::ESBENSEN 19 lines 24-AUG-1989 02:57 -< good luck here >- -------------------------------------------------------------------------------- Hi, We have been using a Megatape GT-88 (Exabyte) drive for almost a year now. We do full backups of all 6 of our 300MB disks each night. Aided by some software that we developed (DTA) we average 230KB/sec...without it, about 150KB/sec. This is much faster than the S L O W TK50 max speed of 45KB/sec. We have had to restore data twice in the last few months. Each time it took about 35 minutes to restore the data (just 10000 blocks or so). In the last 6 months we have had NO reported tape errors on the driver or controller (at least none that VMS saw). We have used the units on both our MVII and our new 3300. We are using Sony 120-MP tapes. Dan E., happy Exabyte user! ================================================================================ Note 444.5 8mm "Unattended" drives? 5 of 6 EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 22 lines 24-AUG-1989 07:49 -< 8mm working well for BACKUP at Fermilab >- -------------------------------------------------------------------------------- We have been using 8mm tapes for backup since early this year. Currently my group has 2 LAVCs with 2 8mm drives on each (each boot node has an 8mm drive) plus 9 standalone systems with Wren-V disks and 8mm drives. On all these systems we do unattended, nightly Backup /Image of all the disks. For our development cluster, this is 4 RA81s + 1 Wren-V + a Wren-IV being split into two jobs going to two separate tape drives simultaneously. About 1 GB+ is being written to each drive; the jobs take about 2.5-3 hours (starting at 3AM). We had quite a few teething pains but after firmware upgrades (Fermilab is working closely with Exabyte, Summus and TD Systems since we are on the verge of standardizing on 8mm tapes for data recording - or so it seems) to the Exabyte drives, problems have gone away pretty much totally. Occassionally a job aborts due to fatal parity error on the tape label processing, but restarting the job on the same tape will almost always succeed. So, as to QBus-based 8mm tapes for BACKUP: they are a major win here at Fermilab. Various experiments are using 8mm tapes for data (mostly 2nd and later processing passes at this time). One experiment plans to take its data on 8mm during the next accelerator run starting in Jan '90. This experiment hopes to write 10,000 8mm tapes of data. ================================================================================ Note 444.6 8mm "Unattended" drives? 6 of 6 EISNER::ANGLEY "Kevin Angley" 4 lines 29-AUG-1989 15:46 -< Another positive experience >- -------------------------------------------------------------------------------- Running two Summus Exabyte systems, we're luvin' it. We have orders of magnitude more confidence in our ability to recover from back than we ever had with 9-track.  ================================================================================ Note 445.0 Need questions answered about the Exabyte tape Drive 13 replies EISNER::DELARISCH "Arnold S. De Larisch" 21 lines 24-AUG-1989 17:13 -------------------------------------------------------------------------------- O.K. .... foolish question #1 of the day... I Purchased a couple of MAXELL P6-90 Metal partical tapes to use in my System Industries Exabyte 8MM tape drive. I attempted to physically load the tape, and the drive. After a minute or so, the drive spit the tape back out at me. After trying a second Maxell tape (which was spit out), I tried a REAL Exabyte tape (it also got spit out). After rebooting the system (and powered it down also), I tried the EXABYTE tape, no problem. I then tried the Maxell tape again, it spit the tape out and refused to load any tapes untill powered down and rebooted. 1. No errors were EVER seen by the controller. 2. I assume an 'ERROR' bit gets stuck 'on' which refuses further tapes. 3. Why doesn't the MAXELL tape work (the only difference I can see between it and the Exabyte tape is a tape clear leader on the MAXELL) Is this the problem? Does the P6-120mp tapes from SONY have a leader? 4. Is there any way of 'resetting' the Exabyte unit from the front pannel? I'm really confused! -Arnold ================================================================================ Note 445.1 Need questions answered about the Exabyte tape Drive 1 of 13 EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 10 lines 25-AUG-1989 08:46 -< Sony tape has clear leader >- -------------------------------------------------------------------------------- >>3. Why doesn't the MAXELL tape work (the only difference I can see between >> it and the Exabyte tape is a tape clear leader on the MAXELL) >> Is this the problem? Does the P6-120mp tapes from SONY have a leader? A spanking new Sony P6-120MP tape does have a clear leader showing on the tapeup reel. A tape which has been used in an Exabyte drive does not show the leader, but has a turn or two (at least) of coated tape on the takeup reel. So the tape obviously has a clear leader, but the drive does not rewind the tape back to it. Could your problem be related to the length of the clear leader? ================================================================================ Note 445.2 Need questions answered about the Exabyte tape Drive 2 of 13 EISNER::DELARISCH "Arnold S. De Larisch" 21 lines 30-AUG-1989 00:06 -< Can you suggest alternatives! >- -------------------------------------------------------------------------------- >> < Note 445.1 by EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" > >> -< Sony tape has clear leader >- >> >>3. Why doesn't the MAXELL tape work (the only difference I can see between >> >> it and the Exabyte tape is a tape clear leader on the MAXELL) >> >> Is this the problem? Does the P6-120mp tapes from SONY have a leader? >> A spanking new Sony P6-120MP tape does have a clear leader showing on >> the tapeup reel. A tape which has been used in an Exabyte drive does >> not show the leader, but has a turn or two (at least) of coated tape >> on the takeup reel. So the tape obviously has a clear leader, but the >> drive does not rewind the tape back to it. Could your problem be related >> to the length of the clear leader? No, I don't think so, the leader is rather short (about 3 inches max). It seems that the TAPE drive was rejecting the tape NOT the controller. It was when the tape was being 'loaded' not during 'VMS' access to the tape. Can you suggest alternative (other than Exabyte's OVERPRICED) tapes? -Arnold (who is still on Vacation, sort of!) ================================================================================ Note 445.3 Need questions answered about the Exabyte tape Drive 3 of 13 EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 9 lines 30-AUG-1989 08:30 -< Need new Exabyte firmware? Version 4$24 >- -------------------------------------------------------------------------------- Fermilab has dozen's of 8mm drives from several vendors (who all get them from Exabyte). Everyone uses the Sony P6-120mp tapes; in fact we get them from the Fermilab stockroom. I've heard of no one with any consistent problems with loading tapes into the drive. I've seen such problems but they've all gone away since the most recent Exabyte PROM upgrade (this is a PROM for the microprocessor ON THE TAPE DRIVE ITSELF) to version 4$24. Since we installed this upgrade (from Exabyte) essentially all our problems have vanished and our 8mm drives have become very stable and reliable. ================================================================================ Note 445.4 Need questions answered about the Exabyte tape Drive 4 of 13 EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 9 lines 30-AUG-1989 08:34 -< Said again: try the Sony tapes >- -------------------------------------------------------------------------------- I forgot to add that the tape loading problems we've had in the past were very infrequent and inconsistent. My group's experience is that we've had more problems with and more bad tapes from the vendor's expensive 8mm tapes than we've had from the Sony's (and we've run a couple of hundred Sony tapes through our drives and a few dozen of the expensive variety). The vendor not named is NOT Exabyte but an OEM of Exabyte's. THe moral is that we just go to the stockroom and get a few boxes (10 tapes each) of Sony's as we need them. Failure rates for the Sony's appear to be in the very low few percent. ================================================================================ Note 445.5 Need questions answered about the Exabyte tape Drive 5 of 13 EISNER::NORTON "Bill Norton" 2 lines 30-AUG-1989 09:18 -------------------------------------------------------------------------------- Sony tapes from a Sony distributor cost us $5.95 when we bought 40, so the $10 price at the supermarkets isn't real either. ================================================================================ Note 445.6 Need questions answered about the Exabyte tape Drive 6 of 13 EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 11 lines 30-AUG-1989 12:51 -< Tape transport sensor problem? >- -------------------------------------------------------------------------------- RE: < Note 445.2 by EISNER::DELARISCH "Arnold S. De Larisch" > -< Can you suggest alternatives! >- |It seems that the TAPE drive was rejecting the tape NOT the controller. |It was when the tape was being 'loaded' not during 'VMS' access to the tape. Although this doesn't add much to the technical line of this thread, my VCR has behaved exactly the same way, accepting some brands of tape and rejecting others. The problem is in one of the sensors that detects tape position and loading. The final answer for me was to replace the tape transport, but it may be a sensor. ================================================================================ Note 445.7 Need questions answered about the Exabyte tape Drive 7 of 13 EISNER::JARZABEK "Simone Jarzabek" 3 lines 30-AUG-1989 17:45 -< Another vote for PROM upgrade >- -------------------------------------------------------------------------------- re .3 :(this is a PROM for the microprocessor ON THE TAPE DRIVE ITSELF) I also experienced "tape" problems which went away when the drive was upgraded. ================================================================================ Note 445.8 Need questions answered about the Exabyte tape Drive 8 of 13 EISNER::DELARISCH "Arnold S. De Larisch" 10 lines 5-SEP-1989 04:34 -< Who supplies the ROM upgrade? >- -------------------------------------------------------------------------------- Frank, Thanks for the info on the PROM upgrade, I'll take a look at the Drive tomorrow (when I'm in the office). Who do I acquire the ROM from? System Industries (who sold the drive and have a 1 Yr. onsite warentee) or from Exabyte? -Arnold About to try Sony P6-120MP tapes! ================================================================================ Note 445.9 Need questions answered about the Exabyte tape Drive 9 of 13 EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 15 lines 5-SEP-1989 08:48 -< WE got the upgrades from Exabyte directly >- -------------------------------------------------------------------------------- < Who supplies the ROM upgrade? > Arnold, I can't answer for your case, though in your shoes I start with SI before bugging Exabyte. In our case, though many/most of our drive were bought from Summus Computers, we got the upgrade PROMs from Exabyte. This was because Exabyte had gotten involved in trying to help Fermilab solve some of its problems with the 8mm drives. For other's information, most of these problems were related to non-BACKUP usage of the tapes for experiment data acquisition and analysis tapes (mostly ANSI tapes). Some of these problems related to high data rates and others to pushing the capacity (trying to use the last 10% between the logical-EOT and the real physical-EOT: important for experiments which plan to wite 10,000 8mm tapes on a limited budget). ================================================================================ Note 445.10 Need questions answered about the Exabyte tape Drive 10 of 13 EISNER::JARZABEK "Simone Jarzabek" 1 line 5-SEP-1989 11:42 -< We got the upgrade from SI >- -------------------------------------------------------------------------------- < Who supplies the ROM upgrade? > ================================================================================ Note 445.11 Need questions answered about the Exabyte tape Drive 11 of 13 EISNER::DELARISCH "Arnold S. De Larisch" 11 lines 6-SEP-1989 17:19 -< Sony P6-120mp's work great! >- -------------------------------------------------------------------------------- I don't know why, but the Sony P6-120mp seem to be working. I really can't figure out why the Maxells were causing trouble? I guess I'll be using Sony from now on! Thanks for all the help and info! -Arnold P.S. Whomever indicated that System Industry provided the ROM upgrade on the Exabyte, did you have to contact them or did they come out and install it as part of your maintenance agreement? ================================================================================ Note 445.12 Need questions answered about the Exabyte tape Drive 12 of 13 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 3 lines 6-SEP-1989 21:33 -< TDK are fine also. >- -------------------------------------------------------------------------------- > -< Sony P6-120mp's work great! >- Here we are using TDK 8mm cassettes, and they work great too. ================================================================================ Note 445.13 Need questions answered about the Exabyte tape Drive 13 of 13 EISNER::JARZABEK "Simone Jarzabek" 7 lines 7-SEP-1989 12:40 -< ROM Upgrade - more info >- -------------------------------------------------------------------------------- > P.S. Whomever indicated that System Industry provided the ROM upgrade on > the Exabyte, did you have to contact them or did they come out and > install it as part of your maintenance agreement? I called them (since the drive was eating tape at the time) and the FS Engineer came out & swapped out the whole drive. Simone  ================================================================================ Note 446.0 CIBCI problems on 85xx series? 1 reply EISNER::BRUCKER "Cyberspace Jockey" 7 lines 24-AUG-1989 20:11 -------------------------------------------------------------------------------- Has anyone had problems using CIBCI-AA or CIBCI-BC CI connections on 8530 or 8550 systems? What are the benifits of using the CIBCA-AA option on these systems? Thanks! Kenb ================================================================================ Note 446.1 CIBCI problems on 85xx series? 1 of 1 EISNER::FULLER_S "Funky cold medina" 23 lines 24-AUG-1989 20:26 -< They all work... >- -------------------------------------------------------------------------------- > Has anyone had problems using CIBCI-AA or CIBCI-BC CI connections > on 8530 or 8550 systems? What are the benifits of using the CIBCA-AA > option on these systems? What's a CIBCI-BC? The CIBCI is basically a CI750 with a BI interface (instead of the 11/750 interface). The interface between the BI boards and the CI box is 16 bits. The CIBCA-A is a complete CI port on 2 BI boards, so there is no 16 bit data path to slow things down. That in itself is a benefit, I reckon. There is a CIBCA-B, which is basically the same as the CIBCA-A, except it has the microcode onboard in ROM. It also has a slightly higher throughput than the -A. What kind of problems are you referring to? Are you having problems? Are you soliciting war stories? Or, are you trying to decide which to buy? If deciding which to buy, get the CIBCA-B (in my opinion). Stu  ================================================================================ Note 447.0 DVORAK keyboard, anyone? 7 replies EISNER::WOOD_J "Allan Wood, STS Consultants, Ltd." 19 lines 25-AUG-1989 12:29 -------------------------------------------------------------------------------- I'm not sure if this is the right conference for this question, but here goes ... Does anybody know of a source for a DVORAK terminal/keyboard? One of our users can touchtype with DVORAK but not QWERTY, and wondered why he couldn't configure his terminal to DVORAK (since he could configure it to Flemish/Danish/German/you name it!). I have to admit he has a good point. Even though DVORAK typists are rare birds, I will acknowledge the superiority of the Dvorak set-up to QWERTY. I wouldn't think it that difficult to include the DVORAK keyboard as an alternative in the VT200 Set-Up menu. I've called DEC Direct but couldn't even find anybody who had heard of a DVORAK keyboard. I've also done a couple of searches here on DECUServe for notes about DVORAK, but no luck. So can anbody out there point to a supplier of a DVORAK terminal? ================================================================================ Note 447.1 DVORAK keyboard, anyone? 1 of 7 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 28 lines 25-AUG-1989 15:21 -< How about Dvorak on a VAXstation? >- -------------------------------------------------------------------------------- >< Note 447.0 by EISNER::WOOD_J "Allan Wood, STS Consultants, Ltd." > > -< DVORAK, anyone? >- Does anybody know of a source for a DVORAK terminal/keyboard? Well, if there is a VAXstation available, it's possible to make it understand Dvorak (at least it is under UIS; I'd assume that the Propeller-heads/Tech Weenies at MIT would have included this functionality in X (and hence DECwindows) as well...) The LK201 is intelligent; either you could try to blast a new PROM for the 8051 processor, or you could make a Dvorak converter box that would sit between the LK201 and the terminal and translate the keycodes appropriately (Hmmm, not a bad idea for a product...). > I have to admit he has a good point. Even though DVORAK typists > are rare birds, I will acknowledge the superiority of the Dvorak > set-up to QWERTY. I wouldn't think it that difficult to include > the DVORAK keyboard as an alternative in the VT200 Set-Up menu. I remember reading somewhere that although the Dvorak keyboard is more logical than QWERTY, Dvorak touch typists in general tend to top out at about the same WPM as QWERTY touch typists. Not that's it relevant; just relating an interesting tidbit of info I came across. As to it being difficult to provide Dvorak in a VT200 Set-up, it's probably trivial. If you're DEC, that is.... :-) Ed ================================================================================ Note 447.2 DVORAK keyboard, anyone? 2 of 7 EISNER::WOOD_J "Allan Wood, STS Consultants, Ltd." 19 lines 25-AUG-1989 15:59 -< Nope, VAXStation not the solution >- -------------------------------------------------------------------------------- > Well, if there is a VAXstation available, it's possible to make it > understand Dvorak As a matter of fact there are a couple of VAXstations here, BUT... ...not for this particular user! (I'm still trying to convince my boss *I* could use one!) So we're still looking for a way to provide and/or turn a VT200 compatible terminal into a Dvorak terminal ... > I remember reading somewhere that although the Dvorak keyboard is more > logical than QWERTY, Dvorak touch typists in general tend to top out at > about the same WPM as QWERTY touch typists. That's interesting ... I had always understood that the QWERTY keyboard was *specfically* designed to *slow down* typists, because the early mechanical typewriters tended to get jammed if they went too fast! ================================================================================ Note 447.3 DVORAK keyboard, anyone? 3 of 7 EISNER::PRIGOT "Jonathan M. Prigot" 9 lines 25-AUG-1989 16:07 -< Hen's teeth >- -------------------------------------------------------------------------------- The best solution would be for your DECsalesrep 200 to research this internal to DEC. We have a user here who has a friend who has a friend who works at DEC and knew of a group of engineers at DEC who were into Dvorak keyboards. He was able to ask a favor and get one of these beasties. Since it was a customized (PROM) LK201, the part number is the standard LK201-AA. ================================================================================ Note 447.4 DVORAK keyboard, anyone? 4 of 7 EISNER::COY "Dale E. Coy" 7 lines 25-AUG-1989 21:46 -< Alternatives >- -------------------------------------------------------------------------------- I'm absolutely sure that sometime in the last year I've seen a Dvorak model from one of the VT-xxx-compatible vendors. Since I had no personal interest, I didn't save the literature and don't remember the vendor. Maybe a telephone survey of all the terminal makers? Alternatively, there are terminal emulators for PCs that are capable of doing Dvorak. ================================================================================ Note 447.5 DVORAK keyboard, anyone? 5 of 7 EISNER::GLEASON "CyberPunk" 7 lines 26-AUG-1989 04:45 -< An alternate DVORAK workstation... >- -------------------------------------------------------------------------------- Well, the Apple II/c has a switch on it that allows selecting between Dvorak and QWERTY...sounds like just the terminal you need... ;-) ================================================================================ Note 447.6 DVORAK keyboard, anyone? 6 of 7 EISNER::CROWELL "Shefth of the Fourth Order" 7 lines 26-AUG-1989 19:31 -< NDS terminal does it >- -------------------------------------------------------------------------------- My GP-220 terminal from Northwest Data Systems has DVORAK as one of its setup options (along with all the language setups - except perhaps for Walloon-Belgian). I suppost the keycaps can be moved around accordingly. It also does Tektronix 4014 emulaton for graphics. I've been quite happy with it for about 3 years. (And I have no vested interest in NDS.) ================================================================================ Note 447.7 DVORAK keyboard, anyone? 7 of 7 EISNER::DAVIS "Steven P. Davis" 10 lines 28-AUG-1989 12:11 -< more legends >- -------------------------------------------------------------------------------- > That's interesting ... I had always understood that the QWERTY keyboard > was *specfically* designed to *slow down* typists, because the early > mechanical typewriters tended to get jammed if they went too fast! The story I heard was that the original keyboard layout caused keys to jam because frequently typed keys were not spread out across the carriage. For example, the 'e' and 'i' strikers were physically close to each other. They spread out the frequently used strikers to avoid the jamming.  ================================================================================ Note 448.0 DECserver discussion 2 replies EISNER::KENNEDY "Terry Kennedy" 3 lines 26-AUG-1989 20:48 -------------------------------------------------------------------------------- This topic will be used for discussing technical information related to the DECserver family of products. ================================================================================ Note 448.1 DECserver discussion 1 of 2 EISNER::KENNEDY "Terry Kennedy" 7 lines 26-AUG-1989 20:51 -< DS200 / RTS/CTS flow control >- -------------------------------------------------------------------------------- The DECserver 200 will always honor RTS/CTS flow control, no matter if the port characteristic FLOW is XON, CTS, or DISABLED. RTS/CTS go directly to the serial chip in the DS200. There is a problem, however. Your RTS/CTS flow control must be synchronized with the start of a character emitted from the DS200 or the character being transmitted will be corrupted. ================================================================================ Note 448.2 DECserver discussion 2 of 2 EISNER::KENNEDY "Terry Kennedy" 44 lines 26-AUG-1989 21:11 -< DS200 debugging information >- -------------------------------------------------------------------------------- Warning: This note presents information which; while useful; may corrupt information in your DECserver 200; may require re-booting the DS200 if used; and may compromise the security of your DECnet *if* a user with the server's privileged password can gain access to port 1 of the server. Press KP3 now if you'd rather not see this information...  The DS200 has embedded ODT. To enable it, issued the privileged command SET SERVER ODT ENABLED. To enter ODT, issue the command ODT from port 1 of the server. ODT is a no-op on the other ports. The ODT prompt is a *, and the following commands are available: E n - Examine memory at address n EB n - Like E, but bytes E@ n - Examine indirect through address n $An - Open CPU register n $Dn - " " " n $SR - Open status register RD - Register dump D s e - Dump memmory from start to end G - Exit from ODT G n - Start at address n Bn r - Set breakpoint n at address r BC n - Clear breakpoint n $B - Display breakpoints SS - Set single step NS - Clear single step The addresses in the DS200 are: RAM - 000000-05ffff EEPROM - 080000-080fff EPROM - 180000-187fff Net Adr - 1c0000-1c003f So, for example, if you want to determine the physical address (remember, this is an *example*, there are lots of easier ways to do this), you would do: E 1c0000 Which would show you 6 words. Ignore the FF, and take the low byte as part of the address.  ================================================================================ Note 449.0 MicroVAX II Default Boot Device 1 reply EISNER::KRULEWICZ 12 lines 28-AUG-1989 10:57 -------------------------------------------------------------------------------- Is there a way to set a MicroVAX II default boot device? If for instance, your production system is on DUB0: and your maintanance system is on DUA0:. Is there a way to set up the system so that if there is an automatic reboot, either because of a power outage or because the operator specified REBOOT in SYSHUTDOWN, that the system will reboot on DUB0:? - Jerry ================================================================================ Note 449.1 MicroVAX II Default Boot Device 1 of 1 EISNER::KILGALLEN "Larry Kilgallen" 12 lines 28-AUG-1989 11:28 -< Only by twiddling your disk directories >- -------------------------------------------------------------------------------- You can manipulate the presence or absence of a [SYS0] directory (via RENAME, rather than DELETE ;-) ) on the earlier drive to cause the ROM to boot off the later drive. Do not try this on a disk which is the current system disk. This means REBOOT can go in one direction but not the other. I presume by "maintenance system" you mean a copy of VMS which has the RENAME command to get back, rather than some diagnostic supervisor. I realize you really wanted ROM commands as on succeeding MicroVAXen, but I really wanted them to release the MicroVAX-II when they did, albeit without every feature one might contemplate.  ================================================================================ Note 450.0 LP26 OCR B band search 3 replies EISNER::LEJA "Drew Leja, InterVarsity VAXinator" 8 lines 28-AUG-1989 12:47 -------------------------------------------------------------------------------- I am in search of alternate printer bands for the DEC LP26 (aka DataProducts B600). Specifically, I want to use an OCR B band so we can have stuff automatically scanned. Also, can someone confirm our observations that the LP26 goes through printer ribbons faster than a hot knife through butter? I'll print reports that are 3000-4000 pages long and have to change the ribbon before completion! ================================================================================ Note 450.1 LP26 OCR B band search 1 of 3 EISNER::KENNEDY "Terry Kennedy" 24 lines 29-AUG-1989 01:12 -< Try Dataproducts directly; using recycled paper? >- -------------------------------------------------------------------------------- > I am in search of alternate printer bands for the DEC LP26 > (aka DataProducts B600). Specifically, I want to use an > OCR B band so we can have stuff automatically scanned. Try Inmac. If that fails, try Dataproducts directly. They're in Woodland Hills, CA. (714, I think). Comment about scanning: The data is already in a computer (you're printing it). There is probably a better way to use it than re-scanning it. On the other hand, I have seen LP2x's used to print sequence codes on mark-sense paper, but that is *not* OCR. > Also, can someone confirm our observations that the LP26 > goes through printer ribbons faster than a hot knife through > butter? I'll print reports that are 3000-4000 pages long > and have to change the ribbon before completion! This should only happen if you're using a) Brand-x ribbons or b) recycled paper. Recycled paper is so loose it draws the ink right out of the paper. You may also want to check the "Copies" dial and make sure it is set appro- priately to the number of copies you're printing. We get about 15,000 pages/ribbon here on multiple (over a dozen) LP25 printers, which are the same machine underneath. ================================================================================ Note 450.2 LP26 OCR B band search 2 of 3 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 9 lines 29-AUG-1989 01:23 -< Dataproducts phone numbers >- -------------------------------------------------------------------------------- > Try Inmac. If that fails, try Dataproducts directly. They're in > Woodland Hills, CA. (714, I think). Dataproducts Corporation 6200 Canoga Avenue Woodland Hills, CA (818) 887-8000 Dataproducts Corp. -- Supplies Division: (213) 384-8685 ================================================================================ Note 450.3 LP26 OCR B band search 3 of 3 EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 3 lines 29-AUG-1989 12:26 -< You have a problem that is apparently local to you >- -------------------------------------------------------------------------------- We have an LP26 and change ribbons much less frequently that once per box (2,800 pages) of paper. We change about once a week and use maybe five or six boxes of paper in that week.  ================================================================================ Note 451.0 OCR scanners - mid to high end? No replies EISNER::HASSINGER "Bob Hassinger" 17 lines 29-AUG-1989 10:37 -------------------------------------------------------------------------------- My boss asked me today about information on mid to high end OCR scanners as an alternative to the "Apple-Scanner class" that is now becoming quite common in our company. Specifically, he asks: "I am interested in info on high speed (for large volume), high quality (reliable character recognition), possibly intelligent (like Kirzweil used to have with their reading machine), versatile (many font styles), etc." I have looked at this area just a little, mostly on the low end. I recall finding a couple of years ago that there seemed to be a significant break at the point where you started talking about scanners that where capable of reliably reading proportionally spaced material such as typeset documents - for example government publications, standards, and so on. Any input would be appreciated.  ================================================================================ Note 452.0 Standard Unibus vs Modified Unibus (MUD) 2 replies EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 35 lines 29-AUG-1989 12:16 -------------------------------------------------------------------------------- I am a bit puzzled over the three types of hex Unibus slots: Standard Unibus, Modified Unibus (MUD), and Extended Unibus (EUB). I know that the difference between Standard and Modified Unibus is that 15 pins on the A & B fingers have been re-designated to accommodate MOS memory voltages, parity signals, and test points. I also understand what would happen if the wrong type of board were plugged into the wrong backplane (or is that the right board in the wrong backplane?). I know nothing about the Extended Unibus, but I assume that more pins were re-designated to accommodate 22-bit addressing memory boards. I would appreciate any documentation on the EUB. Is it used on anything besides the PDP-11/44 memory slots? It is obvious that this only applies to hex boards and that they fall into the following categories: 1) Boards which don't care. Several popular boards (RL11, DZ11, KW11-K) don't use the affected pins, so obviously fit here. 2) Boards which MUST be installed in a Modified Unibus slot. Some memory boards obviously fit in this category. The question is: which ones? 3) Boards which MUST be installed in an Extended Unibus slot. I assume that memory boards larger than 256KB would fit here, but are there any others? 4) Boards which MUST be installed in a Standard Unibus slot. Are there any hex boards that fall in this category? ================================================================================ Note 452.1 Standard Unibus vs Modified Unibus (MUD) 1 of 2 EISNER::KENNEDY "Terry Kennedy" 34 lines 29-AUG-1989 20:43 -< More information >- -------------------------------------------------------------------------------- > I am a bit puzzled over the three types of hex Unibus slots: > Standard Unibus, Modified Unibus (MUD), and Extended Unibus > (EUB). You got the explanations right, though, so you can't be *that* puzzled, which is more than I can say for DEC on some occasions ;-) > I would appreciate any documentation on > the EUB. Is it used on anything besides the PDP-11/44 > memory slots? Used on 11/24 when the KT24 module is installed. > 2) Boards which MUST be installed in a Modified Unibus slot. > Some memory boards obviously fit in this category. > The question is: which ones? Memories which hold over 32Kw. Thus, all "current" memories, notably the MS11-L, MS11-M, and MS11-P families. > 3) Boards which MUST be installed in an Extended Unibus slot. > I assume that memory boards larger than 256KB would fit here, > but are there any others? Any memory board which is configured for a 22-bit address space must be either placed in an EUB slot, or have its address comparitor set to 18-bit mode. The MS11-L and MS11-M are of this type, and use jumpers (not switch- es) to set the bus mode. > 4) Boards which MUST be installed in a Standard Unibus slot. > Are there any hex boards that fall in this category? Core memories . Also, any board which uses the core voltages (I know of none, other than the core boards themselves). ================================================================================ Note 452.2 Standard Unibus vs Modified Unibus (MUD) 2 of 2 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 9 lines 29-AUG-1989 23:07 -< I'm not sure >- -------------------------------------------------------------------------------- | > 4) Boards which MUST be installed in a Standard Unibus slot. | | | Core memories . Also, any board which uses the core | voltages (I know of none, other than the core boards | themselves). But it is the MUD slots that have three pins assigned to +20 (core). Did I miss something?  ================================================================================ Note 453.0 Help needed finding rackmount hardware 3 replies EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 7 lines 29-AUG-1989 14:50 -------------------------------------------------------------------------------- Does anyone have a DEC part number (or, perhaps better, an alternate source) for those spring-clip-nuts that they use for rackmounting things? The local hardware store will happily sell me "Tinnerman nuts", but these are greatly inferior, the "threads" consisting of the edges of crudely- stamped sheetmetal. ================================================================================ Note 453.1 Help needed finding rackmount hardware 1 of 3 EISNER::KENNEDY "Terry Kennedy" 7 lines 29-AUG-1989 20:45 -< How many do you need? >- -------------------------------------------------------------------------------- > Does anyone have a DEC part number (or, perhaps better, an alternate source) > for those spring-clip-nuts that they use for rackmounting things? I can get you the part number, but I'm sure you don't want to pay $5.00 *each*!!! and wait 120 days. Drop me a mail message with how many you need and I'll see what I can do... ================================================================================ Note 453.2 Help needed finding rackmount hardware 2 of 3 EISNER::COY "Dale E. Coy" 8 lines 29-AUG-1989 21:03 -< Available commercially >- -------------------------------------------------------------------------------- These are a fairly standard item in the electronics industry. Try any large electronic parts distributor in your area - somebody who carries relay racks. Alternatively, they can be ordered from a number of relay rack manufacturers. My favorite source is AMCO, but probably available from Everest (exhibits at DEXPO every time) and others. You'll probably have to order 50 or 100, but they're not that expensive. ================================================================================ Note 453.3 Help needed finding rackmount hardware 3 of 3 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 7 lines 30-AUG-1989 11:24 -< Get Fastened ! >- -------------------------------------------------------------------------------- If you live in a big city try the Yellow Pages (business-to-business) and look under "Screws" and/or "Fasteners". You should find several distributors who would be willing to sell in low volume (onezie, twozies). Their prices are usually much better that your local hardware store and you would ba **AMAZED** at the selection of screws, nuts, bolts and other "thingies" that they have. (We in the auto business call them "fasteners" instead of nuts and bolts !)  ================================================================================ Note 454.0 Emulex Qbus disk ctlr BR problems 1 reply EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 82 lines 29-AUG-1989 15:17 -------------------------------------------------------------------------------- Some discussion over in PUBLIC_DOMAIN_SOFTWARE wrt DECUS uucp has brought up some Qbus compatibility issues with older Emulex disk controllers. The problem results in a system crash during periods of high activity by other controllers on the bus. Here is Mark Pizzolato's explanation. Note that while he specifically references the Emulex QD21, the same problem has been noted with other Emulex disk controllers. ---------------------------------------------------------------------- The Emulex QD21 as originally shipped (prior to about January 1988) generated interrupts at BR5. This emulates exactly the Unibus UDA interrupt request behavior, but NOT the KDA or RQDXn, which generate their interrupts at BR4. The crash situation in VMS will arise when - the QD21 disk controller is placed behind a BR4 device, in this scenario a terminal mux, in the bus - the driver for the mux has raised IPL to device IPL to lock out interrupts, and then - the QD21 and the terminal mux both request interrupts at (effectively) the same time Device IPL for the terminal mux happens to be correspond to BR4, so the internal IPL will reflect that. Now, the QD21 asserts BR5 AND the terminal mux asserts BR4. The CPU interrupt arbitration logic compares the interrupt request level to the current CPU IPL and if the CPU IPL is less than the interrupt request level, the interrupt IS granted. In this case, CPU IPL=4, and highest BR = 5, so the grant happens. Unfortunately, the Qbus, unlike the Unibus, has ONLY ONE interrupt grant line, and given the current implementation logic on the requesting boards the one physically closest to the CPU will be the one to TAKE the interrupt. So, although the grant was intended for the BR5 device, the fact that the BR4 device is closer to the CPU than the BR5 device, allows the BR4 device to interrupt the CPU! Do you see it yet? The terminal driver has attempted to lock out interrupts from its controller, but the hardware has allowed one in anyway! This can cause many imaginable tangled results, one of which is eventually noticed as the crash you saw. This problem is really a design flaw in the Qbus and/or its interrupt arbitration specifications. If things behaved as above AND ALL boards were REQUIRED to monitor the other (higher) interrupt lines before taking an interrupt request (i.e. the would not take it if they saw a higher request line asserted), then things would be fine. This, however, is NOT in the specification and is NOT implemented on ANY boards (including the ones made by Digital). Emulex boards have been the most frustrating with respect to this problem. It took Emulex more than 6 months to come up with a fix after I reported the problem. It took me several months to convince them that there actually WAS a problem. Emulex boards DO NOT have user selectable BR levels. Dilog and other vendor's usually have either jumpers, or simple etch cuts that can be made to select the BR level, but not Emulex. To fix it: Your QD21 board needs a jumper cut, and a different one inserted, and is probably out of rev firmware wise anyway. To work around: Move the QD21 ahead of the DZQ in the bus. If I were to do it, I would put the DEQNA in Slot 4 AB, the QD21 in Slot 4 CD, and the DZQ in slot 5 AB. You might then say that the same could happen with the DEQNA, but as it turns out DEC found this problem themselves when they started to cause HEAVY DEQNA interrupt loads when LAVCs first started, so they ended up changing the DEQNA drivers to use IPL 31 as DEVICE IPL, which is a software work around for the problem. The recommended position for the DEQNA is closest to the CPU (after memory). I have configured almost 100 MicroVAXen this way with successful results. Be warned, the same BR level problem used to exist with the Emulex TC03, but tape interrupt activity is usually MUCH lower than disk so the likelihood of this very timing-dependent scenario is much lower. ================================================================================ Note 454.1 Emulex Qbus disk ctlr BR problems 1 of 1 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 17 lines 29-AUG-1989 23:01 -< It IS in the specification >- -------------------------------------------------------------------------------- > This problem is really a design flaw in the Qbus and/or its > interrupt arbitration specifications. If things behaved as > above AND ALL boards were REQUIRED to monitor the other > (higher) interrupt lines before taking an interrupt request > (i.e. the would not take it if they saw a higher request line > asserted), then things would be fine. This, however, is NOT in > the specification and is NOT implemented on ANY boards > (including the ones made by Digital). On page 122 of the PDP-11 Bus Handbook, it specifies this monitoring protocol as the "4-Level Interrupt Configuration". It is also described in every MicroVAX CPU Module User's Guide (EK-KA630-UG for the MicroVAX-II). True, it is not REQUIRED that ALL boards do this, but I consider it a design flaw if they do not. Unfortunately, there are a lot of lazy engineers out there.  ================================================================================ Note 455.0 LN03+ Toner Light 14 replies EISNER::HEROUX "Bob Heroux" 3 lines 29-AUG-1989 15:59 -------------------------------------------------------------------------------- Has anyone ever experienced the add toner light staying on constantly?? Our's was fine until the first time it needed toner and it's been on ever since. ================================================================================ Note 455.1 LN03+ Toner Light 1 of 14 EISNER::MAYHEW "Bill Mayhew" 5 lines 29-AUG-1989 17:01 -< "Too empty?" >- -------------------------------------------------------------------------------- Nope, I haven't seen this. But note that if the toner hopper gets *really* empty, it takes two full cartridges to refill it, and I believe that adding just 1 will not be sufficient to turn out the light. Did you possibly let your LN03 get *very* low on toner the first time, such that it's never been "full" ever since? ================================================================================ Note 455.2 LN03+ Toner Light 2 of 14 EISNER::BURNET "John Burnet" 0 lines 29-AUG-1989 20:41 -< Try stirring the toner to cover the spring thing >- -------------------------------------------------------------------------------- ================================================================================ Note 455.3 LN03+ Toner Light 3 of 14 EISNER::COY "Dale E. Coy" 12 lines 29-AUG-1989 21:11 -< It COULD be a stuck light! >- -------------------------------------------------------------------------------- > Did you possibly let your LN03 get *very* low on toner the > first time, such that it's never been "full" ever since? CAREFULLY - I once had an LN03 that experienced a hardware problem that caused the "add toner" light to be STUCK on. I also had fairly-well-trained users. Guess what - the third user to "ADD TONER" was very disturbed when all of that black stuff (with nowhere else to go) got all over her dress. ================================================================================ Note 455.4 LN03+ Toner Light 4 of 14 EISNER::MAYHEW "Bill Mayhew" 8 lines 29-AUG-1989 22:17 -< Necessity is the mother of... >- -------------------------------------------------------------------------------- > Guess what - the third user to "ADD TONER" was very disturbed when > all of that black stuff (with nowhere else to go) got all over her > dress. Aha, now I know why DEC has recently introduced the LN03 Cleaning Kit, which includes everything from white cotton gloves to plastic "drop cloths"... ================================================================================ Note 455.5 LN03+ Toner Light 5 of 14 EISNER::HEROUX "Bob Heroux" 10 lines 30-AUG-1989 08:10 -< blinking red light >- -------------------------------------------------------------------------------- As soon as the light went on the first time, I added toner. I've also tried adding two cartridges on occasion but that has not worked either. This condition has been going on for about a year but the light is always on. I guess it's just stuck. If it ever needs service I'll have it fixed at that time. Since I do not have a maintenance contract on it I don't want to call DEC Field Service just for this. ================================================================================ Note 455.6 LN03+ Toner Light 6 of 14 EISNER::DRIESMAN "Debbie Driesman" 4 lines 30-AUG-1989 11:30 -< possibly a dirty sensor? >- -------------------------------------------------------------------------------- I don't know if it's relevant, but we had a copier that insisted it needed toner until the service tech. came out and cleaned the sensor. Apparently when toner was added the previous time, some spilled onto the sensor. ================================================================================ Note 455.7 LN03+ Toner Light 7 of 14 EISNER::ROECKEL "Bruce W. Roeckel" 13 lines 1-SEP-1989 10:52 -< Under Warranty? >- -------------------------------------------------------------------------------- > This condition has been going on for about a year but the > light is always on. I guess it's just stuck. If it ever > needs service I'll have it fixed at that time. Since I do > not have a maintenance contract on it I don't want to call > DEC Field Service just for this. DEC does have a one year warranty on all hardware purchases. Did you buy it directly from DEC? If so, the warranty should cover the fix, free of chardge to you. -Bruce ================================================================================ Note 455.8 LN03+ Toner Light 8 of 14 EISNER::MAYHEW "Bill Mayhew" 7 lines 1-SEP-1989 14:22 -< Good point! >- -------------------------------------------------------------------------------- > Did you buy it directly from DEC? Actually this shouldn't matter. If you bought it from any authorized sales channel (distributor, etc.) you should have the one-year warranty. (There are some *possible* exceptions but I've never run into one.) The warranty is probably return-to-factory, though, which may make the cure worse than the disease ;-}... ================================================================================ Note 455.9 LN03+ Toner Light 9 of 14 EISNER::CONROY "Alan Conroy" 8 lines 2-SEP-1989 05:04 -< One possibility >- -------------------------------------------------------------------------------- > I don't know if it's relevant, but we had a copier that insisted it > needed toner until the service tech. came out and cleaned the sensor. We had this exact problem with one of our LN03's about a year ago. CDC came out and vacuumed the sensor and the problem went away. It's been fine ever since... NOTE: If you decide to vacuum the thing yourself, do it CAREFULLY, I think it is on the fragile side... ================================================================================ Note 455.10 LN03+ Toner Light 10 of 14 EISNER::BURNET "John Burnet" 10 lines 4-SEP-1989 00:41 -< Toner-proof vacuum cleaner recommended for laser-printer service >- -------------------------------------------------------------------------------- > NOTE: If you decide to vacuum the thing yourself, do it CAREFULLY, I think it > is on the fragile side... Not only that -- be sure to use a vacuum cleaner with a toner-proof dust container. 3M, Hoover (I think), and a few other manufacturers make vacuum cleaners that are built specifically for use in servicing copiers and laser printers. If you vacuum toner with a normal vacuum cleaner, you may be surprised by the results. :-} ================================================================================ Note 455.11 LN03+ Toner Light 11 of 14 EISNER::MAYHEW "Bill Mayhew" 11 lines 4-SEP-1989 12:19 -< Why? >- -------------------------------------------------------------------------------- > If you vacuum toner with a normal vacuum cleaner, you may be surprised by > the results. :-} OK, I'll bite. A few months ago, I sucked up the last toner in the hopper of our LN03 with a garden-variety Kenmore household-type vacuum. No clouds of dust appeared out of the vacuum's back end... it still works fine... and the toner did indeed get sucked out of the hopper. What fun was I lucky enough to miss? ================================================================================ Note 455.12 LN03+ Toner Light 12 of 14 EISNER::BURNET "John Burnet" 11 lines 4-SEP-1989 20:36 -< a fine mess >- -------------------------------------------------------------------------------- > A few months ago, I sucked up the last toner in the hopper of our > LN03 with a garden-variety Kenmore household-type vacuum. No clouds > of dust appeared out of the vacuum's back end... it still works > fine... and the toner did indeed get sucked out of the hopper. > What fun was I lucky enough to miss? Exactly what you mentioned: clouds of toner dust in the vacuum cleaner's exhaust. The bag in your vacuum must have pores that are small enough to contain the toner particles. Lots of vacuums don't have such bags, so there's at least the potential for creating a real mess. I guess this also depends on how powerful the cleaner's motor is. ================================================================================ Note 455.13 LN03+ Toner Light 13 of 14 EISNER::HEROUX "Bob Heroux" 14 lines 5-SEP-1989 08:43 -< Where's the sensor??? >- -------------------------------------------------------------------------------- > DEC does have a one year warranty on all hardware purchases. Did > you buy it directly from DEC? If so, the warranty should cover > the fix, free of chardge to you. I did purchase the LN03+ from a distributor which should not affect the one year warranty from DEC. (It's really been about 13 months anyway.) As far as I know it was a return to factory deal which wasn't worth the trouble or the loss of use for whatever time it takes. PS: Can anyone tell me where the sensor is so that I can try to clean it. ================================================================================ Note 455.14 LN03+ Toner Light 14 of 14 EISNER::BURNET "John Burnet" 8 lines 5-SEP-1989 21:52 -< Get out the plastic gloves >- -------------------------------------------------------------------------------- > PS: Can anyone tell me where the sensor is so that I can try to > clean it. Ouch. Have fun. It's underneath all that toner. 8-} The sensor consists of two or three metal pieces and a small spring, located at the bottom of the toner reservoir about a third or half the way across the reservoir from the left side.  ================================================================================ Note 456.0 MVII - Machine check in Kernel mode 4 replies EISNER::GRATTAN "Michael Grattan (Who #316)" 37 lines 5-SEP-1989 10:57 -------------------------------------------------------------------------------- We have had a serious problem with our uVAX II crashing regularly (every day) since May. (The actual crash is "Machine check while in kernel mode.") As it turns out there are a couple of known problems with MicroVAX II's. 1) Apparently, the cable that connects the CPU board to the two memory expansion cards must be shielded. According to DEC: "signal crosstalk is causing a memory parity error to be selected intermittently, resulting in an apparent parity error, cauing the system to crash. Error log entries reflect a Machine Check 80, 81,or 82 and an MSER of 341." The replacement memory cable is part number 17-02174-01. 2) #1 above, didn't solve my problem, but I thought that I would pass it along. After much frustration, DEC finally traced the problem to old versions of the 8Mb boards. (we are running VMS 5.1 with 16Mb) The version of memory boards that solve the problem are version A2. (Also, please note that some of the boards are mislabeled and are not A2 boards. True A2 boards have terminating resistors for data lines 15 and 16. Our service rep. had to get a guru to double check the boards we got from the factory. The first set was bad.) Since the new boards have been installed, we have been stable for several weeks. Hope this keeps someone else from having this nightmare... Of course, there's a 'gotcha'. One of the boards taken out of my machine is from Clearpoint. Do third party vendors get or have access to up-to-date technical information from DEC? The person to whom I spoke at Clearpoint didn't seem to know anything about terminating resistors. (Neither do I for that matter.) Since DEC is 'hinting' that they'd like one of their boards back, I would like to be sure that Clearpoint has the latest information. Thanks, Mike Grattan ================================================================================ Note 456.1 MVII - Machine check in Kernel mode 1 of 4 EISNER::MCCORMICK_B "Bob McCormick" 24 lines 5-SEP-1989 17:43 -< Is it live or is it DEC memoreyx >- -------------------------------------------------------------------------------- Thanks Mike for the info... no matter how horrible it sounds! | 1) Apparently, the cable that connects the CPU board to the two memory | expansion cards must be shielded. According to DEC: | | "signal crosstalk is causing a memory parity error to be selected | intermittently, resulting in an apparent parity error, cauing the | system to crash. | | Error log entries reflect a Machine Check 80, 81,or 82 and an MSER | of 341." | | The replacement memory cable is part number 17-02174-01. Boy, I've had my hands in over 40 MicroVAX II's all over the continent since they were announced and never saw a shielded memory interconnect cable... Do you know if the part noted above is provided on demand as an ECO/FCO? I've got a 16mb (two DEC MS650-CA boards) BA123 system, 13 months old, running V5.2 in my office. I'll pull it apart and see if the parts and revs you note apply ... but I have not seen any problems, either. Do you know where the two resistors are located on the board (etch location?) ================================================================================ Note 456.2 MVII - Machine check in Kernel mode 2 of 4 EISNER::KILGALLEN "Larry Kilgallen" 2 lines 5-SEP-1989 18:01 -< You are not alone >- -------------------------------------------------------------------------------- I had such a cable replacement gratuitously 3 months ago when a field servant was in on an unrelated problem. ================================================================================ Note 456.3 MVII - Machine check in Kernel mode 3 of 4 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 3 lines 6-SEP-1989 12:00 -< DEC Field Service on top of it ! >- -------------------------------------------------------------------------------- My PMI (Private Memory Interconnect) cable was upgraded also during an unrelated service call. This definitely is an "across the board" FCO along with changing the power cable on the BA123. ================================================================================ Note 456.4 MVII - Machine check in Kernel mode 4 of 4 EISNER::GRATTAN "Michael Grattan (Who #316)" 21 lines 8-SEP-1989 14:55 -< A fix! A fix! >- -------------------------------------------------------------------------------- Just as a follow-up, I thought that I would pass this along. Two people from Clearpoint came to my site to exam the DEC memory boards and look into my problem. One was a board designer who seemed very knowledgeable about DEC products. (I was impressed!) He examined the A2 rev DEC boards and noted the location of the termination resistors. According to him, they are there to eliminate line noise problems on the board-level bus and are strictly a DEC phenomena. (BTW, if anyone would like I will be happy to FAX a "copy" I took of my boards which show the resistor locations.) At this time it looks like my system was crashing due to a bad batch of DRAM chips on the Clearpoint board. Clearpoint will fix this problem. It turns out that they have had some pretty sever problems with this chip vendor and they no longer get chips from them. From what I now understand, if you have DEC memory boards and you have this problem, it's probably due to the termination resistors (the lack thereof..). If you have third party boards it might be the chips. I'll post a note when I get the new Clearpoint board.  ================================================================================ Note 457.0 The future of DSSI 5 replies EISNER::KILGALLEN "Larry Kilgallen" 16 lines 5-SEP-1989 11:15 -------------------------------------------------------------------------------- > <<< EISNER::DUA0:[NOTES$LIBRARY]LOCAL_HAPPENINGS.NOTE;1 >>> > -< LOCAL_HAPPENINGS >- >=============================================================================== >Note 5.31 NY Metro LUG 31 of 31 >EISNER::TINKELMAN "Bob Tinkelman - CCA" 19 lines 25-AUG-1989 13:46 > -< 08/29/89 Meeting notice >- >------------------------------------------------------------------------------- > > Date: TUESDAY, AUGUST 29, 1989 at 5:30 pm > > ** RICH WRENN of DEC will speak on DSSI VAX clusters. Learn all about > the DSSI bus: how it allows dual-access to disks from two microVAXes, > why the limit is 8 nodes per DSSI, why the limit is 2 CPUs per DSSI, > and how likely it is any of these will change. So what were the answers? ================================================================================ Note 457.1 The future of DSSI 1 of 5 EISNER::TINKELMAN "Bob Tinkelman - CCA" 26 lines 6-SEP-1989 18:29 -< Rumors, rumors and more DSSI rumors >- -------------------------------------------------------------------------------- > ** RICH WRENN of DEC will speak on DSSI VAX clusters. Learn all about > the DSSI bus: how it allows dual-access to disks from two microVAXes, > why the limit is 8 nodes per DSSI, why the limit is 2 CPUs per DSSI, > and how likely it is any of these will change. The eight-node per DSSI limit is architectural. But you can have more than one DSSI per machine. You can have only 2 CPUs per DSSI due to current packaging. (Boxes which house CPUs don't have an "in" and "out" for the DSSI, CPU integrated DSSI interfaces are terminated, ...) Some restrictions, such as clustering with different DSSI adapters on the 2 CPUs are really software restrictions, related to the way that the disks are named by VMS (and the requirement that all nodes on the cluster use the same name to refer to the same disk). These will be eased (solved?) with VMS 5.3. V5.3 (I think) will also support clusters with more than one DSSI. Currently multiple DSSIs are supported only on nonclustered systems. Starting with V5.2, both host-disk and host-host cluster traffic will run over the DSSI (with the integrated adapters). Currently the host to host stuff (eg lock manager) runs over the Ethernet. Future microVAXes will all support >2 CPUs per DSSI. Probably more such stuff. All rumors until official release dates. ================================================================================ Note 457.2 The future of DSSI 2 of 5 EISNER::KILGALLEN "Larry Kilgallen" 7 lines 7-SEP-1989 09:54 -< BA123 is flexible >- -------------------------------------------------------------------------------- > current packaging. (Boxes which house CPUs don't have an "in" and "out" > for the DSSI, CPU integrated DSSI interfaces are terminated, ...) This would seem to be a KFQSA design flaw, since the MicroVAX II (for example) was designed long before DSSI. It's box has general purpose bulkhead openings. Was he indicating that the KFQSA has only one set of wires? ================================================================================ Note 457.3 The future of DSSI 3 of 5 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 15 lines 7-SEP-1989 10:26 -< ? >- -------------------------------------------------------------------------------- ! -< BA123 is flexible >- ! ! > current packaging. (Boxes which house CPUs don't have an "in" and "out" ! > for the DSSI, CPU integrated DSSI interfaces are terminated, ...) ! ! This would seem to be a KFQSA design flaw, since the MicroVAX II ! (for example) was designed long before DSSI. It's box has general ! purpose bulkhead openings. Was he indicating that the KFQSA has ! only one set of wires? To my knowledge *NO* DSSI disk can mounted in a BA23/BA123 box. DEC has a BA213/214 based tower that you place next to your BA123 and mount the DSSI disks in. A cable set from the KFQSA runs out of the BA123 into the BA213/214 tower. ================================================================================ Note 457.4 The future of DSSI 4 of 5 EISNER::KILGALLEN "Larry Kilgallen" 8 lines 7-SEP-1989 13:41 -< KFQSA cables must be "in and out" >- -------------------------------------------------------------------------------- Yes, but to put more than three CPUs on a DSSI, as was indicated, you need to run the DSSI cable "in and out" of the KFQSA and the containing cabinet. That is the limitation which I believe was described. Whatever they do in the way of bulkhead connectors on the BA123 to connect the KFQSA to the disks they can do twice (so far as the BA123 and the FCC are concerned) since both shapes of cutouts come in multiple instantiations. So is the "in and out" limitation on the KFQSA board itself? ================================================================================ Note 457.5 The future of DSSI 5 of 5 EISNER::TINKELMAN "Bob Tinkelman - CCA" 14 lines 8-SEP-1989 01:03 -< There's a 4" limit >- -------------------------------------------------------------------------------- RE: < Note 457.4 by EISNER::KILGALLEN "Larry Kilgallen" > -< KFQSA cables must be "in and out" >- > So is the "in and out" limitation on the KFQSA board itself? I believe there is only a single connector on the KFQSA. The slides indicated that there was a "T" off the DSSI to each "drop". The specs limit the length of this "T" to 0.1 meter (4 inches). My impression was that the problem was that you couldn't get from the bulkhead to the KFQSA in 4". This it's a problem with the design of the boxes into which you put Q-buses. [I feel really uncomfortable as I'm doing this from memory and might be getting it wrong. Can anyone confirm or correct me here?]  ================================================================================ Note 458.0 ANA/DISK/REP Error 1 reply EISNER::KRULEWICZ 16 lines 5-SEP-1989 22:21 -------------------------------------------------------------------------------- Ok, I typed ANA/DISK/REP DUA0: and the system responded: %VERIFY-W-ALLOCCLR, blocks incorrectly marked allocated LBN 15975 to 16292, RVN 1 %VERIFY-E-SORTERROR, error sorting multiply allocated block list -RMS-E-WLK, device is write locked $ Now what? BTW, I don't think the device is write locked... - Jerry ================================================================================ Note 458.1 ANA/DISK/REP Error 1 of 1 EISNER::BURNET "John Burnet" 35 lines 5-SEP-1989 22:44 -< Just define SYS$SCRATCH to point to some other disk >- -------------------------------------------------------------------------------- This is a very frustrating misfeature of VMS V5.0. Yes, the device is in fact software write-locked; the VERIFY program very kindly does it for you while it's analyzing the disk. (That in itself is a very good idea, if not an essential step in the analysis/verification; but the fact that VERIFY needs scratch space to sort the multiply-allocated block list was apparently overlooked by the developers.) If you have another disk (*any* other disk -- even a floppy drive) on the system, make sure that you have write access to some directory on that disk, and do this immediately before the ANALYZE/DISK/REPAIR command: $ define/user sys$scratch some_disk:[some_directory] If you don't have any other disk on the system, and if your system is a network node, make sure that you have write access to some directory on some disk on some node in the network, and do this: $ define/user sys$scratch some_node::some_disk:[some_directory] before the ANALYZE/DISK/REPAIR. If you are the proud owner of a standalone, single-disk system, congratulations. You get to edit SYS$UPDATE:STABACKIT.COM and pull out SETSIZE.MAR, assemble and link it, and (with the help of the DEC-supplied PDDRIVER.EXE) set yourself up with a (rather inefficient) RAM-disk. Details of that procedure, and complete instructions, should be somewhere on this system or on the Pageswapper system. Once you have the RAM-disk set up (you may need to increase NPAGEDYN using Autogen and reboot in order to get enough nonpaged pool for the memory disk), you can do: $ allocate pda0: $ init pda0: whosis $ mount pda0: whosis $ define/user sys$scratch pda0:[0,0] $ analyze/disk/repair dua0: ! or whatever  ================================================================================ Note 459.0 HSC 3.90-VMS BACkUP PROBLEM 4 replies EISNER::JENSEN_M 11 lines 6-SEP-1989 00:43 -------------------------------------------------------------------------------- Not sure if this is right conference, but here goes. We have encountered a problem with an HSC70 running 3.90 code. The problem is crunching a disk when VMS BACKUP is running and a BBR (bad block replacement) is called for. The problem occurs on a non-DEC Disk drive, but the Field Service rep indicates that any disk is suseptible(sp) to this bug. He also indicates that HSC 3.94 code may fix it. Anyone else experience this problem ? ================================================================================ Note 459.1 HSC 3.90-VMS BACkUP PROBLEM 1 of 4 EISNER::FULLER_S "Funky cold medina" 14 lines 6-SEP-1989 15:48 -------------------------------------------------------------------------------- >Not sure if this is right conference, but here goes. We have encountered >a problem with an HSC70 running 3.90 code. The problem is crunching a disk >when VMS BACKUP is running and a BBR (bad block replacement) is called for. First, define "crunching". Does the disk report a fault? Does the error log indicate a bad block replacement has occured? What error messages do you get? Does the disk start making strange (crunching?) noises? There is a known problem with the HSC code that will cause it to crash if it has to do BBR on a shadow set, but there is a patch for it available from the CSC. Stu ================================================================================ Note 459.2 HSC 3.90-VMS BACkUP PROBLEM 2 of 4 EISNER::JENSEN_M 1 line 6-SEP-1989 22:55 -< More info >- -------------------------------------------------------------------------------- ================================================================================ Note 459.3 HSC 3.90-VMS BACkUP PROBLEM 3 of 4 EISNER::JENSEN_M 15 lines 6-SEP-1989 23:06 -< more information >- -------------------------------------------------------------------------------- No "new" loud noises. The crunch is in the bits. The FE indicated that critical areas of the disk were corrupted - FCT,etc. - he didn't specify. Attempting to reformat with the HSC formatter was unsucessful. New information received today from the FE is that 3.94 is really 4 patches. One of the 4 deals with the BBR/BACKUP issue. Others deal with the TA90, etc. A floopy was delivered today by the FE containing patches thru 94. It will be installed tonight via cold start of the HSC. Mike ================================================================================ Note 459.4 HSC 3.90-VMS BACkUP PROBLEM 4 of 4 EISNER::JENSEN_M 14 lines 7-SEP-1989 21:50 -< An update... >- -------------------------------------------------------------------------------- An update... Colorado indicates that .94 is indeed focused at bad blocks with shadow sets. Today the FE cleared one or more sectors of COPY 1 of the FCT. He then sucessfully formatted the drive, verified it, and ran ILEXR on it for 2 hours. He proclaimed it healed and turned it over to us. We'll see what happens with the drive in the next couple of days. His expertise was recently acquired at a Level 2 class on MSCP,HSC,SDI etc. He also had a notebook from the class with a substantial amount of information on all those topics.  ================================================================================ Note 460.0 Soon to be Obsolete Hardware? 9 replies EISNER::FERGUSON "Linwood Ferguson, ARA Services" 49 lines 6-SEP-1989 07:47 -------------------------------------------------------------------------------- There are some other topics where individual aspects are discussed but I thought I would start a new topic for the overall problem. At least we think it is a problem. We're hearing that alot of hardware we have will not longer be supported in the near future, and cannot get a straight answer from Dec even as to whether it is true, much less whether they are going to do anything to help $$-wise if so. Here's the list we have so far. Note that most of this is rumor or implication, but some fairly substantial rumor. I would be interested if anyone else is concerned that some fairly new hardware is being obsoleted, and whether you have had any luck holding Dec responsible for the problem. Also, if you know any of these are NOT true, let me know so I can stop worrying. If you know any others, maybe we can build a full list. - DEQNA support dropped in near future (rumored 1 year support past V5 is up). This was a formal Dec announcement. Dec apparently has no program for upgrades, at least at present (I don't consider "buy a new DELQA" a program for upgrading). - DHU-11 support on Unibus on BI bus: also announced with V5's release, I've heard nothing else about this. I also cannot tell whether this is Decnet-only (i.e. if used for async Decnet, which we don't) or all usage. - (I don't consider this a reliable rumor) Last week a Dec "support specialist" our Dec folks brought in to address the above two issues said that DHU-11 support was being dropped on everything, and that Dec's answer (in regard to our 785's) was to upgrade to a 3400 or 6000 machine. (we just kind of stopped listening to her at that point). - Phase V's preliminary literature says DMV-11's will not be supported. We have a LOT of those since DSV11's didn't exist early-on. Dec's answer so far is "replace with DSV11's", but again no upgrade program (yet?). - Recurring rumor: I hear long ago, and keep hearing occasionally again that Dec plans to drop Multipoint Decnet support. For those of us unlucky enough to use Multipoint all over the country, it will be VERY expensive to change (and not in Dec hardware - which Dec might like - but in ATT or other carrier charges). I have not heard this from Dec at all. Non-Hardware concerns: - I have been told that Phase V requires Distributed Name Service. That's a product I have no other use for, and it is not cheap. I have not been able to find out if we can avoid going to Phase V while continuing to stay current with VMS, nor have I been able to confirm if DNS is really needed and/or included. ================================================================================ Note 460.1 Soon to be Obsolete Hardware? 1 of 9 EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 2 lines 6-SEP-1989 11:40 -< I'd wager DNS will be included >- -------------------------------------------------------------------------------- If phase V required DNS, I can't believe DEC won't include it. However, I can here the screeching of all those who purchased DNS. ================================================================================ Note 460.2 Soon to be Obsolete Hardware? 2 of 9 EISNER::MCMICHAEL "Chuck McMichael" 14 lines 6-SEP-1989 11:57 -< DECnet-VAX Cover Letter V5.2 >- -------------------------------------------------------------------------------- > DEQNA support dropped in near future The DECnet-VAX Cover Letter (AV-DL04H-TE) for VMS 5.2 states "In the next 9 to 15 months...software support for the DEQNA Ethernet adapter will be withdrawn. For 24 months after this software support is withdrawn, the DEQNA will be accessible only by user applications using the $QIO interface to the Q-bus Ethernet device driver (XQDRIVER). > DHU-11 support on Unibus on BI bus: also announced with V5's release, The DECnet-VAX Cover Letter (AV-DL04H-TE) for VMS 5.2 states that prior DECnet-VAX cover letters were incorrect on the subject of the DHU11 and the DMZ32. "Please ignore it, and note that these adapter configurations are still supported." ================================================================================ Note 460.3 Soon to be Obsolete Hardware? 3 of 9 EISNER::HASSINGER "Bob Hassinger" 13 lines 6-SEP-1989 16:10 -< More fuel for DNS speculation (& D/N phase V futures) >- -------------------------------------------------------------------------------- Note too on page 3-46 of the VMS V5.2 Release Notes, section 3.17: "A number of files in the VMS Version 5.2 kit refer to a system service $DNS, a feature that may be included in a future release of VMS. Use of these files under VMS V5.2 is not supported. Under VMS Version 5.2, the DNS$ system service will return SS$_UNSUPPORTED to all calls. "Note: The Remote System Manager and VAX Distributed File Service products contain files with similar names. The files mentioned above will not interfere with the operation of these products." What should we make of this? ================================================================================ Note 460.4 Soon to be Obsolete Hardware? 4 of 9 EISNER::KILGALLEN "Larry Kilgallen" 5 lines 6-SEP-1989 17:38 -< No speculation involved >- -------------------------------------------------------------------------------- It was announced at the Atlanta Symposium that SYS$DNS was the system service to be used in the future for calling the local DNS clerk. At the time, they said that sites which had DNS would be able to use it in VMS V5.2. Obviously that changed, but sometimes it is good that they release no software before its time. ================================================================================ Note 460.5 Soon to be Obsolete Hardware? 5 of 9 EISNER::LEDERMAN "Bart Z. Lederman" 3 lines 6-SEP-1989 20:27 -< Problems with DNS >- -------------------------------------------------------------------------------- Also note that if you are using DNS now there is a rather persistant FLASH on DSIN that says they have some big problems with it now (like crashes). ================================================================================ Note 460.6 Soon to be Obsolete Hardware? 6 of 9 EISNER::MERRIMAN "George Merriman - CCA/NY" 8 lines 6-SEP-1989 20:50 -< No software, please, we're hardware weenies >- -------------------------------------------------------------------------------- It seems to me that this topic started out as a DEC business practices question and has deteriorated to a "did you hear the latest rumour about topic, where has little to do with hardware. I will suggest that the BUSINESS_PRACTICES moderators "import" the .0 note so you can all carry on there. In the mean time, please remember this is HARDWARE_HELP. Thanks. George Merriman, HARDWARE_HELP co-moderator. ================================================================================ Note 460.7 Soon to be Obsolete Hardware? 7 of 9 EISNER::MCCORMICK_B "Bob McCormick" 20 lines 8-SEP-1989 13:11 -< DMV11 --> DSV11 >- -------------------------------------------------------------------------------- DMV11's utilize DDCMP all done in microcode, part of the architecture of the module. I field tested this things _years_ ago! DSV11's get downline loaded when you do the SYSGEN AUTOCONFIG/ALL. Today they speak real nice to DMV11's and the like. If you look on a DSV11 module you'll see a chip that has what looks like an Ethernet address ... why on a WAN sync board, you say? Well for Phase V that's the kind of architecture changes they're making! This is detailed in the various specs, which have been available at DECUS, or you can purchase. Sorry to say, but I believe that DEC did clearly indicate the DMV11's would only be Phase IV compliant, and that the DSV would be the answer for today and upgradable to tomorrow. There's a big price difference, and I know of the lead time problems, too. As for the DNS stuff (not hardware, but I'll tack it on here...) if you look at your VMS V5.2 mandatory update tape, one of the things it does is DISABLE the DNS provided with V5.2 ... nothing before its time! ================================================================================ Note 460.8 Soon to be Obsolete Hardware? 8 of 9 EISNER::TINKELMAN "Bob Tinkelman - CCA" 18 lines 9-SEP-1989 20:54 -< DNS, DECnet Phase V, and DNS servers >- -------------------------------------------------------------------------------- > I have been told that Phase V requires Distributed Name Service. That's > a product I have no other use for, and it is not cheap. I have not been > able to find out if we can avoid going to Phase V while continuing to > stay current with VMS, nor have I been able to confirm if DNS is really > needed and/or included. The way I remember it being explained at the recent Symposia, the client side of DNS (called the clerk) resides on all the Phase V DECnet nodes and will be bundled into DECnet. It's the server side (of which you need at least one per network, more for fault tolerance, etc.) that's the layered product. And here, the feeling I get is that you may run it initially on one of your existing machines, like you can run DECnet routing on "one of your time sharing VAXes", however as time goes on, just as networks are moving to using dedicated routing boxes, there will be dedicated DNS server boxes. [I mentioned "boxes". So this *is* a HARDWARE_HELP topic after all! :-) ] ================================================================================ Note 460.9 Soon to be Obsolete Hardware? 9 of 9 EISNER::FERGUSON "Linwood Ferguson, ARA Services" 28 lines 10-SEP-1989 10:25 -< There was no DSV11 when we got the DMV11's >- -------------------------------------------------------------------------------- >Sorry to say, but I believe that DEC did clearly indicate the DMV11's >would only be Phase IV compliant, and that the DSV would be the >answer for today and upgradable to tomorrow. There's a big price >difference, and I know of the lead time problems, too. If my poor memory serves, we ordered these in spring of '88. At that point there was no such thing (at least announced) as a DSV11 for 3xxx machines. We had to have something for WAN or we would not order the 7 3xxx machines, so Digital proudly offered the DMV11's to us, I believe a bit before they were even announced. No indication they were not THE reengineered WAN product for the 3xxx. They were announced, and for a period of months were the only things avaialble. And at no time during that period did anyone with Dec (or anyone else for that matter) suggest they were anything but fine. Then the DSV11 came out a few months later (around fall Decus I think) and we were told the DMV11 was a rushed substitute because the "real" WAN product, the DSV11, was late in being released. Great. The DMV11-SA was a freshly reengineered product (we were assured over and over that No, we could not use an old DMV11 on the 3xxx). "Everything reworked to make it compatible and more efficient (never explained) with the 3xxx machines". Dec knew it would be obsolete this soon, why didn't they say? The answer to that is obvious: they wanted to see seven 3xxx machines! Sorry, Sir Moderator, maybe this is getting into Business Practices. But it seems you cannot worry purely about hardware with Digital without getting tangled up in their Business Practices.  ================================================================================ Note 461.0 TK50 "Volume not software enabled" ?? 2 replies EISNER::BARASCH 17 lines 7-SEP-1989 16:36 -------------------------------------------------------------------------------- I'm seeing same TK50 problem reported in MICROVAX_SYSTEMS 73.0 which was %INIT-F- Volume not software enabled. (TK50 is on uvax2) This problem shows up now and then. First time it happened I simply had FS replace TQK50, next time a drive. This time however I was prudent and just bounced power, which seemed to clear up problem. Probably would have saved time doing this in 1st place. ;-) If this is a software problem, I'd sure like a fix. If it's a media problem, why does it persist to other tapes? I'm running 5.1-1. ANAL/ERROR shows "EDGES OF CAL TRK 1 OUT OF SPEC" Clues, anyone? ================================================================================ Note 461.1 TK50 "Volume not software enabled" ?? 1 of 2 EISNER::TERRAZAS "Mike Terrazas" 7 lines 7-SEP-1989 17:05 -< Probably doesn't apply >- -------------------------------------------------------------------------------- I know this isn't YOUR problem, but every time one of our international offices calls in with this problem, it's the red button wasn't pushed in, or there was an DISMOUNT that didn't specify /NOUNLOAD and then the red button not cycled. It's always seemed strange to me that this had anything to do with software at all. ================================================================================ Note 461.2 TK50 "Volume not software enabled" ?? 2 of 2 EISNER::BARASCH 0 lines 7-SEP-1989 19:21 -< You're right, it doesn't :-) >- --------------------------------------------------------------------------------  ================================================================================ Note 462.0 Cpu to HSC without star coupler? 4 replies EISNER::GRAUE "Geoff, Security Pacific Bank" 20 lines 8-SEP-1989 23:44 -------------------------------------------------------------------------------- Can anyone shed some light...? We want to connect our 6310 directly to an HSC70. Since we do not, at this time, have any other CI-based systems in this cluster, (this is a separate system used just for development) , we would like to avoid having to buy a star coupler just to connect one cpu to one HSC (we had a spare HSC not being used and our 6310 does not have any KDB hardware). Originally, Field Service told us this would work. We tried it and the HSC refused to boot. We asked Field Service what the problem was and they said that they had seen this work with a 780, so the problem must be with the newer type of CI. First, has anyone made this work? Second, if a star coupler is needed, does anyone know why it would be needed in this configuration? I thought it was a passive device and was invisible to any hardware or software. ================================================================================ Note 462.1 Cpu to HSC without star coupler? 1 of 4 EISNER::COY "Dale E. Coy" 25 lines 8-SEP-1989 23:59 -< Passive -- but not invisible >- -------------------------------------------------------------------------------- > Second, if a star coupler is needed, does anyone know why it would > be needed in this configuration? I thought it was a passive device > and was invisible to any hardware or software. Somebody will show up with an answer to why/why not/how. In the meantime - A star coupler is a passive device - primarily consists of a fancy set of coupling transformers. Just because it's passive, that doesn't mean it's invisible. Example: If you take your 110-volt appliance (say a hair dryer) to europe, and plug it into their usually-220-volt outlets, it's not going to work very well (or very long). What you need is some kind of PASSIVE DEVICE (a transformer, for example) to convert 220 to 110. (possibly applicable) Example: hooking two things together through a transformer allows the AC to pass, but does not allow the DC to pass. If it turns out that you do need a Star Coupler, and you want to do it "cheap", and you don't want beauty - you could order just the Star Coupler upgrade kit that adds additional "ports". Fundamentally, the upgrade kit is a Star Coupler without a cabinet. ================================================================================ Note 462.2 Cpu to HSC without star coupler? 2 of 4 EISNER::FULLER_S "Funky cold medina" 20 lines 9-SEP-1989 14:31 -< Won't work - here's why >- -------------------------------------------------------------------------------- This configuration will not work. The HSC, in its self test, wants to talk on the CI and see its responses. If it's connected directly to the 6310, it'll never see its responses, and will therefore assume that the HSC's CI port is broken. The CI port on the VAX will be the same way. Also, the SC (Star Coupler) provides attenuation of the signal. If you connect the XMIT directly to the RECV ports, the RECV ports will be "bombarded" with a much larger signal than they were designed to accept, and may fail quickly. As Dale suggested, you could possibly use just the SC panels, and not use the cab. Just remember to use the plastic mounts for the panels. Where you put them is up to you - try the dead space in the bottom of the 6310 or the top of the HSC. No guarantees of whether it'll work (reliably) or not, but if your primary interest is cost savings at the possible expense of reliability... Stu ================================================================================ Note 462.3 Cpu to HSC without star coupler? 3 of 4 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 5 lines 11-SEP-1989 10:48 -< dead to you, maybe >- -------------------------------------------------------------------------------- > ... try the dead space in ... the top of the HSC. But if you do that, where will FS store all its tapes, manuals, lunchboxes, etc.? :-) ================================================================================ Note 462.4 Cpu to HSC without star coupler? 4 of 4 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 1 line 11-SEP-1989 15:11 -< Stu goes OUT to lunch ! >- -------------------------------------------------------------------------------- sorry I couldn't resist.  ================================================================================ Note 463.0 SC02 and TC02 from PDP to MVII? 1 reply EISNER::KOZAM 11 lines 10-SEP-1989 14:24 -------------------------------------------------------------------------------- Has anyone used an Emulux SC02 controller with a MVII running VMS? The SC02 hooks up to some CDC drives and emulates a string of RK06/RK07 drives for a total of about 240MB. The SC02 and drives come off a PDP that is being decommissioned, and I'd like to move the drives and controller to a MVII that is pressed for disk space. The plan would be to bind all drives into a volume set, and thus obtain a decent amount of storage space. While I'm at it, is anyone using an Emulux TC02 with Cipher F880 on a MVII? It too comes from the PDP. ================================================================================ Note 463.1 SC02 and TC02 from PDP to MVII? 1 of 1 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 17 lines 10-SEP-1989 22:40 -< FROM 11 LAND - THERE ARE PROBLEMS >- -------------------------------------------------------------------------------- > Has anyone used an Emulux SC02 controller with a MVII running > VMS? The SC02 hooks up to some CDC drives and emulates a string of > RK06/RK07 drives for a total of about 240MB. You have a problem with SC02's on PDP-11's with more than a 1/4 meg that I assume will effect MV's. DEC never produced a Q-BUS version of the RK6/7 controller. Therefore DEC assumes 18 bit addressing for RK6/7. There is a spare register where you can place the additional address bits - but that means a driver modification. > While I'm at it, is anyone using an Emulux TC02 with Cipher F880 > on a MVII? It too comes from the PDP. There is an emulation problem with the TC02 before REV J. I have heard this caused a problem with V5.  ================================================================================ Note 464.0 Need Talaris EXCL help 1 reply EISNER::DENOVO "Jay DeNovo" 4 lines 11-SEP-1989 11:06 -------------------------------------------------------------------------------- Has anyone worked with the Talaris EXCL printer language to create forms in the VAX/VMS environment? Specifically, I'd like to know about using overlay forms into which data can be merged at print time. Reply here, or call me directly at 608-266-8902. Thanks! ================================================================================ Note 464.1 Need Talaris EXCL help 1 of 1 EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 2 lines 11-SEP-1989 14:45 -< Consider cross posting >- -------------------------------------------------------------------------------- Personally, Jay, I would pose this question in DESKTOP_PUBLISHING or 3RD_PARTY_VMS_SOFTWARE.  ================================================================================ Note 465.0 VS3100 and the graphics decelerator No replies EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 5 lines 11-SEP-1989 20:54 -------------------------------------------------------------------------------- We have several VS3100s. Some run VWS, some run DECwindows. We have found that putting in the graphics accelerator card from DEC really helps the VWS machines. The DECwindows machines, however, *slow up* when the "accelerator" (oops, I should say co-processor), is used. DEC hardware folks can't explain it yet.  ================================================================================ Note 466.0 TK70 can read TK50, but can it write? 3 replies EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 4 lines 13-SEP-1989 12:48 -------------------------------------------------------------------------------- We all know the party line: TK70s can't write TK50 compatible tapes. Is this *really* true? Could there be some secretive jumper that could be inserted/deleted to make this go? Has anyone *tried* to make this work? ================================================================================ Note 466.1 TK70 can read TK50, but can it write? 1 of 3 EISNER::HASSINGER "Bob Hassinger" 6 lines 13-SEP-1989 17:38 -< "No" - "different head geometry" >- -------------------------------------------------------------------------------- I asked a DECie who knows. He said no and he reminded me of an underlying answer that explains at least part of the reason. You come across the same thing in a lot of these cases. It has to do with reduced head geometry to get the higher density, reduced magnetic flux strength on write, resulting inability to write a tape a TK50 can read, etc... ================================================================================ Note 466.2 TK70 can read TK50, but can it write? 2 of 3 EISNER::OSUDAR "John Osudar" 0 lines 13-SEP-1989 18:23 -< That's OK, my TK50's often write tapes a TK50 can't read :-) >- -------------------------------------------------------------------------------- ================================================================================ Note 466.3 TK70 can read TK50, but can it write? 3 of 3 EISNER::TERRAZAS "Mike Terrazas" 9 lines 14-SEP-1989 09:15 -< Geometry notwithstanding... >- -------------------------------------------------------------------------------- re: .1 Yeah, yeah, That's what IBM said when they came out with the 1.2 meg 5.25" diskette drives. But for non-critical copying, we all have done it one way or another. The problems that arose were usually related to the use of a previously used diskette. I agree that it would be helpful to find a way (modified driver?, new ROM in the '70?) to be able to write a TK50 format tape with a TK70. Reliability could not be expected, but it may not be an issue.  ================================================================================ Note 467.0 Exabyte 8mm drive on VAXstation 3100 SCSI? 1 reply EISNER::BURNET "John Burnet" 18 lines 13-SEP-1989 21:44 -------------------------------------------------------------------------------- A couple of questions on SCSI cabling from a hardware illiterate: We've got a couple of VAXstation 3100s (SCSI/ST506 with disks but no tape drives), and also an SI59 8mm drive on an 11/750 (Exabyte drive, of course, and SI SCSI-to-Unibus adapter). 1. If we were to hang the Exabyte drive on a VAXstation's SCSI (assuming that the drive's SCSI ID is set to whatever the VAXstation wants), would the VAXstation recognize it as a TK50 (or TZ30) the way the 750 does? 2. The reason we haven't tried this yet is that the cables are incompatible. I'm not at work and I've forgotten the details, but I think the Exabyte drive's cable has a 50-pin IDC connector on each end, while the VAXstation uses one smaller IDC connector on the VAX end and a Centronics/telco-type connector at the drive. Question: Would an adapter cable work, or are there electrical differences as well as connector differences? Thanks for any and all help. ================================================================================ Note 467.1 Exabyte 8mm drive on VAXstation 3100 SCSI? 1 of 1 EISNER::CHADWICK "Keith Chadwick" 13 lines 14-SEP-1989 11:33 -< I don't beleive that this works yet >- -------------------------------------------------------------------------------- >1. If we were to hang the Exabyte drive on a VAXstation's SCSI (assuming > that the drive's SCSI ID is set to whatever the VAXstation wants), would > the VAXstation recognize it as a TK50 (or TZ30) the way the 750 does? The rumor that I have heard is that the VAXstation will not recognize the Exabyte until VMS V5.3. According to the VMS V5.2 release notes, third party disks (such as the CDC WREN V) are now able to be served to the cluster (there also was a patch to DKDRIVER that was posted on INFO-VAX distribution list, which performs the same function). -Keith.  ================================================================================ Note 468.0 VAX 6000 Memory Card Count 7 replies EISNER::BRUNS "Alan Bruns" 16 lines 14-SEP-1989 14:21 -------------------------------------------------------------------------------- I'm looking at the probable necessity of adding memory to my VAX 6000 next year. From the owner's guide I get the impression that they really want you to have either 1, 2, 4, 6, or 8 32 Mb boards. It says that this is because interleaving (only possible with these configurations?) gives faster access time. That's greek to me. I remember hearing something like that (having to add memory in pairs of boards) at the 6000 announcement, so apparently DEC means it... I've already got 64 Mbytes (2 boards). I don't really think I'd have any use for more than an additional 32 Mbytes for at least a year, and at DEC's prices (~$25K/board) I'm not really eager to add a bunch of capacity that I won't use. Just what do I lose by *only* buying one 32Mbyte board? VPA says my peak CPU usage is averaging in the low 30s, so I've got lots of cycles to burn. ================================================================================ Note 468.1 VAX 6000 Memory Card Count 1 of 7 EISNER::HANRAHAN "Jamie Hanrahan, jeh@crash.cts.com" 23 lines 14-SEP-1989 15:38 -< memory interleaving explained. >- -------------------------------------------------------------------------------- > ...interleaving (only possible with these > configurations?) gives faster access time. That's greek to me... From the VAX Hardware Handbook, 1982-1983 edition (I will give it up when they pry it from my, etc.), page 243: "Interleaving improves memory subsystem throughput on the bus... "...most memory operations are performed on consecutive memory locations. While one controller is fetching data, the other controller is available to decode an address for the next operation.... on the 11/780, the two memory controllers access alternate quadwords." If I'm not mistaken, in a 6000-series machine, each memory board is its own controller. > I've already got 64 Mbytes (2 boards). I don't really think I'd have > any use for more than an additional 32 Mbytes for at least a year. What the heck, use the RAM disk software on recent SIG tapes until you need the memory as memory... ================================================================================ Note 468.2 VAX 6000 Memory Card Count 2 of 7 EISNER::FULLER_S "Funky cold medina" 44 lines 14-SEP-1989 21:04 -< Why you WANT to interleave! >- -------------------------------------------------------------------------------- > Just what do I lose by *only* buying one 32Mbyte board? When you have an even number of memory boards, the system will configure the boards to interleave in pairs. If you have an odd number of boards, the system will configure all the boards to interleave in pairs, except for the last one. That one will not interleave with any other board, and will therefore be the slowest board, and will be addressed to the highest physical addresses. The kicker is, where do you think most of VMS resides in physical memory? You're right: the highest physical addresses in memory. The reasoning behind memory interleaving comes from the old core memories. The read of a core memory was destructive. If you read a memory location, the act of reading would zero that location. So, there was some additional circuitry in the memory controller that would immediately write the data back. If you were accessing memory sequentially, then the cpu would have to wait for the memory controller to re-write that location. So the concept of interleaving was developed. This was basically manipulation of the address bits, such that sequential addressing was guaranteed to access one controller then the other. This re-write cycle is also the idea behind the "read-modify-write" cycle. If the cpu knew, in advance that a location would be read and immediately re-written with new data, it would tell the memory controller (via some bus signal), so that the controller would not waste time re-writing the memory contents that the read destroyed (remember the destructive readout from above). So, the memory controller would simply wait for the updated data. We no longer use core memory, but the theory still applies. MOS dynamic memory has 2 times associated with it: Access time and cycle time. Access time refers to the amound of time that must pass between when a memory address is presented to the chip, and the chip responds with valid data. Cycle time refers to the amount of time that you must wait between accesses of the chip (I don't remember why), and is larger than access time. So, if we're sequentially reading memory locations, we can access a location in 100ns (nano-seconds), but we may have to wait another 100ns before we can access the next sequential location (since with 1Mb chips, it's very likely that we'll want the same chip). So, if we can interleave the boards, we can read one location (in 100ns) and read the next location (which is actually on another memory board) in the next 100ns (instead of waiting for the cycle time of the chip). Stu ================================================================================ Note 468.3 VAX 6000 Memory Card Count 3 of 7 EISNER::MCCORMICK_B "Bob McCormick" 8 lines 15-SEP-1989 20:45 -< May work but be invalid... >- -------------------------------------------------------------------------------- If my memory ECC bits are not corrupted... I believe a past conversation with a DEC engineering type generated a response that although an odd number of memory modules on the 6000's CPU-memory bus may indeed work, it may not be supported. (i.e. invalid configuration) Anyone running a system with an _odd_ number of modules (other than one)? ================================================================================ Note 468.4 VAX 6000 Memory Card Count 4 of 7 EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 3 lines 18-SEP-1989 12:28 -< It's valid >- -------------------------------------------------------------------------------- We have a 6000 Model 220, and we've been running with three boards since we bought the thing from DEC. They sold us this config and they maintain it, so it's certainly legal. ================================================================================ Note 468.5 VAX 6000 Memory Card Count 5 of 7 EISNER::OSUDAR "John Osudar" 4 lines 18-SEP-1989 15:05 -< should work -- does if one fails >- -------------------------------------------------------------------------------- If one (or an odd number) of your 6000's memory boards fails self-test at startup, the system is supposed to be able to come up with what's working -- implying an odd number of boards will work. It *does* leave the un-interleaved memory at the top, as was mentioned before! ================================================================================ Note 468.6 VAX 6000 Memory Card Count 6 of 7 EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 11 lines 29-SEP-1989 12:35 -< The last card or the first card can ne noninterleaved >- -------------------------------------------------------------------------------- RE: < Note 468.5 by EISNER::OSUDAR "John Osudar" > -< should work -- does if one fails >- | working -- implying an odd number of boards will work. It *does* | leave the un-interleaved memory at the top, as was mentioned before! This is not entirely true, I've found out. As I said in .4, we have three memory cards in our 6000 series. We have altered the default interleaving, though, to interleave the second two cards and skip the first one. This can be done in a support manner, although I'm unfamiliar with the details. I'll find out how if anyone is interested. ================================================================================ Note 468.7 VAX 6000 Memory Card Count 7 of 7 EISNER::OSUDAR "John Osudar" 6 lines 29-SEP-1989 12:49 -< right; you can control it >- -------------------------------------------------------------------------------- Yes. I should have said that the un-interleaved memory ends up at the top *if you let the system decide interleaving by itself*. There are console commands to set the memory interleaving. I don't have my 6200 book handy, but if you have one then you'll find it in there -- and if you want me to look it up, I can do that too...  ================================================================================ Note 469.0 Swap Maxstor RD54 into MV2000? 3 replies EISNER::HUDGINS "Jerry Hudgins" 12 lines 14-SEP-1989 16:03 -------------------------------------------------------------------------------- I recently obtained an rtVAX 2002, which for the uninitiated is just an 4MB MV2000 with TK50 and RD53, VMS 4-user, DECnet, and ELN toolkit, plus a KA620 with DELQA in a BA23. This is meant to be a cheap ELN development environment (and it is), but it's a bit shy on disk space. I ordered the RD54 expansion box with it, but the special price didn't allow swapping an RD54 for the RD53 in the system unit. Obviously, it doesn't many layered products on top of VMS to fill up the RD53; I've already pretty much done so. However, I DO have a Maxstor RD54-equivalent in an 11/73 that I could swap for it. Are there any special considerations in swapping a drive in an MV2000? Size jumpers or switches (doubtful)? Anyone know of any pitfalls? ================================================================================ Note 469.1 Swap Maxstor RD54 into MV2000? 1 of 3 EISNER::MAYHEW "Bill Mayhew" 5 lines 14-SEP-1989 16:44 -< Nothing too tough... >- -------------------------------------------------------------------------------- No big deal. Physical installation is the worst part. You'll need to reformat it of course, but the 2000 has the formatting "utility" in ROM. There's some information on this on DSIN in the 2000 Hardware database, if you have access to it, though I don't think you particularly need it. ================================================================================ Note 469.2 Swap Maxstor RD54 into MV2000? 2 of 3 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 1 line 14-SEP-1989 18:24 -< PULLING AN MV2000 APART >- -------------------------------------------------------------------------------- Also check the MICROVAX conference topic 59 ================================================================================ Note 469.3 Swap Maxstor RD54 into MV2000? 3 of 3 EISNER::HUDGINS "Jerry Hudgins" 1 line 15-SEP-1989 11:11 -< It worked >- -------------------------------------------------------------------------------- Thanks, gents. Swap occurred uneventfully.  ================================================================================ Note 470.0 Connecting DSA disks uvax 3xxx 3 replies EISNER::HAMPARIAN "Don Hamparian - IDI" 16 lines 15-SEP-1989 10:57 -------------------------------------------------------------------------------- We are looking at moving towards a distributed computing environment. We are looking at giving up our big 8700 in exchange for purchasing 3100s and uvax file servers. The only problem is that we are not giing up our DSA disks (RA82s/90s) and would like to move them into our distributed environment. I know that uVAX 3800/3900 can serve them but 3800s/3900s are expensive. Are there any controllers (3rd party is fine) that are out there that would allow DSA disks to talk to 3300s/3400s? I am confused about the KFQSA available in the 3300/3400s. Does this provide any way to connect a Q bus DSA adapter to these machines? Maybe I'm just *plain* confused. Don ================================================================================ Note 470.1 Connecting DSA disks uvax 3xxx 1 of 3 EISNER::MAYHEW "Bill Mayhew" 20 lines 15-SEP-1989 12:05 -< KDA50-SE >- -------------------------------------------------------------------------------- The KFQSA is a DSSI disk controller. You use it to put the new-ish DSSI disks on a Q-bus system, or to put *MORE* DSSI disks on a 3300/3400 (which have an on-board DSSI controller; you can now add a KFQSA to the system as well... on other Q-bus machines you can have up to two KFQSAs... none of which is particularly relevant to your need ). I have a client who just spent some time with the DECdirect Technical Sales Assistance group (or whatever it's called) researching the use of RA-type disks on a 3300/3400. The answer that came back is, "It's supported" via the KDA50-SE. This is a pretty new development. I thought I had read about it somewhere but can't locate it. I advised the client, and would suggest the same for you, to get this statement of support in writing from DEC somehow. But, it does appear to be the answer, at least to the question you asked. Whether a 3300/3400 has the oomph to service your network, though, is another question that you should consider. Not enough info here for me to guess one way or the other. ================================================================================ Note 470.2 Connecting DSA disks uvax 3xxx 2 of 3 EISNER::MCCORMICK_B "Bob McCormick" 10 lines 15-SEP-1989 20:51 -< More thoughts >- -------------------------------------------------------------------------------- Bill's point about umph is a good one. If you are going to beat on the Etherhose a lot, there is a big difference between the KA650/KA655 (3500,3600/3800,39600) _plus_ a dedicated Ethernet controller DELQA/DESQA and the KA640 with its integrated Ethernet interface on the CPU board. The 3300 box I feel is quite small for most implementations. Only 6 slots and one power supply. If you were to stuff a dual board KDA controller on it... not much left for other stuff. Don't forget user licenses! That can add up to a lot of money. DEC recently announced more users on 3800/3900's for the same $$$! ================================================================================ Note 470.3 Connecting DSA disks uvax 3xxx 3 of 3 EISNER::HAMPARIAN "Don Hamparian - IDI" 21 lines 17-SEP-1989 21:08 -< Cluster Building >- -------------------------------------------------------------------------------- Re: .1, .2. Thanks for the comments. What we are looking at doing is having one server contain our development code and then 5 or so other servers providing system disks for about 30 3100s. The 3100s would have local disks for page/swap. There would also be about 30 character cell terminals connected - mostly for mail and simple editting functions. We then would separate the servers/systems with bridges. Couple issues 1. Maybe DECWindows terminals could be interesting but no info from DEC :-( . 2. Certainly e-net bandwidth (at the server e-net boards) is a major issue. People used to CI cluster performance could be surpised. But, then people have a 2.7 vup processor on their desk. 3. Should this be one cluster or many? I know by requirements are vague, but if anyone has any feedback - fire away.  ================================================================================ Note 471.0 PDP-11 AND VAX MAINTENANCE/TECHNICAL MANUALS 3 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 55 lines 16-SEP-1989 12:02 -------------------------------------------------------------------------------- Do you like real tech manuals? Not the ones DEC normally ships these days that tell you where the power cord goes and that's about it. When DEC first came out with the MicroSystem series they shipped real manuals with them that did neat things like tell you what all the jumpers are for. About 18 months ago DEC converted these to the here's where the power cord goes and call field service useless type manuals. Well DEC has made amends with two new kits for PDP's and VAX's. ZYABX-GZ MicroPDP-11 Maintenance Information Kit CLP $105.00 Vol 1 KDF11-B CPU system maintenance EK-245AA-MG KDJ11-B CPU system maintenance EK-247AA-MG KDJ11-D/S CPU system maintenance EK-246AA-MG Vol 2 BA23 Enclosure Maintenance EK-186AA-MG BA123 Enclosure Maintenance EK-188AA-MG BA213 Enclosure Maintenance EK-189AA-MG BA214 Enclosure Maintenance EK-190AA-MG BA215 Enclosure Maintenance EK-191AA-MG H9642-J Enclosure Maintenance EK-187AA-MG H9644 Enclosure Maintenance EK-221AA-MG Vol 3 Microsystems options EK-192AB-MG ZNABX-GZ MicroPDP-11 Maintenance Information Kit CLP $305.00 Vol 1A KA630 CPU system maintenance EK-178AA-MG KA640 CPU system maintenance EK-179AA-MG KA650 CPU system maintenance EK-180AA-MG MicroVAX Special system maintenance EK-181AA-MG Vol 2B KA655 CPU system maintenance EK-306AA-MG KN210 CPU module set system maintenance EK-329AA-MG Vol 2 BA23 Enclosure Maintenance EK-186AA-MG BA123 Enclosure Maintenance EK-188AA-MG BA213 Enclosure Maintenance EK-189AA-MG BA214 Enclosure Maintenance EK-190AA-MG BA215 Enclosure Maintenance EK-191AA-MG H9642-J Enclosure Maintenance EK-187AA-MG H9644 Enclosure Maintenance EK-221AA-MG Vol 3 R215F Expander Maintenance EK-271AB-MG R213F Expander Installation EK-310AA-IN R215F Expander Installation EK-317AA-IN R213A Expander Installation EK-254AA-IN H9644 Expander Installation EK-312AA-IN Vol 4 Microsystems options EK-192AB-MG I would encourage anyone who works with MicroSystem's hardware to purchase these. They are geared towards field service techs and self maintenance customers. Who know's we buy enough of these and DEC may feel producing tech manuals has value. ================================================================================ Note 471.1 PDP-11 AND VAX MAINTENANCE/TECHNICAL MANUALS 1 of 3 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 9 lines 16-SEP-1989 15:41 -< Very good manuals >- -------------------------------------------------------------------------------- > ZNABX-GZ MicroPDP-11 Maintenance Information Kit CLP $305.00 With the exception of Vol 1B (?) (KA655 & KN210), DEC was selling these at the Spring '89 symposium for $15 per volume! This included a binder. They did not have the PDP-11 one, but I am hoping they will have it at Anaheim in November. Even at $305 for the set of three, it is still a good deal. Some information is still lacking, but this is a *big* improvement over the pabulum we were getting. ================================================================================ Note 471.2 PDP-11 AND VAX MAINTENANCE/TECHNICAL MANUALS 2 of 3 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 7 lines 16-SEP-1989 18:08 -< SHOULD HAVE DOUBLE CHECKED >- -------------------------------------------------------------------------------- > ZYABX-GZ MicroPDP-11 Maintenance Information Kit CLP $105.00 > > ZNABX-GZ MicroPDP-11 Maintenance Information Kit CLP $305.00 Cut and Paste got me. That second item should be.... ZNABX-GZ MicroVAX Maintenance Information Kit CLP $305.00 ================================================================================ Note 471.3 PDP-11 AND VAX MAINTENANCE/TECHNICAL MANUALS 3 of 3 EISNER::REYNOLDS_J "John Reynolds, Eastman Kodak" 13 lines 19-SEP-1989 12:30 -< Clarification, please. >- -------------------------------------------------------------------------------- >When DEC first came out with the MicroSystem series they shipped real >manuals with them that did neat things like tell you what all the >jumpers are for. About 18 months ago DEC converted these to the here's >where the power cord goes and call field service useless type manuals. >Well DEC has made amends with two new kits for PDP's and VAX's. was that : call field service (and tell them they're) useless call field service (because the uV is) useless call field service (because you're) useless or useless ... manuals? :-)  ================================================================================ Note 472.0 Need anti-static terminal 3 replies EISNER::ROLLER "Sir Osis of Liver" 6 lines 18-SEP-1989 15:46 -------------------------------------------------------------------------------- I manage a cluster that collects data from factory floor that produces electronic products. Recently there has been a big push on to reduce ESD (Electro Static Discharge) on the floor. I had many VT terminals out there, and they are no longer acceptable due to their plastic enclosures which allow the buildup of static charges. Does anyone know of any VT compatible terminals which are anti-static? ================================================================================ Note 472.1 Need anti-static terminal 1 of 3 EISNER::MAYHEW "Bill Mayhew" 3 lines 18-SEP-1989 18:04 -< Digital has it now >- -------------------------------------------------------------------------------- DEC makes "industrial" versions of the VT series which, I believe, come in metal cabinets. See section 8 of the Jul-Sep '89 Systems and Options Catalog. ================================================================================ Note 472.2 Need anti-static terminal 2 of 3 EISNER::LEDERMAN "Bart Z. Lederman" 3 lines 19-SEP-1989 07:08 -< How about treating the existing terminals? >- -------------------------------------------------------------------------------- Have you considered talking to companies like (I think) Aquadag, which make conductive coatings? It might be easier (and cheaper) to spray the plastic cases to make them conductive and/or anti-static. ================================================================================ Note 472.3 Need anti-static terminal 3 of 3 EISNER::STONE_L "Larry Stone" 7 lines 19-SEP-1989 22:33 -< Anti-static screens >- -------------------------------------------------------------------------------- How about an anti-static screen on the terminals. Our terminals (VT220s and 320s) are approved in anti-static environment with anti-static screens. We use the Inmac Glare Sentry Plus. Inmac is a general purpose computer supplies company with offices in about a dozen different cities. Expect to pay about $100 per screen.  ================================================================================ Note 473.0 725 upgrade questions 10 replies EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 13 lines 19-SEP-1989 12:01 -------------------------------------------------------------------------------- A friend of mine has a VAX 11/725 with two RC25's and a TU80 (stop laughing). He wants to put some bigger disk on it for as little as possible. Can anyone recommend a good third party controller/disk(s) that will give about 0.5 - 1.0 Gbytes ? We may be able to scrounge up a UDA50. Is this controller supported on a 725 ? What DSA disks (if any) will fit inside a 725 housing ? Finally, if he decides to cuck the 725 and get a microVAX 3xxx (with a Q-Bus) what is the part number of the controller for the TU80. ================================================================================ Note 473.1 725 upgrade questions 1 of 10 EISNER::MAYHEW "Bill Mayhew" 15 lines 19-SEP-1989 13:42 -< Entering the world of unsupported hardware >- -------------------------------------------------------------------------------- *Nothing* is "supported" on the 11/725 other than the RC25s... and that includes the TU80. (Well, if you count the TU58s, they're supported.) It is, however, a standard 11/730 CPU, with a UNIBUS backplane, so anything UNIBUS-compatible should _work_, as long as nobody tells the FCC about you when you bring the cables outside the box. If he wants to chuck the RC25s, it might be possible to put a modern 8" rackmount drive in that space. Otherwise, he's going to have to go to something external. Of course, VMS no longer "supports" the 725 either, due to the disk space limitations of the RC25. That doesn't mean it won't work, though. ================================================================================ Note 473.2 725 upgrade questions 2 of 10 EISNER::LEDERMAN "Bart Z. Lederman" 7 lines 19-SEP-1989 14:40 -< I think it's time to trade. >- -------------------------------------------------------------------------------- Don't get offended by this, but I think what you really want to do is trade it in on a newer machine. DEC is offering what look like reasonable allowances (like, more than the machine goes for on the used market) for the 725, and I think the software licence is considered a high enough class that you can trade them straight for a 3100 or something like that. Plus you will get a system which is going to be supported for a while. ================================================================================ Note 473.3 725 upgrade questions 3 of 10 EISNER::BURNET "John Burnet" 16 lines 19-SEP-1989 18:24 -< Keep it as a pet and get a real computer for work >- -------------------------------------------------------------------------------- The 11/725 is a nice hacker's machine, fun to play around with and all that... it's probably ideal for debugging device drivers if you don't mind waiting a long time for reboots. However, if the intention is to do any serious work on the computer, and especially if two or more people will need to be logged in at the same time, then you may as well stop dreaming. There was probably a time when this wasn't necessarily so (VMS 3.x), but VMS V5 will just eat the 725 for breakfast, belch, and ask for more. We've got an office that (even as I write this) is having its 11/725 swapped for a MicroVAX II. The few people in that office who use the VAX are so used to the slow response time on the 725 that I know we're going to be getting calls in the next few days saying "Wow! Thanks for the upgrade! This new system is lightning-quick!" When a MicroVAX II looks like a Cray by comparison, it's time to move up. ================================================================================ Note 473.4 725 upgrade questions 4 of 10 EISNER::DELARISCH "Arnold S. De Larisch" 13 lines 19-SEP-1989 19:23 -< No DEC made Q-BUS controller for the TU80! >- -------------------------------------------------------------------------------- >> < Note 473.0 by EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" > >> -< 725 upgrade questions >- >> Finally, if he decides to cuck the 725 and get a microVAX 3xxx (with a >> Q-Bus) what is the part number of the controller for the TU80. So sorry ... I had a client in a similar situation, she liked the TU80 an DEC does not make a controller for it on the Q-Bus. There are 3rd party alternatives out there. Check the HARDWARE_HELP conference for references to the TU80 and a PDP 11/44. -Arnold ================================================================================ Note 473.5 725 upgrade questions 5 of 10 EISNER::GLEASON "CyberPunk" 20 lines 20-SEP-1989 00:14 -< Yesterdays VAX >- -------------------------------------------------------------------------------- I think the 11/725 makes for a great toy VAX - they are cheap, and there is plenty of used UNIBUS gear available to hack around with. As a home hobby system, it is impressive looking to guests who know nothing about computers. The cheapest way to add some space to the 725 is to take advantage of the fact that is is really a 730 in a little box, and get a used RB730 controller and an R80 or RM80 (the same, really) disk drive. These two items with some careful shopping can be had for approx. $1000. I've been calling mine an 11/727 since I did this - it's halfway between an 11/725 and an 11/730. If you like, you can also add a couple of RL02s as well. Note, that I said cheapest - the RB730 is not the fastest controller in the world, nor is the R80 the smallest, quitest, or most energy efficient drive in the world. ================================================================================ Note 473.6 725 upgrade questions 6 of 10 EISNER::LEDERMAN "Bart Z. Lederman" 5 lines 20-SEP-1989 07:27 -< I'd still prefer a Micro-VAX >- -------------------------------------------------------------------------------- > RB730 controller and an R80 or RM80 (the same, really) disk drive. These I don't think they are the same drive. We had a 730 go out of service because the disk died, and were told that an RM80 could not be substituted (the interface is different). ================================================================================ Note 473.7 725 upgrade questions 7 of 10 EISNER::RENES "Stuart Renes, DFWLug Chair" 15 lines 20-SEP-1989 14:19 -< I beg your pardon? >- -------------------------------------------------------------------------------- Whenever 11/725's are discussed, as a proud? owner of one, I must interject: VMS 5 doesn't eat a single-user 725 alive. I currently have VMS 5.1-1 running and will soon have 5.2 up on my 725. It's one way to restore the challenge of VMS upgrades lost since the early days. Of course, these comments are only applicable to homw-type systems where COST is the ONLY factor. Stuart. (busy tailoring a virgin 5.2 system) ================================================================================ Note 473.8 725 upgrade questions 8 of 10 EISNER::KENNEDY "Terry Kennedy" 6 lines 20-SEP-1989 16:07 -< Rx80 drive differences >- -------------------------------------------------------------------------------- > I don't think they are the same drive. We had a 730 go out of service > because the disk died, and were told that an RM80 could not be > substituted (the interface is different). They're not an exact swap, but most of the parts are common. The 3 variants are RM80 - Massbus drive, RA80 - "Advanced" drive (SDI), R80 - nothing drive. ================================================================================ Note 473.9 725 upgrade questions 9 of 10 EISNER::GLEASON "CyberPunk" 18 lines 20-SEP-1989 21:14 -< Experimental method ... >- -------------------------------------------------------------------------------- >> I don't think they are the same drive. We had a 730 go out of service >> because the disk died, and were told that an RM80 could not be >> substituted (the interface is different). > They're not an exact swap, but most of the parts are common. The 3 variants >are RM80 - Massbus drive, RA80 - "Advanced" drive (SDI), R80 - nothing drive. Spinning noisily in my kitchen/computer room is an RM80, ordered from a used equipment dealer, and apparently unmodified (though I have never seen a brand new one), attached to an RB730 - works great, lasts a long time. I didn't know for sure it would work when I got it, but, I was willing to do a little experimentation to avoud the cost of a UDA 50 and a real RA drive - those were (at the time I bought the RM80) significantly more expensive. ================================================================================ Note 473.10 725 upgrade questions 10 of 10 EISNER::BURNET "John Burnet" 4 lines 20-SEP-1989 21:24 -< Emphasis on *single-user* :-( >- -------------------------------------------------------------------------------- > < Note 473.7 by EISNER::RENES "Stuart Renes, DFWLug Chair" > > > VMS 5 doesn't eat a single-user 725 alive. I currently have > VMS 5.1-1 running and will soon have 5.2 up on my 725.  ================================================================================ Note 474.0 RA90 Failures/Crashes....Resolutions? 10 replies EISNER::KMIECIK 22 lines 22-SEP-1989 16:17 -------------------------------------------------------------------------------- Has anyone had problems with RA90 drives ? The 4 RA90's that I have are attached, via KDB50's, to a VAX 8530 (NO HSC involved). We are running version 4.7 of VMS. Other disks attached to KDB50s are: 4- RA82 and 1- RA81. I have had no problems with either the RA82's or 81. THe RA90's were installed in March of 89 and since then DIGITAL has replaced 6 of the HDA's and a few of the ECMs. These replacements, for at lease half, have been as the result of HARD disk failures occuring in a relatively short timeframe. Most of the disk failures have occured within 30 minutes of receiving the first error on the disk. The most recent of the 2 failures has been after a complete power shutdown for a weekend and then when powered up two of the RA90 drives could not be mounted - and the RA90 HDA needed to be replaced on both of these drives. Digital performed power down and the power up (field service engineers onsite) Has anyone else had this type of problem/ have any suggestions? Digital is at a loss here....... thanks. ================================================================================ Note 474.1 RA90 Failures/Crashes....Resolutions? 1 of 10 EISNER::KENNEDY "Terry Kennedy" 8 lines 22-SEP-1989 18:02 -< Air conditioner "breath fresheners" in use? >- -------------------------------------------------------------------------------- > Digital is at a loss here....... Hmmm. The RA90 is a thin-film head design, right? Are you using anti- fungus tablets in your computer room air conditioners? This will cause thin-film heads to deteriorate... If this is the problem, and Digital really didn't know of it, you should be able to hear fists hitting foreheads all the way from Colorado Springs! ================================================================================ Note 474.2 RA90 Failures/Crashes....Resolutions? 2 of 10 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 8 lines 22-SEP-1989 20:03 -< A fungus amung us >- -------------------------------------------------------------------------------- > The RA90 is a thin-film head design, right? Are you using anti- > fungus tablets in your computer room air conditioners? This > will cause thin-film heads to deteriorate... Do the RA90 drives really "breath" enough to let in something like that (and what *is* "that")? This is certainly one of the stranger problems I have heard of. How did you find out about it and where can I find more information? ================================================================================ Note 474.3 RA90 Failures/Crashes....Resolutions? 3 of 10 EISNER::KENNEDY "Terry Kennedy" 9 lines 23-SEP-1989 01:26 -< Probably >- -------------------------------------------------------------------------------- > Do the RA90 drives really "breath" enough to let in something > like that (and what *is* "that")? This is certainly one > of the stranger problems I have heard of. How did you > find out about it and where can I find more information? I don't know if they do, but I think they do. I was on the design team for the "failed" RD52 drive. The other folks were mostly ex-IBM 3380 design folks, and this was a problem (and still is) with the IBM 3380's. IBM specifically distributes a warning with the drives... ================================================================================ Note 474.4 RA90 Failures/Crashes....Resolutions? 4 of 10 EISNER::MADISON "Matt Madison" 7 lines 23-SEP-1989 19:11 -< "thermal stabilization"? >- -------------------------------------------------------------------------------- Did the installers let the drives sit for a day in your machine room before they were spun up? When our local F/S guys installed our SA600 they mentioned something about high failure rates in the 90's happening if they weren't "thermally stabilized" before being used. It may sound like an obvious thing (since it's clearly stated in the installation guide), but they said that they had actually witnessed this kind of failure (implying that they didn't read the manual :-). ================================================================================ Note 474.5 RA90 Failures/Crashes....Resolutions? 5 of 10 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 2 lines 25-SEP-1989 00:10 -< OK by us... >- -------------------------------------------------------------------------------- Well.. we've had two RA90s installed since June 26th. We've logged exactly one error in total. 24 hours/day 7 days/week. ================================================================================ Note 474.6 RA90 Failures/Crashes....Resolutions? 6 of 10 EISNER::HAMPARIAN "Don Hamparian - IDI" 1 line 25-SEP-1989 13:44 -< RA90s working fine here >- -------------------------------------------------------------------------------- We have had 3 of them since May - no problems at all. ================================================================================ Note 474.7 RA90 Failures/Crashes....Resolutions? 7 of 10 EISNER::MATUSCAK "Joe Matuscak, B&W, (216)860-1865" 1 line 25-SEP-1989 20:20 -< 2 drives since may, 1 on 3rd HDA >- -------------------------------------------------------------------------------- ================================================================================ Note 474.8 RA90 Failures/Crashes....Resolutions? 8 of 10 EISNER::HUNT_AL "Alan B. Hunt (Ford Motor Co.)" 3 lines 26-SEP-1989 22:52 -< Not too bad here so far >- -------------------------------------------------------------------------------- We have had twelve of them for several months. One HDA and one card failure so far. The one HDA went bad the first time it was powered down and back up after its installation. ================================================================================ Note 474.9 RA90 Failures/Crashes....Resolutions? 9 of 10 EISNER::CANNELL "Larry" 4 lines 10-OCT-1989 16:13 -< Problems with ra90s here as well >- -------------------------------------------------------------------------------- Our site has had two RA90s fail on us. Both were after a full power down that lasted several hours. Another site within the company had 3 of out 8 fail after a similar power down. ================================================================================ Note 474.10 RA90 Failures/Crashes....Resolutions? 10 of 10 EISNER::KENNEDY "Terry Kennedy" 9 lines 10-OCT-1989 20:29 -< A request... >- -------------------------------------------------------------------------------- > Our site has had two RA90s fail on us. Please, when reporting these problems could you state one of: a) Our site uses anti-fungal air conditioning tablets b) Our site does not use anti-fungal air conditioning tablets c) I don't know if our site uses anti-fungal air conditioning tablets By doing this we can see if there is a correlation between the two events.  ================================================================================ Note 475.0 Q-BUS EXPANSION BOXES No replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 7 lines 23-SEP-1989 00:34 -------------------------------------------------------------------------------- OK Q-BUS wizards. I have come into a large supply of BA11-SA boxes as a result of upgrades. DEC sold a BA11-SC version of this box to use as an expansion box. Leaving the issue of the cable interconnects aside - I assume if I cut the H9276 Backplane Jumper for the LTC signal I can use this former system box as an expansion box? BTW - I have a BCV2A-03 interconnect cable set  ================================================================================ Note 476.0 MV3000 CPU'S WITH STANDARD Q-BUS INTERFACES 2 replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 1 line 23-SEP-1989 00:37 -------------------------------------------------------------------------------- MV3000 and standard Q-BUS stuff ================================================================================ Note 476.1 MV3000 CPU'S WITH STANDARD Q-BUS INTERFACES 1 of 2 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 162 lines 23-SEP-1989 00:40 -< H9642-S >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! H9642-S MicroVAX II MIGRATION CABINETS - 21 September 1989 ******************************************************************************* o An expander cabinet for MicroVAX 3000 series systems that adds two BA23 backplanes onto the BA213 Q-bus o A migration tool for OEMs wishing to migrate their installed base to the new, more powerful MicroVAX 3000 series systems o CSS custom engineering services are available o Available October 1989 The H9642-S Expansion Cabinet allows you to upgrade to the powerful MicroVAX 3000 series of CPUs without sacrificing all the popular peripherals and modules supported by the MicroVAX II system. If you have a substantial investment in MicroVAX II peripherals, or in other modules designed for the BA23 chassis, the H9642-S offers the benefits of the newest and most powerful Q-bus systems, while extracting maximum value from current investment and minimizing upgrade costs. PRODUCT DESCRIPTION The H9642-S is a 40-inch high (105 cm) cabinet containing two BA23 chassis, power supplies, control and I/O panels and cooling fans. It also includes the Q-bus inter-cabinet cabling and connector modules to connect the H9642-S cabinet to a MicroVAX 3000 series system housed in a BA213-based cabinet or pedestal. The H9642-S can connect BA23-based devices to the MicroVAX 3400, 3500, 3600, 3800, and 3900 systems, but not to the MicroVAX 3200 or 3300 system, which are based on different chassis. It offers 13 usable BA23 Q-bus slots for supported modules, plus space for four 5.25-inch storage devices and one 13-inch storage device. Both 120 V and 240 V versions are available as Field add-ons for new or existing systems. While Digital cannot directly support, nor accept responsibility for, all of the many Q-bus compatible modules manufactured by third parties or OEMs for traditional BA23, BA123, and H9642 cabinets, it may be possible to use the H9642-S as a tool to develop migration solutions for such devices. CSS would be happy to quote custom engineering services to support specific modules or configurations. FEATURES/BENEFITS Feature Benefit o A hybrid, BA23 to BA213 Q-bus o Allows one Q-bus system solution to cable connection share the benefits of both MicroVAX II and MicroVAX 3000 series options. Feature Benefit o Thirteen BA23 module slots, four o Allows for a wide array of 5.25-inch storage spaces, one configurations for peripherals and 13-inch storage space, and modules to meet specific needs. extended I/O panel cut-outs o H9642 style packaging and o Maximizes the investment in existing compatibility with many MicroVAX II systems. traditional Q-bus modules o Installed as a Field Add-on o Allows connection to existing or new MicroVAX 3000 series systems by local customer service. o CSS pre-sales support is o Compatibility or configuration provided to assist in module questions can be resolved by calling compatibility, determination and the CSS Sales Support Hotline. configuration o CSS can provide customization or o Custom engineering services to compatibility testing for volume support specific modules or special purchasers. configurations can be arranged. Where Are Your Resources? The CSS Sales Support Hotline at 1-800-832-6277 or your local CSS representative should be contacted for more information or configuration support. Documentation An H9642-Sx Expansion Cabinet Product Information Sheet, EC-F0232-49 is available. The H9642-S Expansion Cabinet Installation Guide, EK-H9642S-IN, details the configuration rules and installation processes for the product. PRODUCT POSITIONING There are three choices available for migration from the MicroVAX II system: The MicroMOVE-VUP Program -- should be used when the entire MicroVAX I/II system can be replaced. This program provides hardware trade-ins, plus software and service incentives. The H9642-S Expander Cabinet solution -- should be used when a significant number of modules cannot be traded in. A new MicroVAX 3000 series system is purchased (with or without MicroMOVE-VUP incentives), and the retained Q-bus modules are connected to it from the H9642-S. The CSS Custom Upgrade Program -- should be used when space or packaging limitations require that a single system enclosure, in addition to specific Q-bus modules, be retained. CONFIGURATION INFORMATION The H9642-S Expansion Cabinets are configured using the same processes as other traditional Q-bus systems. However, specific rules need to be followed because of the nature of the hybrid connection to a MicroVAX 3000 series system. These rules may be obtained by contacting the CSS Hotline. Part of the configuration is dependent on the quantities and specific revision levels of each of the Q-bus modules that need to be placed in the H9642-S. Customer Services must perform a pre-sales site inspection and revision level audit of the specific modules to be placed in the H9642-S. SPECIFICATIONS H9642-SF H9642-SG PHYSICAL Height: 41.7 in. (105 cm) Width: 25.7 in. ( 65 cm) Depth: 36.0 in. ( 90 cm) Weight: 288 lb. (139.6 Kg)* ELECTRICAL Voltage 110 VAC 220 VAC Frequency: 60 Hz 50 Hz Power Consumption: 2100 W 2100 W (Maximum) ENVIRONMENTAL Operating Temperature: 15-32oC (59-90oF) Relative Humidity: 20-80% (Non-condensing) Heat Dissipation 1000 BTU * Operating Environment: Class B (Office Environment)** NOTES: *Exclusive of installed options **or Class A (Computer Room) if Class A options are installed PRICING/ORDERING INFORMATION Model No. Description U.S. List Price BMC H9642-SF Dual BA23 Expander for BA213 $11,995 $32 Systems - 120 V H9642-SG Dual BA23 Expander for BA213 * * Systems - 240 V *Country-specific pricing AVAILABILITY: October 1989 ================================================================================ Note 476.2 MV3000 CPU'S WITH STANDARD Q-BUS INTERFACES 2 of 2 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 147 lines 23-SEP-1989 00:45 -< VAXstation 3500/3520/3540 DWFQA Q-BUS ADAPTER >- -------------------------------------------------------------------------------- .___.___.___.___.___.___.___. ! ! ! ! ! ! ! ! | d | i | g | i | t | a | l | PROD - N e w s !___!___!___!___!___!___!___! VAXstation 3520/3540 OPTIONS - 21 September 1989 ******************************************************************************** =================================================================== | | | o VAXstation 3520/3540 DWFQA Q-bus adapter | | | | o VAXstation 3520/3540 32 MB memory module | | | | o Available now | | | =================================================================== PRODUCT DESCRIPTION VAXstation 3520/3540 DWFQA Q-BUS ADAPTER The DWFQA (Q-bus adapter) module makes it possible to use Q-bus communications and data storage options with the VAXstation 3520/3540 systems. The Q-bus adapter achieves this by providing a communications link between the M-bus and Q-bus modules. The DWFQA module occupies slot 9 of the system backplane and provides three Q-bus option module slots. The Q-bus options will occupy slots 10 through 12. The following Q-bus options work with the DWFQA module: o DSV11-SA/SF/SB/SG controller: a quad-height module providing two synchronous line interfaces allowing high-speed connection to SNA networks, X.25 networks and DECnet networks. o CXY08-AA/AF asynchronous multiplexer: performs data concentration, real-time processing, and interactive terminal handling. The CXY08 module is quad-height, with a BA213-style handle. o TK70E-SA/SF: a magnetic streaming tape drive subsystem that provides up to 296 megabytes of backup data storage on a tape cartridge. The TQK70 controller is a tape mass storage control protocol device that provides the interface between the TK70-AA tape drive and the Q-bus. o TSV05-SE/SF: a 26.25-centimeter (10.5-inch) 1600 bit, 9-track, industry standard magnetic streaming tape drive that provides 40.5 megabytes of backup data storage. The TSV05 controller is a modular, microprocessor- based quad-height module. VMS AND ULTRIX SUPPORT FOR Q-BUS OPTIONS The following lists which Q-bus options are supported on VMS and ULTRIX for the VAXstation 3520/3540 systems. (NOTE: VMS 5.2 or ULTRIX UWS 2.1 is required for DWFQA support.) Option Module VMS 5.2 ULTRIX UWS 2.1 DSV11 Yes No CXY08 Yes Yes TQK70 Yes Yes TSV05 Yes Yes Like other Q-bus machines, Digital does NOT support third-party options. If third-party manufacturers follow the Q-bus spec, then the option should work. 1. The DWFQA does not fully support block-mode DMA nor extended block-mode (Hog-mode). Because of the M-bus architecture, the maximum DMA rates that are achievable are in the 480-650 Kbytes/sec. range. 2. Highest M-bus priority 3. Must adhere to 8.2us Q-bus timing spec 4. Diagnostics are in 4 ROMs (256 KB total) 5. BA213 expander box not supported 32 MB Memory Module o The VAXstation 3520/3540 system has four M-bus slots. The slots are used for interconnecting CPU, memory, and the DWFQA. o The MS60-CA will allow the VAXstation3520/3540 to have up to 128 MB of ECC memory. o In dual CPU configurations with no other options attached to the M-bus, up to four 32 MB memory modules can be installed (128 MB). o The addition of a second CPU and the DWFQA will each take up a M-bus slot. In this case, only two memory modules can be installed (64 MB). VAXstation 3520/3540 BACKPLANE For more information, see the Systems & Option Catalog, page 1C.7. -------------------------------------------- 1 | 16-plane graphics module module (optional) | |--------------------------------------------| 2 | 8-plane graphics module | |--------------------------------------------| 3 | Graphics module | |--------------------------------------------| 4 | Input/output | |--------------------------------------------| 5 | Memory (optional) | |--------------------------------------------| 6 | Memory or CPU module (optional) | |--------------------------------------------| 7 | CPU module | |--------------------------------------------| 8 | Memory (8 or 16 or 32 MB) | |--------------------------------------------| 9 | Memory or VS60B-AA or DWFQA-AA (optional) | |--------------------------------------------| 10 | Q-bus slot (optional) | |--------------------------------------------| 11 | Q-bus slot (optional) | |--------------------------------------------| 12 | Q-bus slot (optional) | -------------------------------------------- PRICING/ORDERING INFORMATION Model No. Description U.S. List Price BMC Instal. DWFQA-AA Q-bus adapter, factory $ 4,000 $5 $510 installed DWFQA-AF Q-bus adapter 4,000 5 510 MS60-CA 32 MB memory 26,400 N/C 560 QUESTIONS/ANSWERS WILL THE VS60B-AA OPTIONAL TAPE KIT STILL BE AVAILABLE WITH THE ANNOUNCEMENT OF THE DWFQA-AA? Yes, in fact, all VS60B-AA shipped after September 11, 1989 will include the DWFQA. WHAT ARE THE MAXIMUM NUMBER OF Q-bus OPTIONS SUPPORTED PER OPTION? Up to three of any options are supported. CAN THE Q-bus BE EXTENDED OR EXPANDED BEYOND THE VAXstation 3520/3540 BOX OR CABINET? No. WILL ANY COMBINATION OF Q-bus OPTIONS EXCEED THE POWER AVAILABLE? No.  ================================================================================ Note 477.0 is a CLONEd DS200 or DS300 possible 2 replies EISNER::BRUCE "Barton F. Bruce - CCA/MA" 33 lines 25-SEP-1989 13:30 -------------------------------------------------------------------------------- I ran into a former DECie at a HAM/COMPUTER flea market at MIT who KNEW why DS200s have performance problems. The memory refresh for the DRAMS is the responsibility of the S/W! He said DEC saved about only peanuts per box that way. On the DS300, the memory used was SRAMS simply because in DEC's standard parts book (that designers almost HAVE to select parts from) the DEC cost for SRAMS was incredibly LOW ("lower than anyone else gets them for"). The DS200 used a 68000 @ 10mhz, and the DS300 uses a 16mhz 68000 (at the time it was designed this was not publicly available) running at 14.7mhz because the 14.7 easily divides down for the standard baud rates and so saved a second crystal. There is NOTHING in either design that is sneeky, and a cookie cutter copy should run with the standard down loaded S/W. The prom code for diags + booting obviously can't be simply copied, but would have to be functionally replaced. If I remember correctly, DEC was charging $50. per DS200 box for the S/W lic to run it. NOW they are asking $50/port of anyone getting a legal LAT lic. It would seem to be cheaper to simply knock of copies and suggest that customers 'write their own TCP/IP S/W' or, if they choose to try DEC S/W, pay DEC the $50 (per box), than to write ones own S/W for ones own BOX and then scramble to keep up with new features when DEC ships new S/W. Since none of the LAT boxes I have seen went this way, it seems there must be a reason, but is there one? The DS300 is obviously the one to clone. ================================================================================ Note 477.1 is a CLONEd DS200 or DS300 possible 1 of 2 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 10 lines 26-SEP-1989 11:38 -< TCP/IP can slow down DS200/300 >- -------------------------------------------------------------------------------- The thread 148.* in the DEC_NETWORKING conference stated that one thing that will surely choke a DS200 is lots of workstations running TCP/IP ! The TCP/IP protocol does not have any provision for a broadcast message for updating routing/available node lists so it uses the Ethernet broadcast message id. This can not (or may I should say "should not") be filtered out by hardware in the controller and must be passed on to the host (the 68000 in the case of the DS200/300) who then throws it away. This will show up as a large number of UNRECOGNIZED DESTINATION counts and will eventually clog/stop as DS200 from "doingit's thing". ================================================================================ Note 477.2 is a CLONEd DS200 or DS300 possible 2 of 2 EISNER::KENNEDY "Terry Kennedy" 31 lines 27-SEP-1989 05:53 -< Assorted comments >- -------------------------------------------------------------------------------- A few things (likely to generate flames, but what the heck): 1) DS200's have performance problems? News to me. I haven't seen any in networks I feel are properly configured. LAT is *Local* Area Trans- port. If you have a large net, and all servers see all hosts, and there are bridges, you can get hurt. That means it's time to think about whether that configuration is necessary. 2) FF-FF-FF-FF-FF-FF packets - The Blue Book says they are a BAD idea and should only be user sparingly and wne absolutely necessary. What are your DS200's doing listening for TCP traffic they can't handle? Time for some agressive filtering. A DS200 with TCP code doesn't exist right now outside of DEC, and I suspect if such a product was released it would want a high- er-performance box. 3) License issue - The license for a DS200 is *not* available seperately. There was an error a while back which implied it was, and there was a price for it - but unless you stocked up back then, you're out of luck now - it is bundled with the server. 4) Cloning the firmware - Not that easy. The firmware remains as operating system and device driver support even after the server gets the load image. It isn't just a case of turning over control to the load image. 5) DS300 being the one to clone - sorry, I disagree. If you need full modem control, you need a DS200 (or a 5x0, but that's another issue). 6) Secrecy of DS200 information - I have both the Technical manual and the Service manual as well as the User's guide - they're orderable parts. I can post the #'s when I get back to the office on Monday if anybody wants them.  ================================================================================ Note 478.0 AVALON'S VACCELERATOR 2 replies EISNER::CLOSE "Eddie Close" 11 lines 26-SEP-1989 09:54 -------------------------------------------------------------------------------- Can anyone share any experiences/comments on Avalons VACCELERATOR AP/30 (or others in the series AP/10 OR AP/20)? The AP/30 is supposed to deliver approx 15 MVUP's of performance for compute intensive fortran programs for approx $15K. We have several jobs that run between and use 1-8 days of CPU so this product looks attractive to us. Are there other products that we should look into? Thanks for the help. EGC ================================================================================ Note 478.1 AVALON'S VACCELERATOR 1 of 2 EISNER::CLOSE "Eddie Close" 12 lines 24-OCT-1989 08:12 -< Going....Going....Gone? >- -------------------------------------------------------------------------------- Reposting as a final attempt for assistance. > Can anyone share any experiences/comments on Avalons VACCELERATOR > AP/30 (or others in the series AP/10 or AP/20)? The AP/30 is supposed > to deliver approx 15 MVUP's of performance for compute intensive fortran > programs for approx $15K. > We have several jobs that run between and use 1-8 days of CPU so > this product looks attractive to us. > Are there other products that we should look into? Thanks for the > help. Eddie ================================================================================ Note 478.2 AVALON'S VACCELERATOR 2 of 2 EISNER::ROBERTS_D "don roberts #241" 3 lines 24-OCT-1989 10:20 -< Well, does this help?b :-} >- -------------------------------------------------------------------------------- Uh, well, one of the principals and one of the employees of Avalon are family members of mine, so you are indirectly benefitting me if you buy their product...  ================================================================================ Note 479.0 Third party SCSI disk experiences wanted 8 replies EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 12 lines 26-SEP-1989 11:56 -------------------------------------------------------------------------------- I'm looking for first hand experience here so no speculation please. Is anyone using **any** third party SCSI disks on a VAXstation 3100 ? Which ones ? What are their sizes ? Does it use the standard DEC device driver ? Any magic words required to INIT and/or MOUNT the disk ? I've heard some rumors that you'll have to wait until VMS 5.3, but what is it that we are waiting for ? I am especially interested in **LARGE** extrenal 5.25" devices (both Imprimis and Maxtor have SCSI drives that are > 1 Gbyte !) ================================================================================ Note 479.1 Third party SCSI disk experiences wanted 1 of 8 EISNER::KILGALLEN "Larry Kilgallen" 25 lines 26-SEP-1989 13:26 -< VMS V5.3 should be immaterial >- -------------------------------------------------------------------------------- > I'm looking for first hand experience here so no speculation please. > I've heard some rumors that you'll have to wait until VMS 5.3, but > what is it that we are waiting for ? I will defy your first request, presuming that it applied to your second paragraph (disk experience) rather than your third. If anyone had first hand experience with VMS 5.3, they would not be allowed to say so anyway :-). What is supposed to be coming in a future version of VMS, according to the SCSI driver talk given in Atlanta, is a supported Port/Class driver interface for SCSI devices. This is the opposite of the terminal driver Port/Class driver interface where DEC supplies the class driver and you supply the port driver (for your strange connection mechanism). In this case, DEC will supply the port driver (for the SCSI bus) and you will supply the class driver, for your unsupported device type (e.g., paper tape punch, thermonuclear reactor, etc.). DEC already supplies class drivers for disk and tape. Thus at least this future software development has nothing at all to do with third party disks which are intended to emulate DEC disks. That concern (which is central to .0) is tied up in the business of how DEC interprets the SCSI electrical specifications and how that matches against the way the third parties interpreted them. ================================================================================ Note 479.2 Third party SCSI disk experiences wanted 2 of 8 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 20 lines 26-SEP-1989 17:56 -< SCSI later than 5.2 will support DISCONNECT >- -------------------------------------------------------------------------------- > I'm looking for first hand experience here so no speculation please. Have none, so I will be brief, but the following is worth noting. > I've heard some rumors that you'll have to wait until VMS 5.3, but > what is it that we are waiting for ? NEWS just had a piece from FORREST@SYBASE.COM that DEC isn't supporting DISCONNECT as of 5.2 for non DEC disks. They were afraid some vendors would not implement it properly, and were NOT trying to screw 3rd party disk users. Troops @ Harvey Mudd have modified the driver to work with device returned strings other than DEC's. The performance impact is in configs with >1 drive. Dec has decided to support it from other drives. In 5.3??? Bob Tinkelman had mailed me a copy of it, and maybe he will be so kind as to post the whole text here. ================================================================================ Note 479.3 Third party SCSI disk experiences wanted 3 of 8 EISNER::TINKELMAN "Bob Tinkelman - CCA" 73 lines 28-SEP-1989 16:12 -< 3rd party SCSI disks, news item by Jon Forrest >- -------------------------------------------------------------------------------- Attached is the news posting to which Barton was referring. It was from Jon Forrest at Sybase and was dated 19 Sep 89 16:15:57 GMT  After posting a series of questions about 3rd party disks on Vaxstation 3100s I received a phone call from a DEC engineer who worked on the VS3100 disk system. He told me several interesting things: 1. It is true that VMS 5.2 (and lower) doesn't support the DISCONNECT feature with 3rd party drives. (DEC drives DO do disconnect). The reason for this is that when DEC was first developing the disk system for the VS3100 they believed that not all disks would implement DISCONNECT correctly. Rather than taking a chance, they just chose not to do DISCONNECT on 3rd party drives. Because of potential legal concerns, DEC has changed its mind on this and will support DISCONNECT on 3rd party drives in a "future release" of VMS. By the way, the way DEC determines if a drive is 3rd party is by strings returned by the drive. As the boys down at Harvey Mudd discovered, you can patch the SCSI driver to think that 3rd party drives are DEC drives by changing various character string constants in the driver. (Don't ask me exactly how to do this because I don't know). He also said that the DISCONNECT function makes the biggest difference when you have more than one drive. 2. Disks connected to a VS3100 use the asynchronous communication method. This is because the local port hardware in the VS3100 can't handle the higher speeds of the synchronous mode. The Vaxstation 3510 and 3520 can do synchronous data transfer. 3. DEC did not cripple the VS3100 disk system because VS3100 performance would be too close to higher priced Vaxes. The decision to not support DISCONECT was purely a technical decision. It is my guess that the decision to not support synchronous mode was due to cost considerations. What all this means is that, as of now, drives that look like 3rd party drives to VMS will not do DISCONNECT, resulting in reduced performance. There is a potential for problems with 3rd party drives not doing DISCONNECT right. This sounds to me like the kind of problem that used to exist with attaching 3rd party devices to the Unibus. Most devices work but sometimes there's one that's in the twilight zone. But, until the future release of VMS that does support DISCONNECT comes out, there'll be no problems since you couldn't use DISCONNECT with 3rd party drives unless you get a new (or patched) driver. What surprises me is the lack of 3rd party drives for VS3100s. I only saw two places advertising such drives in the latest issue of Digital Review. On the other hand, since there are also only two places I'm aware of that make 3rd party memory for VS3100s maybe it's just too early. Jon Forrest forrest@sybase.com {pacbell,sun,{uunet,ucbvax}!mtxinu}!sybase!forrest 415-596-3422 --- News headers attached ---  X-NEWS: ccavax comp.os.vms: 2633 Relay-Version: VMS News - V5.8A 21/06/89 VAX/VMS V5.2; site ccavax.camb.com Path: ccavax!adelie!mirror!bu-cs!oliveb!apple!gem.mps.ohio-state.edu!tut.cis.ohio-state.edu!ucbvax!mtxinu!sybase!forrest@sybase.com Newsgroups: comp.os.vms,comp.sys.dec Subject: Answers to Questions About 3rd Party Disks on Vaxstation 3100s Message-ID: <6131@sybase.sybase.com> From: forrest@sybase.com Date: 19 Sep 89 16:15:57 GMT Reply-To: forrest@sybase.com Sender: news@sybase.sybase.com Organization: Sybase, Inc. Lines: 52 Xref: ccavax comp.os.vms:2633 comp.sys.dec:339 ================================================================================ Note 479.4 Third party SCSI disk experiences wanted 4 of 8 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 4 lines 29-SEP-1989 11:20 -------------------------------------------------------------------------------- Okay, I'm ignorant. What is this DISCONNECT thing that is causing all the problem ? Does it have to do with DISMOUNT or SYSGEN CONNECT or what ? ================================================================================ Note 479.5 Third party SCSI disk experiences wanted 5 of 8 EISNER::KENNEDY "Terry Kennedy" 13 lines 30-SEP-1989 09:14 -< Explanation >- -------------------------------------------------------------------------------- > What is this DISCONNECT thing that is causing all the problem ? > Does it have to do with DISMOUNT or SYSGEN CONNECT or what ? Nope, it's a SCSI-ism. Original SCSI (and SASI) required the host adaptor (or controller, in DECspeak) to remain in contact with the drive for the duration of the command. So, on a read with an implied seek, you had to hold the path open until the read started. Same for other commands. DIsconnect is used for performance enhancement - If the drive logic "knows" that the command it just got will take a while, it and the host adaptor can agree to disconnect, letting the host adaptor talk to other drives, until the first drive is ready to transfer some data. If you think of it as overlapped seek support you won't be far off. ================================================================================ Note 479.6 Third party SCSI disk experiences wanted 6 of 8 EISNER::HAHN "DECUServe Vice Chair" 7 lines 10-OCT-1989 21:17 -< SCSI for mVAX II >- -------------------------------------------------------------------------------- What about a SCSI controller for the Q-BUS mVAX II. Who has such a device with capability of attaching 300-800 Mbyte disk and 4 or 8 mm backup tape? Can you give me the name of the vendor(s) which have same and do you have any experience with them? So far Winchester Systems is the only name I have. ================================================================================ Note 479.7 Third party SCSI disk experiences wanted 7 of 8 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 19 lines 10-OCT-1989 21:54 -< CMD supports tape and disk on SCSI >- -------------------------------------------------------------------------------- >What about a SCSI controller for the Q-BUS mVAX II. Who has such a device >with capability of attaching 300-800 Mbyte disk and 4 or 8 mm backup tape? NO experience, and the following is from bad memory notes... You may find you need a seperate controller for disk and tape at least for the current drivers, and later just for performance. CMD currently has a controller that can handle both disk and tape, but I think they have 2 seperate controllers on the same card. Dilog and Emulex have SCSI controllers. DILOG expects to have some sort of dual support shortly. A big bummer is the cost of those tiny connectors with many wires. The cab kit becomes a significant $ component, and just gets worse if the card has 2 seperate SCSI buses. ================================================================================ Note 479.8 Third party SCSI disk experiences wanted 8 of 8 EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 9 lines 11-OCT-1989 08:19 -< Summus or TD Systems >- -------------------------------------------------------------------------------- >>What about a SCSI controller for the Q-BUS mVAX II. Who has such a device >>with capability of attaching 300-800 Mbyte disk and 4 or 8 mm backup tape? Summus Computer 442Q and TD Systems Viking controllers. Both are in massive use here at Fermi. Both controllers connect a single SCSI bus with up to 7 devices (partitioned as 3 tape and 4 disk) to the QBus, appearing on the QBus as separate MSCP and TMSCP controllers. The most common installation here is a Summus/TD Systems controller with an 8mm tape and a 600 MB CDC/Imprimis Wren-V disk.  ================================================================================ Note 480.0 "TK50 AND TK70 on a mVAX II? mVAX 3300?" 1 reply EISNER::HUHN "Eric Huhn, ARA/Phila." 15 lines 26-SEP-1989 12:13 -------------------------------------------------------------------------------- We are supporting several dozen stand-alone MicroVAX systems (II, 2000, 3100, 3300) in our division offices around the country using a MicroVAX II here in our Philadelphia headquarters. Because of the variety of systems, we'd like to add a TK70 to our system, but we keep getting conflicting information about whether this can be done. Does anyone know for sure? Our configuration is: a MicroVAX II, with a BA123 cabinet holding 2 RD54's and a TK50. If this can't be done with a MicroVAX II, we'd be open to buying a MicroVAX 3300 (and shipping the mVAX II to one of the field offices). So the same question for a 3300: Can it have both a TK50 and a TK70? In either situation, internal or external drive doesn't matter to us. ================================================================================ Note 480.1 "TK50 AND TK70 on a mVAX II? mVAX 3300?" 1 of 1 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 18 lines 26-SEP-1989 13:55 -< No problem on MicroVAX-II in BA123 >- -------------------------------------------------------------------------------- > a MicroVAX II, with a BA123 cabinet holding 2 RD54's and a > TK50. I see no problem here. You have enough physical space to mount the TK70. I assume you have a spare bus slot, since the controller can handle only one drive. VMS will quite happily configure two MU-type device, so you are OK there. Just a bit of trivia picked up from a DEC TK50/TK70 expert at the last symposium: The TQK70 controller will work with either the TK50 or TK70 drive (and give you better TK50 performance). Also, it is possible to rig up a switch to allow use of either drive. (I don't know if you would have to re-boot when you switch.) This would not be a trivial job since you have to switch 13 wires (the other 13 are grounds), but a determined hacker should be able to do it.  ================================================================================ Note 481.0 LETTERWRITER 100 7 replies EISNER::TATRAI 8 lines 28-SEP-1989 16:09 -------------------------------------------------------------------------------- I have a Letterwritter 100 that was connected to a VT100. A new VT320 waand connected to the printed using tnter with DEC cables. When I go to print it prin ok for a while then the printer begins to beep and send out garbage. The settings on the VT320 are: speed 4800 xoff 8 bits no parity terminato= FF. I do not have a manual to the printer so if any switches need to be set can you let me know what they are. Thanks. ================================================================================ Note 481.1 LETTERWRITER 100 1 of 7 EISNER::BONE "George Bone, MINSY" 25 lines 29-SEP-1989 10:22 -< No switches - done through keyboard >- -------------------------------------------------------------------------------- < and send out garbage. The settings on the VT320 are: < speed 4800 xoff 8 bits no parity terminato= FF. I do not have < a manual to the printer so if any switches need to be set can you < let me know what they are. Thanks. No switches - this is all done through the keyboard. Enter SETUP by pressing both SETUP and CTRL keys simultaneously - SETUP light should be flashing Check current settings by pressing the "8" key (not the keypad - the one on the top of the keyboard). This will give you a printout of current settings. If any need to be changed, press the appropriate key (P for parity, S for speed, etc) and . THis will show you the available settings. To change them, type =. For example, to change speed to 4800 baud, type S=L Change all required settings, then store them by pressing 9. Hope this helps. By the way, the LETTERWRITER 100 OPERATOR GUIDE has a part number of EK-LW100-OP-001. You might want to order one. GBB ================================================================================ Note 481.2 LETTERWRITER 100 2 of 7 EISNER::HAHN "DECUServe Vice Chair" 2 lines 29-SEP-1989 15:33 -< Letterprinter 100 = LA 100 >- -------------------------------------------------------------------------------- My Ltterprinter 100 (AKA LA100) has no keyboard and baud, parity etc is done with switch setting. Next note will tell setting ================================================================================ Note 481.3 LETTERWRITER 100 3 of 7 EISNER::HAHN "DECUServe Vice Chair" 87 lines 29-SEP-1989 16:08 -< LA100 sitch setting >- -------------------------------------------------------------------------------- Baud Switch A 1-5 (R = right; L = Left) 1 2 3 4 5 50 R R R R R 75 L R R R R 110 R L R R R 134.5 L L R R R 150 R R L R R 200 L R L R R 300 R L L R R 600 L L L R R 1200 R R R L R 1800 L R R L R 2400 R L R L R 4800 L L R L R 7200 R R L L R 9600 L R L L R There are more baud setting for split speed Parity Switch A 6-7 7 Bit 8 Bit 6 7 Space None R R Mark None L R Even Even R L Odd Odd L L Data bit selection Switch A 8 7 Bit per character R 8 Bit per character L Switch B Switch B 1 - Speed Control/Restraint Switch 1 R = restraint 1 L = speed Control Switch B 2 - Modem Control 2 R = no modem control 2 L = modem control Switch B 3-4 Paper out fault Feature response Switch setting 3 4 No Action stop printing R R Paper Low drop DTR L R (after CD is dropped) Paper out drop DTR R L Paper out send 275 ms Break Signal L L Switch B 5 Coded Disconnect Switch 5 R = coded disconnect feature off 5 L = coded disconnect feature on at EOT auto disconnect communication line Switch B 6 XON/XOFF switch 6 R = auto XON/XOFF disabled 6 L = auto XON/XOFF enabled Switch B 7 Receiver Error Processing Switch 7 R = feature off 7 L = feature on printing of error character on receipt of parity error, framing error UART overrun Switch B 8 Store Feature Switch move switch from R to L stores whatever settings you have this cannot be changed by programming and can be reset by power off-on - Moving switch from L to R does nothing. ================================================================================ Note 481.4 LETTERWRITER 100 4 of 7 EISNER::HARVEY "Jack Harvey" 12 lines 29-SEP-1989 20:06 -< Flow Control Problem? >- -------------------------------------------------------------------------------- > it prints ok for a while then the printer begins to beep > and send out garbage. This *sounds* very much like a flow control problem. When the printer buffer fills up at 4800 baud, and is unable to get the terminal to stop sending, this is what I would expect to happen. The mystery, of course, is why it worked with the VT100 and now fails with the VT320. Could it be that the VT100 supported RTS/CTS or DTR for flow control? I don't recall that. If the printer doesn't have XOFF enabled, that would be something to try. ================================================================================ Note 481.5 LETTERWRITER 100 5 of 7 EISNER::RICE "Gary Rice" 13 lines 30-SEP-1989 08:42 -------------------------------------------------------------------------------- > This *sounds* very much like a flow control problem. > When the printer buffer fills up at 4800 baud, and is > unable to get the terminal to stop sending, this is what > I would expect to happen. The mystery, of course, is > why it worked with the VT100 and now fails with the VT320. A VT-100 if it is doing its "smooth scrolling" thing, is actually operating at around 1200 baud, well within the LA100's print speed rangs. Tha LA100 is rated at 180-240 characters per second in draft mode. Gary ================================================================================ Note 481.6 LETTERWRITER 100 6 of 7 EISNER::KENNEDY "Terry Kennedy" 4 lines 30-SEP-1989 09:09 -< Possible bug in 320 >- -------------------------------------------------------------------------------- This may very well be a firmware bug in the 320. I recall several, but then I got rid of all my 320's. When you power up or enter Setup (I don't remember which) it will display a version number. If you post that, maybe somebody with a newer 320 can tell you if you're current. ================================================================================ Note 481.7 LETTERWRITER 100 7 of 7 EISNER::TATRAI 5 lines 3-OCT-1989 11:55 -< Thank you >- -------------------------------------------------------------------------------- Thanks for all the help especially the switch settings from Hahn. It seems the printer was set up at 2400 baud 7 bit even parity. It would have taken me a while to try all of the parity bit and speed combos and since the VT100 had broke I coul't tell what its settings were. All seems to be OK now. Again thanks.  ================================================================================ Note 482.0 Cipher 880 read error 3 replies EISNER::POWELL_M 7 lines 30-SEP-1989 15:19 -------------------------------------------------------------------------------- We have a Cipher 880 with a problem. The first large (8192 byte) datablock written to the tape can't be read without getting retry errors. Other blocks seen to read just fine. Has anyone heard of this problem...if it is any help, this seems to have just started when we installed a EMULEX QT13 (we had been using a TC03...but it stopped working with VMS 5.0). ================================================================================ Note 482.1 Cipher 880 read error 1 of 3 EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 0 lines 30-SEP-1989 19:19 -< IS IT A 3200 BPI TAPE? IF SO YOU ARE IN TROUBLE >- -------------------------------------------------------------------------------- ================================================================================ Note 482.2 Cipher 880 read error 2 of 3 EISNER::SMITH_PA "Paul Smith" 14 lines 1-OCT-1989 22:32 -< Lost bytes. Related? >- -------------------------------------------------------------------------------- > We have a Cipher 880 with a problem. The first large (8192 byte) > datablock written to the tape can't be read without getting retry > errors. Other blocks seen to read just fine. > > Has anyone heard of this problem...if it is any help, this seems > to have just started when we installed a EMULEX QT13 (we had been > using a TC03...but it stopped working with VMS 5.0). We also upgraded to QT13 from TC03 and had problems with losing the first byte of a block. The problem was with the default set-up of the strobing parameter (I believe - I don't have notes with me). Easily changed with the menu, works like a champ now. ================================================================================ Note 482.3 Cipher 880 read error 3 of 3 EISNER::POWELL_M 6 lines 4-OCT-1989 11:06 -< Problem solved >- -------------------------------------------------------------------------------- Thanks for your input. We tried changing the strobing parameter but it made the problem worse. Since the sub system-origionally worked for about three weeks we looked at the novram parameters again and they were different from when they were first set. One of the parameters was set for CDC tape drive. We set the controller for Cipher 880 and life was good. Thanks again for the help.  ================================================================================ Note 483.0 Do dual-ported TA78s work? 5 replies EISNER::TACKETT "Galen Tackett" 12 lines 30-SEP-1989 23:51 -------------------------------------------------------------------------------- I'm having problems with dual-ported TA78 drives. The problem is that if both port buttons are pushed in while a tape is being accessed, sooner or later one or the other HSC will report several formatter errors and then declare the formatter offline. The other HSC continues processing it with no problem. THe errors reported usually have a device type of "????" or something like that and a unit number on the order of 65535. My DEC CE says that pushing in both port buttons is unsupported. I thought it was supposed to be supported! We have HSC V3.94 and VMS V5.1-1, and the drives have had the TX78U upgrade. Does anyone know the real story on support for dual-ported TA78 drives? ================================================================================ Note 483.1 Do dual-ported TA78s work? 1 of 5 EISNER::AZZOLI 9 lines 1-OCT-1989 14:57 -< >- -------------------------------------------------------------------------------- > My DEC CE says that pushing in both port buttons is unsupported. > I thought it was supposed to be supported! We have HSC V3.94 and > VMS V5.1-1, and the drives have had the TX78U upgrade. Does anyone > know the real story on support for dual-ported TA78 drives? It is unsupported, but I think I saw something in the VMS V5.2 release notes that said 5.2 supported dynamic failover of tapes. I don't have the release notes handy so I am not positive. ================================================================================ Note 483.2 Do dual-ported TA78s work? 2 of 5 EISNER::FULLER_S "All stressed up + no one to choke" 6 lines 1-OCT-1989 20:44 -< Works fine here! >- -------------------------------------------------------------------------------- Well, our drives have been dual-ported (with both buttons in) for quite some time - at least as early as V5.0 days. We have been having some problems of late with the drives, but this appears to be unrelated to the dual-porting. Stu ================================================================================ Note 483.3 Do dual-ported TA78s work? 3 of 5 EISNER::HUNT_AL "Alan B. Hunt (Ford Motor Co.)" 10 lines 2-OCT-1989 23:39 -< It do work! >- -------------------------------------------------------------------------------- Mount verification for tapes with failover has been supported since V5.0 as stated by Stu. We have used it and as one guy here said: "It saved my ..." You may find the failover one way, however. After it fails over from A to B or B to A seems like I recall something about having to do some reboots of the HSCs before it can failover again. This may not be true now but I thought there was some limitations in the early release. ================================================================================ Note 483.4 Do dual-ported TA78s work? 4 of 5 EISNER::TACKETT "Galen Tackett" 14 lines 7-OCT-1989 15:33 -< What's your rev level? >- -------------------------------------------------------------------------------- The CE seems to think the restriction has something to do with the HSC and not with VMS. He hasn't shown much willingness to do anything about it, either, since usually there's no serious interference with availability of the drives. If there's really some problem with the HSC that prevents this from working, he should at least supply me with a good written statement of the limitation. If you have dual-ported TA78 drives that work OK with both buttons in, could you post your HSC software version, K.STI rev level, VMS version, and whatever else seems appropriate (maybe the formatter rev level)? That might help me persuade him to do something. ================================================================================ Note 483.5 Do dual-ported TA78s work? 5 of 5 EISNER::FULLER_S "All stressed up + no one to choke" 8 lines 7-OCT-1989 23:10 -------------------------------------------------------------------------------- > If you have dual-ported TA78 drives that work OK with both buttons > in, could you post your HSC software version, K.STI rev level, VMS > version, and whatever else seems appropriate (maybe the formatter > rev level)? That might help me persuade him to do something. We're running HSC V390, K.STI is mc=26 (or 27), VMS V5.anything, formatter rev 4 or 5. In other words, nothing special in the revs.  ================================================================================ Note 484.0 Changing RF disk parameters 3 replies EISNER::GEIER "Jim Geier, General Instrument" 58 lines 3-OCT-1989 18:20 -------------------------------------------------------------------------------- [This may be well documented, but I could not find it easily, so I thought I would spare someone else the searching I did to get the information.] If you have a MicroVAX 3300 or MicroVAX 3400 with the RF disks, you have probably noticed that their integral controller comes with a name. The controller's name is rather odd, but must be unique in a cluster, so all RF disks are shipped with unique 6 character names pre-set. The names I have seen were similar to R3UIFB, R4SCST, R2RBTS, etc. These devices also come with an allocation class of 0, which must be changed to a non-zero value in order to be served by the MSCP server. This is necessary if the RF disk is to become a system disk for a Local Area VAXcluster. You can access the parameter setting mode of the RF disk controllers either at the console in console mode, or from VMS using the SET HOST/DUP command. From console mode, I did the following to change an RF disk's (R3UIFB$DIA0) allocation class to 1 and its nodename to RF. Note that the item name to change for the allocation class is "ALLCLASS" which is NOT spelled the same as the Sysgen parameter. >>> SET HOST/DUP/DSSI 0 PARAMS PARAMS> SET ALLCLASS 1 PARAMS> SET NODENAME RF PARAMS> WRITE The WRITE command takes almost a minute. Then the system asks whether you want to proceed with initialization of the controller to which I responded with YES. I then booted the system and both the allocation class and the name were changed. From the PARAMS> prompt, there is reasonably good HELP available. The same changes can theoretically be done from VMS. I have not done this, but the commands are: 1. First connect the FYA0 device using SYSGEN to allow use of the SET HOST/DUP command: $ RUN SYS$SYSTEM:SYSGEN CONNECT FYA0:/NOADAPTER EXIT 2. Set Host to the controller using SET HOST/DUP. In this example, the controller (node) name is R3UIFB: $ SET HOST/DUP/SERVER=MSCP$DUP/TASK=PARAMS R3UIFB PARAMS> SET ALLCLASS 1 PARAMS> SET NODENAME RF PARAMS> WRITE Learning how to change the name and the allocation class cost me about 3 hours on a Sunday during what was supposed to have been a two hour upgrade. This note is intended to spare you that effort. Hope it helps! ================================================================================ Note 484.1 Changing RF disk parameters 1 of 3 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 6 lines 3-OCT-1989 23:58 -< Thanks for giving the answer before the question >- -------------------------------------------------------------------------------- > < Note 484.0 by EISNER::GEIER "Jim Geier, General Instrument" > > -< Changing RF disk parameters >- Now that is real DECUServe service! I was going to ask that question in about three weeks, but didn't know it until I read your note. (I'll be installing a customer's 3300.) ================================================================================ Note 484.2 Changing RF disk parameters 2 of 3 EISNER::STERN "Seth Stern, (216) 266-6721" 6 lines 8-OCT-1989 17:39 -< Naming Conventions? >- -------------------------------------------------------------------------------- I had rather a fight with my first 3400. There seems to be some confusion in the documentation about whether the name you assign can have 6 or 8 characters. It is also not clear what characters are allowed. I finally settled on short alpha names since I got tired of having waiting for S/A backup to boot only to tell me that I didn't have any disk drives. ================================================================================ Note 484.3 Changing RF disk parameters 3 of 3 EISNER::MCCORMICK_B "Bob McCormick" 40 lines 14-OCT-1989 07:51 -< More thoughts on ISE configs >- -------------------------------------------------------------------------------- Both the KA640 and KA650 have pretty good help at the local console (>>>) prompt. Remember this for starters... On the KA640 we use the following: >>> SET HOST/DUP/DSSI n PARAMS PARAMS> SET ALLCLASS x PARAMS> WRITE On the KFQSA host adaptor we use: >>> SET HOST/DUP/UQSSP/DISK n PARAMS PARAMS> SET ALLCLASS x PARAMS> WRITE Where 'n' is the disk unit number and 'x' is the allocation class, which we always set to 1 on our non-clustered systems. The RF30's always come through with a node name, where the RF71's don't seem to. I tried to perform the function with the DCL SET HOST/DUP command, I think it was on VMS V5.0-2a, but couldn't get it to work properly. Have not tried since... Because the WRITE command resets the ISE I don't think its the kind of thing you want to be doing from a running system... We usually set the boot device to a specific drive. This is due to the fact that sometimes I've seen the system restart (like after a power failure) and skip over the drive, only to try tapes and/or Ethernet without success. Hardcoding the boot to a specific drive stops this problem. Use these commands: KA640 KFQSA (assume KA650 class system) >>> SET BOOT DIA0: >>> SET BOOT DUA0: Also note: We've seen a significant amount of problems on KA650 systems with more than one ISE when mounting them from SYSTARTUP_V5.COM. Inserting a 10 second wait prior to the mount seems be enough time for the ISE to be recognised.  ================================================================================ Note 485.0 DEC Goes into the Parts Business? 1 reply EISNER::RICE "Gary Rice" 15 lines 7-OCT-1989 08:00 -------------------------------------------------------------------------------- Last week, the power supply on my PRO380 died. Looking to obtain the lowest possible repair price, I contacted the local field service office for their opinion on the "way-to-go". During the conversation, the field service person that I was talking to told me that DEC Logistics was in the process of setting up a policy that would allow me (or anyone) to buy ANY part listed in the logistics parts list, to include sub-assembiles, ROMS, cables, boards, etc. No schedule information was available nor was any written document that described the proposed plan. However, I thought I would mention it so that you parts "bloodhounds" could "sniff" around for more information. Gary ================================================================================ Note 485.1 DEC Goes into the Parts Business? 1 of 1 EISNER::KENNEDY "Terry Kennedy" 8 lines 7-OCT-1989 17:31 -< BUSINESS_PRACTICES Note 19.0 >- -------------------------------------------------------------------------------- > During the conversation, the field service person that I was talking to > told me that DEC Logistics was in the process of setting up a policy > that would allow me (or anyone) to buy ANY part listed in the logistics > parts list, to include sub-assembiles, ROMS, cables, boards, etc. This is the new unlisted parts plan which took effect recently. Take a look in the Business_Practices conference (see this note's title for the exact note number) for a comment (positive, even!) on this new policy.  ================================================================================ Note 486.0 SQE and Transceivers as DESTAs. 2 replies EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 25 lines 11-OCT-1989 17:14 -------------------------------------------------------------------------------- Arrgh! We have a problem here, with some Ethernet connections to new (Sun) equipment. It seems that Sun-4 workstations currently come with only an AUI connector for Ethernet and are supplied (at least when purchased from Valid Logic) with a piece of AUI cable and a Cabletron transceiver, the kind with the AMP vampire tap. We are now sitting on an even dozen Sun-4/60 and 4/330 stations which need to be connected into our Ethernet -- which is a thickwire backbone and thin net distribution on each floor of our building (through DEMPRs and Cabletron MR-9000Cs). Some Ethernet equipment vendors specify the use of a transceiver (with BNC connectors replacing the vampire tap -- an option supplied by AMP) in places where DEC would specify a station adaptor (DESTA). I was very suspicious of this when first I heard of it, but we do have three instances on our network where this kludge is installed on one of our thin net strands, and I am amazed to report that they all appear to work. We have two questions: 1), is the repeater-as-station-adaptor approach *really* legitimate, or will there be potential problems later; and 2), in this approach, does the presence or absence of SQE 'heartbeat' in these devices have any significance? (Must or should we leave it on or off?) I have reviewed all of the 'heartbeat' notes that I can find, but can find little that bears on our problem: we have no DELNIs and will not. Can anyone help us here? We're all stumped. ================================================================================ Note 486.1 SQE and Transceivers as DESTAs. 1 of 2 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 36 lines 11-OCT-1989 18:09 -< DESTA <> DESPR >- -------------------------------------------------------------------------------- >Some Ethernet >equipment vendors specify the use of a transceiver (with BNC connectors >replacing the vampire tap -- an option supplied by AMP) in places where >DEC would specify a station adaptor (DESTA). This is totally 'ok'. A DESTA is just a thinwire only tranceiver. H4000s make dandy DESTAs if you buy the BNC head to replace the vampire one. You can also buy a head with two N connectors to fit between 2 pieces of EtherHose. The Cabletron unit is conveniently smaller and probably has nice LEDs. The BNC heads come in 2 flavors. 1 BNC connector, and 2 BNC conectors. The 2 BNC version implements an electrical T inside, and the distance from each BNC to that T is irrelevant (the T itself must be close to the actual chip). The 1 BNC flavor gets my vote because you get to use an external T which allows you to remove the tranceiver without killing the segment. Of course you can leave the 2 BNC head in the cable and remove it from the tranceiver, or connect an external T into one of its 2 BNCs, but the single BNC heads often have a central recessed BNC closer to the chip, and leave no second port for some turkey to plug something else into while your T is on the other side. >network where this kludge is installed on one of our thin net strands, Its not a kluge, its legal, but the spare BNC heads take a bit of looking to find. Some quote near $35 for them, but I bet you can get them for near $15 as an AMP part if bought through the right distributor, or even less if you qualify to buy directly from AMP at OEM pricing. >We have two questions: 1), is the repeater-as-station-adaptor approach I don't understand the question. Are you confusing DESTA (a tranceiver - Dec Ethennet STation Adapter) with DESPR (the one eyed DEMPR) - Dec Ethernet Single Port Repeater? ================================================================================ Note 486.2 SQE and Transceivers as DESTAs. 2 of 2 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 12 lines 12-OCT-1989 10:59 -< Oops. And thanks. >- -------------------------------------------------------------------------------- >> We have two questions: 1), is the repeater-as-station-adaptor approach ^^^^^^^^ Brain fade again, or something like that. This should be TRANSCEIVER- as-station-adaptor, of course. Sorry about that, Chief. I appreciate the information; the concept of the 'station adaptor' (DESTA) as being a thinwire-only transceiver clears up a whole lot of questions. (I therefore infer that SQE rules would be the same as for DESTAs.) Thanks again.  ================================================================================ Note 487.0 BA23 PINOUTS No replies EISNER::KILLEEN "Jeff Killeen DECUServe Chair" 3 lines 13-OCT-1989 08:35 -------------------------------------------------------------------------------- I just tripped over the print set for the BA23 System interconnection pinouts. It includes the routing signals from the RQDX to internal RD/RX drives. Send me a mail message if you need a photocopy.  ================================================================================ Note 488.0 How to add Quad Unibus controller 3 replies EISNER::JENSEN_M 28 lines 13-OCT-1989 23:46 -------------------------------------------------------------------------------- Very recently tried to install a quad unibus controller board into a Unibus on a VAX 8650. It's been a number of years since I had fooled with a Unibus. I chose an unoccupied slot in a 9-slot backplane after first removing the "flip-chip". I put the board in the rear four connectors. I didn't have a BA-11 book handy, but I thought that was the proper orientation. No such luck. The system wouldn't boot. Smelled that awful burnt elec. smell. After removing the board and replacing the flip-chip, no problem with a boot-up. After examining the apparently separate Unibus for the 8650's RL02 and seeing it's controller set to the rear as I had placed my board, I was/am confused as to what went wrong. Doesn't an SPC slot mean "most any" slot in a standard BA11-K ? The controller "was" a TD systems Unibus SCSI adapter and does (did) need DMA capability. OK, Unibus experts, I guess I blew this one ? And, NO, the component side was not facing the wrong way. Any help would be appreciated. Please hold the chuckles. ================================================================================ Note 488.1 How to add Quad Unibus controller 1 of 3 EISNER::KENNEDY "Terry Kennedy" 22 lines 14-OCT-1989 03:05 -< Console RL02 is Q-Bus, use *front* 4 for quad >- -------------------------------------------------------------------------------- > After examining the apparently separate Unibus for the 8650's RL02 and > seeing it's controller set to the rear as I had placed my board, I was/am > confused as to what went wrong. Surprise! The console RL02 backplane is Q-Bus. Just one of those DEC odd- ities... > Doesn't an SPC slot mean "most any" slot in a standard BA11-K ? Yes, given that you have a DD11-CK/-DK system unit. An easy way to remember this is that the quad slot you need to use is the part left over when you stick a Unibus cable or M9202 in the back dual slot. If you need NPR capability (DMA) you'll usually have to remove a wire-wrap wire in the backplane. It's a short one, and well illustrated in just about every DEC book dealing with backplanes. If you need more help, rather that try to explain it in words, I'll FAX you a diagram. If you see a dual G7273 in the middle 2 slots (instead of the scrape-your- knuckles G727), it may indicate that the slot has already had the backplane wire removed. Also note that some new Unibuses (Unibii? 8-) have DIP switches instead of the jumper wires (the 11/84, for example). ================================================================================ Note 488.2 How to add Quad Unibus controller 2 of 3 EISNER::JENSEN_M 9 lines 15-OCT-1989 00:10 -< Does Flip-Chip = DMA ? >- -------------------------------------------------------------------------------- Thanks for quick and helpful response. One further point if I may - All (most) of the unused slots in -DK (9 slot) had the single width Flip Chip in one of the middle connectors. Most of the open slots did have DZ32's in them. Think I still need to go after the NPR jumper ? ================================================================================ Note 488.3 How to add Quad Unibus controller 3 of 3 EISNER::HARVEY "Jack Harvey" 11 lines 15-OCT-1989 00:53 -< NPR Jumpers Exist by Default >- -------------------------------------------------------------------------------- >Think I still need to go after the NPR jumper ? Probably, but it's such a pain to find the jumper unless you do it every day, you might just plug in and try... I guess the real question is: What was in the slot before? If it was unused and had only the single-width buss grant card, then the jumper is almost surely there. If it was used by a DMA device, then it's almost surely *not* there. (Assuming of course, it's not a new-fangled type with a dip-switch :-)  ================================================================================ Note 489.0 CVAX-II ? 4 replies EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 27 lines 16-OCT-1989 10:31 -------------------------------------------------------------------------------- Here is one for all you SERIOUS hardware buffs ! We all know that the CVAX chip set that is used in the 3xxx and the 6xxx is made in CMOS, but using what level of technology (ie. 2.0 micron, 1.5, 1.0 ?) Is the second chip strictly for floating-point ? Just F and D formats, right ? How much cache is on the chip ? The reason I'm asking all these questions is look at the 80486 chip especially when compared to a '386/'387. It has floating point, memory mangement and cache all on on chip. And it's FAST ! Now I know DEC is no INTEL when it comes to IC process technology and they lag the IC industry by 3 to 5 years, but imagine what CVAX-II will do ! An interesting quote that I have picked up from several IC vendors lately. "We will always be able to build CPUs that will be faster than the fastest cost-effective memories. (Check pg 1 of Digital Review Oct 9, 1989; Solbourne is using GaAs RAMs and believe me GaAs isn't cheap!!) Because of this CISC architectures which can do many functions per instruction cycle will win out in the long run !" Sounds like VAX isn't dead yet ! ================================================================================ Note 489.1 CVAX-II ? 1 of 4 EISNER::SHANNON "Terry C. Shannon, erstwhile Matco" 10 lines 16-OCT-1989 21:01 -< VAX to the Future >- -------------------------------------------------------------------------------- < Note 489.0 by EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" > -< CVAX-II ? >- Sounds like VAX isn't dead yet ! Nope, DEC has publically stated that there are two more CVAXes (CMOS-3 and CMOS-4 implementations running at 10-12 and 20-25VUPS respectively) and two Aquarius followons (an ECL-based Aquarius II and the GaAs-based Centaurus) in the pipeline. Beyond those systems, however, the future of the VAX looks, well, RISCy. ================================================================================ Note 489.2 CVAX-II ? 2 of 4 EISNER::NAGY "Frank J. Nagy, VAX Guru/Loose Cannon" 32 lines 17-OCT-1989 07:24 -< Answers about the CVAX chip >- -------------------------------------------------------------------------------- >> Is the second chip strictly for floating-point ? Just F and D formats, >> right ? Yes, the CFPA chip is for floating point. It handles the F_Floating, G_Floating and D_Floating formats. >> How much cache is on the chip ? 1 KByte, two-way associative, write through with a 90 nsec cycle time. The translation buffer is 28-entry, fully associative. >> The reason I'm asking all these questions is look at the 80486 chip >> especially when compared to a '386/'387. It has floating point, >> memory mangement and cache all on on chip. And it's FAST ! The CVAX already has memory management on the chip (the 78032 MicroVAX chip also). The CVAX operates at 22 MHz and has about 180,000 transistors on the chip. >> Now I know DEC is no INTEL when it comes to IC process technology >> and they lag the IC industry by 3 to 5 years, but imagine what CVAX-II >> will do ! I remember reading someplace a year or so ago that Digital was one of the world's 10 largest IC manufacturers. The article went on to say that Digital concentrated on producing highly complex chips for their internal use (as opposed to zillions of memory chips) and was one of the leaders in IC technology. I admit that I was a bit surprised by the article at the time. ** Much of the information given above was taken from the KA650 CPU Module Technical Manual (EK-KA650-UG-001). ================================================================================ Note 489.3 CVAX-II ? 3 of 4 EISNER::PATTEEUW "Jack Patteeuw, Ford Motor Company" 17 lines 17-OCT-1989 10:45 -------------------------------------------------------------------------------- >> Nope, DEC has publicly stated that there are two more CVAXes Okay Terry, what "public announcements" do you get that the rest of the world doesn't. The comment about the '486 had more to do with having on board floating point. Estimate (from various IC vendors) indicate that on chip FP is 2 - 4 times faster than off chip (do to having to load/unload the co-processor). Also as a comparison, a '486 has about 1.2 million transistors and is made on a 1 micron process yielding a die that is approximately 500 sq. mils. I'm guessing that the current CVAX is made on 2 micron process. With only 180,000 transistors it seems that DEC could put FP and a MUCH bigger cache on one chip even using a 1.5 micron process. BTW the 1kbyte cache uses about 80,000 of those transistors (assuming a "dynamic" RAM cell design). ================================================================================ Note 489.4 CVAX-II ? 4 of 4 EISNER::SHANNON "Terry C. Shannon, erstwhile Matco" 37 lines 17-OCT-1989 22:05 -< Glad You Asked!!! >- -------------------------------------------------------------------------------- >> Okay Terry, what "public announcements" do you get that the rest of the world doesn't. I'm glad you asked that question! An April 1989 presentation entitled "VLSI VAX Technology," included comments to the effect that three generations of CVAX were under development (including the 7VUPS Rigel). By extrapolating density, number of transistors, pin count and CMOS-n fabrication process, one could reasonably conclude that the last two CVAX generations would come in at 10-12 and 20-25VUPS respectively. With respect to Aquarius followons, Bob Glorioso and other Digital participants in the 25 September Aquarius Unannouncement in Cupertino alluded to "future generations" of high performance systems. A comment about the future replacement of the HDSC "cold plate" heat sink with a water cooling jacket lends credence to the existence of an Aquarius followon. And another intriguing comment about gallium arsenide makes it seem likely to me that a GaAs-based system--which would be a third-generation Aquarius--is also planned. This latter bit o' speculation is, incidentally, supported by the fact that Digital has an equity position in GaAs purveyor GigaBit Logic. All of the foregoing is publically available information. Hope you didn't think I got the inside scoop from Charlie Matco*. *Whoever *she* is this week! ;-)  ================================================================================ Note 490.0 Physical data recovery services 1 reply EISNER::NORTON "Bill Norton" 25 lines 17-OCT-1989 15:16 -------------------------------------------------------------------------------- We had a RA81 HDA fail on us, and priced a data recovery service: Ramdomex 1100 E Willow St. Signal Hill, CA 90806 800-726-3669 (800-RANDOMX) I talked to Ray Rusabe. He said evaluation would be $700, during which they'd see if the data was readable after examining the platters for physical damage. If the platters were OK but the heads were bad they could use a set of external heads to extract the data. If the platters were damaged they could recover data from the undamaged areas. After the evaluation, they'd call to report on the recoverability and get permission to proceed. The actual recovery could cost $3K - $3500, which is a lot cheaper than re-doing several person-weeks of development work. The data would be returned on a VMS-formatted RA60. In our case the managers opted to retype their documents rather than have the disk recovered, so I don't have any results to report, but I figured others might want to know. ================================================================================ Note 490.1 Physical data recovery services 1 of 1 EISNER::MAYHEW "Bill Mayhew" 7 lines 17-OCT-1989 17:03 -< Another vendor >- -------------------------------------------------------------------------------- Another alternative is a company called COMP3, Inc. in Ohio. This company was formerly headed by my friend Terry Compton, who is, regrettably, no longer with us (he passed away suddenly, last Spring). However, as far as I know, they continue to do this type of work, and a couple of folks I know who've used them gave rave reviews. Phone: 614-876-3682.  ================================================================================ Note 491.0 DSD 880 Manuals Needed 2 replies EISNER::MCGLINCHEY "DECUS Board of Directors" 18 lines 17-OCT-1989 19:26 -------------------------------------------------------------------------------- HELP!! I am the, uh, proud owner of a DSD 880 box. It's an RX02 and and RL02 in one box. The RL02 is emulated via an 8-inch Winchester. It has a dual-height Q-Bus controller, and a spiffy-looking front panel with 'hyperdiagnostics'. (I always wonder when a manufacturer puts a lot of emphasis on diagnostics, as if they might be heavily needed in the future). What I need are manuals. Does anyone have a set for this box (I'd be willing to reimburse copying and shipping costs). Many thanks, Jim McGlinchey ================================================================================ Note 491.1 DSD 880 Manuals Needed 1 of 2 EISNER::CROWELL "Shefth of the Fourth Order" 4 lines 19-OCT-1989 01:01 -< The ol; packrat probably has it >- -------------------------------------------------------------------------------- The DSD880 came in three flavors over the years (with different capacity Winchesters). I may still have the manual for the small one (7.8 MB) somewhere. It'll at least describe how to use the 'hyperdiagnostics.' I'll find your address in WHOI_AM_I. ================================================================================ Note 491.2 DSD 880 Manuals Needed 2 of 2 EISNER::MCGLINCHEY "DECUS Board of Directors" 5 lines 25-OCT-1989 17:39 -< Packrats, unite! >- -------------------------------------------------------------------------------- Thanks, John! Jim  ================================================================================ Note 492.0 Booting Diagnostics from disk 3 replies EISNER::SPAIN "Harrison Spain - spain@mdcbbs.com" 17 lines 18-OCT-1989 21:17 -------------------------------------------------------------------------------- Has anyone been able to move the MicroVax Diagnostics from tape to disk? I would like to eliminate the boot time from tape similar to placing Standalone BACKUP on disk. I created a directory: disk:[SYS1.SYSEXE] Then I mounted the diag tape (MOUNT /OVER=ID) and copied the files to the disk:[SYS1.SYSEXE] directory. I then tried to boot: >>> B/10000000 disk: No luck :-(. ================================================================================ Note 492.1 Booting Diagnostics from disk 1 of 3 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 7 lines 19-OCT-1989 00:35 -< I need Tape-Tape before Tape-Disk >- -------------------------------------------------------------------------------- > Has anyone been able to move the MicroVax Diagnostics from tape > to disk? First, I would like to know how to copy the darn tape! VMSTPC can't do it because there is something on it with a record > 32k. Our F/S guy hasn't yet got a spare he can leave here, but would love it if we could make a copy for him. ================================================================================ Note 492.2 Booting Diagnostics from disk 2 of 3 EISNER::NORTON "Bill Norton" 11 lines 19-OCT-1989 09:35 -< Possible diag tape copy method >- -------------------------------------------------------------------------------- I used to copy the tapes like this: 1. $Mount/over=id/nowrite mua0: 2. $Copy/log=future_command_file mua0:*.* [some_empty_directory] 3. Edit the log into a command file to copy back to tape, in the exact same order. 4. $INI/OVER=(ACC,EXP) MUA0:same_label_as_original 5. $MOUNT/BLOCK=512 MUA0: label 6. $@command_file_from_step_3 It's been a couple years since I've seen any tapes to practice on, but it might still work. ================================================================================ Note 492.3 Booting Diagnostics from disk 3 of 3 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 6 lines 19-OCT-1989 09:59 -< Yup, it's the order that's important! >- -------------------------------------------------------------------------------- As the .-1 note implied, you have to put the files back in the same order you took them off the original tape. I have a command procedure that does essentially what's done in .-1. It's overkill if you only plan on doing this once in a while, but you're welcome to it... Ed  ================================================================================ Note 493.0 New firmware for SI 93C drives 2 replies EISNER::WOOD "Jim Wood (609)423-1040" 10 lines 18-OCT-1989 22:02 -------------------------------------------------------------------------------- System Industries just installed new PROMs in our model 93C 842MB drives. The new firmware will (hopefully) solve a burst mis-seek problem that is crashing our HSCs approximately once per week. (The HSCs **are** crashing; the analysis of the problem courtesy of SI) The three new PROMS are labeled: 8700-8939 8700-8940 8700-8935 REV A REV A REV A LOC L LOC U ================================================================================ Note 493.1 New firmware for SI 93C drives 1 of 2 EISNER::DELARISCH "Arnold S. De Larisch" 12 lines 18-OCT-1989 23:07 -< Does this effect Non-HSC systems? >- -------------------------------------------------------------------------------- >> < Note 493.0 by EISNER::WOOD "Jim Wood (609)423-1040" > >> -< New firmware for SI 93C drives >- >> System Industries just installed new PROMs in our model 93C 842MB drives. >> The new firmware will (hopefully) solve a burst mis-seek problem >> that is crashing our HSCs approximately once per week. (The HSCs >> **are** crashing; the analysis of the problem courtesy of SI) Is this 'fix' wanted/needed on Non-HSC uses of this drive? -Arnold {Who has all his Boss' files on an SI 93C} ================================================================================ Note 493.2 New firmware for SI 93C drives 2 of 2 EISNER::WOOD "Jim Wood (609)423-1040" 4 lines 19-OCT-1989 21:10 -< SI 93C prom upgrade for non-HSC drives >- -------------------------------------------------------------------------------- >> -< New firmware for SI 93C drives >- >> -< Does this effect Non-HSC systems? >- Sorry, Arnold, I don't know. We're all HSC-based here.  ================================================================================ Note 494.0 Know about ZR-TS05Z-AA for MV3100? No replies EISNER::STRIEGEL "Alan Striegel" 25 lines 19-OCT-1989 10:36 -------------------------------------------------------------------------------- Has anyone else gotten information on this product? I have heard from two sources inside Digital that this product is on the way, but the last I heard from CSS was that it was not due until January or so. The following is taken from a message I got from someone inside our company. Can anyone answer the questions with authority? >> From: 9-OCT-1989 11:55 >> To: STRIEGEL, >> Subj: SCSI DRIVER FOR F880 >> >> ALAN, >> >> x from Tulsa called today with some interesting info. He said that DEC CSS >> now sells the TS05 package for the MV3100. The part number is ZR-TS05Z-AA >> and sells for $8979. This includes the tape drive (desktop enclosure), >> cabinet kit, and SCSI driver. >> >> He also said that as of VMS version 5.2, the SCSI driver for the magtape is >> built in? Do you know if this is true? I've been getting a lot of requests >> for a 9 track tape on a 3100 and have not been able to help anyone. Any real written information sources from a Digital publication would be most appreciated. Alan  ================================================================================ Note 495.0 RA81 puzzlements. 8 replies EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 34 lines 19-OCT-1989 11:52 -------------------------------------------------------------------------------- I am somewhat ashamed to post this one, but here goes... As a result of our company's ever-ending pursuit of the leading edge of technology, we are now "upgrading" some of our RP07's to RA81's. Since we obtain these from the used market, I'm not sure that we are in fact getting all of the documentation that applies to these products, but we have looked through that which we did get and find nothing that applies. We recently installed two RA81-AA's (late models, 1986 manufacture dates) and a UDA50-A into our 780, and find that we consistently cannot mount them cleanly at boot time. Whenever we (in SYSTARTUP) say "$ MOUNT/SYSTEM DUA0: DISK6", the system responds with "MOUNT-F- NOSUCHDEV no such device available". If we stick in a "$ SHOW DEV" just before the MOUNT commands, everything works properly. We are not sure whether it is the additional time taken or the 'access' of the devices by the SHOW process that does the job (and I cannot in good conscience continue to take the machine up and down for a lot of such testing). Our question here is: why does VMS not recognize the RA81's here? Do we need to allow some more time to elapse after SYSGEN/AUTOCONFIG runs, or do we need to do some sort of specific 'access' to the drive before we can mount it? Where should we look to find out the information about this? What are we doing wrong in setting up our drives and/or startup procedures? Question 2. We now always show a PUA1: device, which always shows two errors. Our DEC persons tell us that this is normal, but cannot tell us why. What *is* the PU device, and why does it get errors? Yes, this is all old stuff -- or should be -- but we're stumped, we can't find what we need anywhere in the docs we have got. Can anyone help us? ================================================================================ Note 495.1 RA81 puzzlements. 1 of 8 EISNER::ALDRICH "Jeff Aldrich" 9 lines 19-OCT-1989 12:07 -< Contrary to user belief, computers aren't instaneous. >- -------------------------------------------------------------------------------- We had RA81's through a UDA50 to a 780 (now HSC). We had an 8 second WAIT before mounting our disks and never had any problems. I'm ashamed to admit this but this wait was there when I got here and I'm not sure that it is there to fix the problem you describe. We did have a problem like that with our 6210. The KDB50 on the second BI needed some extra time to get it's life in order. As I remember, 30 seconds did it. ================================================================================ Note 495.2 RA81 puzzlements. 2 of 8 EISNER::BAILEY_E "Ed Bailey - WHO_AM_I #287" 30 lines 19-OCT-1989 13:58 -< UDAs log errors, 2 (do-dah, do-dah) >- -------------------------------------------------------------------------------- >< Note 495.0 by EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" > >Our question here is: why does VMS not recognize the RA81's here? >Do we need to allow some more time to elapse after SYSGEN/AUTOCONFIG >runs, or do we need to do some sort of specific 'access' to the >drive before we can mount it? Where should we look to find out the Basically, the problem is that it takes time for the UDA to figure out A: what it has for drives attached to it, and B: communicating this information back to the system. This same problem can be seen on a cluster with HSCs. You can put a WAIT command in your startup file, but a better method would be to use F$GETDVI("DUA0","EXISTS") to determine if the drive has been configured. If it's there, mount it immediately. If not, try another drive, and loop around until all the drives exist and have been mounted. >Question 2. We now always show a PUA1: device, which always shows >two errors. Our DEC persons tell us that this is normal, but cannot >tell us why. What *is* the PU device, and why does it get errors? PUA1 is your UDA. Yes, the two errors are normal. Depending on who you talk to, the errors are either informational messages sent to VMS when the UDA is being initialized, or they are "real" errors that always occur when VMS gets the UDA running; VMS has to nudge the thing several times to get it going, and it always takes three tries to get it going. I don't know which is true (if either), but they are normal... Ed ================================================================================ Note 495.3 RA81 puzzlements. 3 of 8 EISNER::SHUMAKER "Mark L. Shumaker -{The Arachnid}-" 8 lines 19-OCT-1989 15:06 -< Thanks. (And another question.) >- -------------------------------------------------------------------------------- Thanks for the information, it has helped us a lot. Another question has been posed here -- in a non-clustered 780 with some DSA drives and some Massbus drives, is it necessary to use the MSCP option in any of the SYSGEN commands at boot time (or any other time) to talk to the RA81's? It doesn't seem to be required, but the descriptions in the documentation are sufficiently confusing that we can't figure out when it might be required. ================================================================================ Note 495.4 RA81 puzzlements. 4 of 8 EISNER::FULLER_S "All stressed up + no one to choke" 24 lines 19-OCT-1989 16:18 -< Here's why the 2 errors >- -------------------------------------------------------------------------------- > PUA1 is your UDA. Yes, the two errors are normal. Depending on >who you talk to, the errors are either informational messages sent to VMS >when the UDA is being initialized, or they are "real" errors that always >occur when VMS gets the UDA running; VMS has to nudge the thing several >times to get it going, and it always takes three tries to get it going. There should always be 2 errors on PUA0. Don't know what PUA1 is, tho. When the bootstrap runs, it has to set aside an area of memory for a ring buffer to talk to the UDA. After the bootstrap has loaded the various pieces of VMS into memory, it calls an initialization routine which brings up VMS. Now, VMS is going to use a different area of memory than the bootstrap, and the only way to tell the UDA about it is to reset it, which it does. Now, then, the UDA is reset and wants to tell VMS about it. First, it says "I was reset", and then it says "I'm done initializing". Each event is written to the error log, and the "error" count is incremented. Stu ================================================================================ Note 495.5 RA81 puzzlements. 5 of 8 EISNER::FULLER_S "All stressed up + no one to choke" 13 lines 19-OCT-1989 16:21 -------------------------------------------------------------------------------- >Another question has been posed here -- in a non-clustered 780 with >some DSA drives and some Massbus drives, is it necessary to use the >MSCP option in any of the SYSGEN commands at boot time (or any other >time) to talk to the RA81's? No. The MSCP "stuff" in SYSGEN is to tell your system to "serve" the drive(s) to other systems in a cluster. In effect, your VAX becomes an HSC disk controller that also supports users. Stu ================================================================================ Note 495.6 RA81 puzzlements. 6 of 8 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 6 lines 19-OCT-1989 23:02 -< Yeah.. fooled us for a while. >- -------------------------------------------------------------------------------- It probably would help if you called the PU errors "status messages" instead. Problem is VMS calls everything it gets from devices an 'error'. We've had the f$getdvi code in our startup code for a long time. "Does it exist yet? No? Wait 2 seconds and try again." ================================================================================ Note 495.7 RA81 puzzlements. 7 of 8 EISNER::FULLER_S "All stressed up + no one to choke" 15 lines 20-OCT-1989 00:01 -------------------------------------------------------------------------------- > It probably would help if you called the PU errors "status messages" > instead. Problem is VMS calls everything it gets from devices an > 'error'. Next time VMS Engineering calls me, I'll be sure to mention it to them :^) > We've had the f$getdvi code in our startup code for a long time. > "Does it exist yet? No? Wait 2 seconds and try again." I would hope there is a timeout mechanism in there, to handle the case of the drive never coming up. Stu ================================================================================ Note 495.8 RA81 puzzlements. 8 of 8 EISNER::SCOPELLITI "Whatsa behind is uva no importan" 2 lines 24-OCT-1989 00:01 -< Oh! Yeah, it's there... forgot to mention it. >- -------------------------------------------------------------------------------- >>> I would hope there is a timeout mechanism in there, to handle the >>> case of the drive never coming up.  ================================================================================ Note 496.0 Tell me what it is 1 reply EISNER::TILLMAN "Brian Tillman, DECUServe MoS" 2 lines 24-OCT-1989 12:31 -------------------------------------------------------------------------------- Page 18-3 of the VMS V5.2 Device Driver manual talks about 8600/8650/8670 systems. What the heck is an 8670? ================================================================================ Note 496.1 Tell me what it is 1 of 1 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 6 lines 24-OCT-1989 13:11 -< What I heard in a hallway >- -------------------------------------------------------------------------------- > Page 18-3 of the VMS V5.2 Device Driver manual talks about > 8600/8650/8670 systems. What the heck is an 8670? I believe it is a performance "kicker" that was never released to the public. Would somebody with better information care to comment?  ================================================================================ Note 497.0 DPV11 Knocks out DHV11 2 replies EISNER::GRATTAN "Michael Grattan (Who #316)" 45 lines 24-OCT-1989 15:12 -------------------------------------------------------------------------------- I am having a problem installing a DPV11 in our uVAX II. The uVAX is configured as follows: Slot # Device ------ ------ 1 CPU 2 8Mb memory 3 8Mb memory 4 (AB) DEQNA 5 DHV11 6 DHV11 7 DHV11 8 (AB) TQK50 8 (CD) RQDX2 When I put the DPV11 in slot 4 (as recommended by the technical manual) and move everything back, any terminal connected to the DHV11's no longer work. I am able to access the DPV11 and make it work. I can do things like SET TERM/PERM/SPEED=(4800) for a terminal on the DHV11. However, I cannot do a SET TERM/PERM/DEV=VT200 (comes back with DEVICE TIMEOUT) and striking the RETURN key on any DHV11 terminal produces nothing but frustration. (The system does not come up with USERNAME:.) SYSGEN shows the following as my configuration: System CSR and Vectors on 24-OCT-1989 14:31:39.70 Name: PUA Units: 1 Nexus:0 (UBA) CSR: 772150 Vector1: 774 Vector2: 000 Name: PTA Units: 1 Nexus:0 (UBA) CSR: 774500 Vector1: 260 Vector2: 000 Name: XQA Units: 5 Nexus:0 (UBA) CSR: 774440 Vector1: 120 Vector2: 000 Name: TXA Units: 8 Nexus:0 (UBA) CSR: 760440 Vector1: 300 Vector2: 304 Name: TXB Units: 8 Nexus:0 (UBA) CSR: 760460 Vector1: 310 Vector2: 314 Name: TXC Units: 8 Nexus:0 (UBA) CSR: 760500 Vector1: 320 Vector2: 324 I am trying to put the DPV11 at: CSR: 760270, Vector: 340. Can anyone point me in the right direction? I would geatly appreciate any assistance. Mike Grattan PS: Yes, I have gone to CSC in Atlanta with this one but so far ...  I have been refered to Colorado ...  Who has refered me to Massachusetts...  Who has refered me back to Atlanta! (insert BIG sigh here) ================================================================================ Note 497.1 DPV11 Knocks out DHV11 1 of 2 EISNER::GRANZOW 12 lines 24-OCT-1989 18:22 -< WRONG VECTORS >- -------------------------------------------------------------------------------- < Note 497.0 by EISNER::GRATTAN "Michael Grattan (Who #316)" > -< DPV11 Knocks out DHV11 >- YOUR PROBLEM IS THAT SYSGEN WANTS TO SEE THE VECTOR OF THE DPV11 AT 300 AND THE DHV11'S AT 310,320,330.... THE SYSTEM SEE'S THE ADDRESS OF THE CONTROLLER AND CAN DO BASIC FUNCTIONS BUT WHEN A INTERRUPT OCCURS IT GO'S TO THE WRONG ADDRESS, THERE-BY TIMING OUT. YOU HAVE TO PHY. CHANGE THE VECTOR SWITCHES ON THE DHV11'S AND PUT THE DPV11 VECTOR AT 300. I ALSO ASSUME THAT YOU ARE PUTTING THE BOARD INTO SLOT 4 CD WITH THE DEQNA IN SLOT 4 AB OR HAVE A GRANT CARD IN ANY BLANK AB/CD. ================================================================================ Note 497.2 DPV11 Knocks out DHV11 2 of 2 EISNER::GRATTAN "Michael Grattan (Who #316)" 26 lines 25-OCT-1989 15:03 -< Thanks! >- -------------------------------------------------------------------------------- > < Note 497.1 by EISNER::GRANZOW > > -< WRONG VECTORS >- >YOUR PROBLEM IS THAT SYSGEN WANTS TO SEE THE VECTOR OF THE DPV11 AT 300 >AND THE DHV11'S AT 310,320,330.... Gee, you were right on! I made the changes you suggested and voila! For future reference (for any system manglers still learning, like me!) I was able to go into SYSGEN and then into CONFIGURE, plug in the devices (and how many) that I wanted on my system, ^Z and SYSGEN told me the correct CSR's and vectors. I'm so embarrassed!! >I ALSO ASSUME THAT YOU ARE PUTTING THE BOARD INTO SLOT 4 CD WITH THE DEQNA IN >SLOT 4 AB OR HAVE A GRANT CARD IN ANY BLANK AB/CD. Yup, DPV11 in slot 4 AB and DEQNA in slot 5 AB with a grant card in slot 5 CD. Thanks for your fast reply. I am constantly amazed and grateful for the speed and accuracy of the information one can get here at DECUServe. Almost makes me wonder why we're sending DEC those checks every month. ;-)  ================================================================================ Note 498.0 3500 CPU "Enhancement"? No replies EISNER::FERGUSON "Linwood Ferguson, ARA Services" 12 lines 25-OCT-1989 07:25 -------------------------------------------------------------------------------- Dec showed up recently at one of six 3500 sites with a new CPU board saying it was an "enhancement" that they had to install. We had not asked for it, and as far as I know that site had no problems, so it sounded like a required FCO. Asking FE what it enhanced got nowhere. Nor (by the time I found out) did I get any kind of number associated with it. Word I have (about 4th hand) is they swapped the CPU board itself. Anyone know of any changes to the 3500's that Dec is volunteering to install? And what it's for? And should I worry that only one site got it?  ================================================================================ Note 499.0 Micro-11 Clock Crystal 2 replies EISNER::CHRISTIAN "Michael Christian" 17 lines 25-OCT-1989 15:32 -------------------------------------------------------------------------------- This note should only be read for those of you with strong stomaches ;). We are using Micro-11s (23/73/53) for real-time process control. A few years ago, we modified some of our 11/73 processors and increased the clock crystal from 15 MHz to 18 MHz. We haven't seen any problems at that speed, and we get better throughput. Unfortunately, the people that did the original research on the crystal speed are no longer with the department/company. :( We are considering increasing the crystal even higher and also modifying some 11/53 processors. They both use the same J11 chip, but the 11/73 had cache/Qbus memory while the 11/53 has on board PMI memory. Has anybody else experimented with Micro-11 clock crystal, and if so would you be willing to share your results? ================================================================================ Note 499.1 Micro-11 Clock Crystal 1 of 2 EISNER::FRISBIE "Alan Frisbie - Flying Disk Systems" 25 lines 25-OCT-1989 17:38 -< Faster PDP-11 CPUs, but not from DEC >- -------------------------------------------------------------------------------- > Has anybody else experimented with Micro-11 clock crystal, and > if so would you be willing to share your results? Scott Taylor of Douglas Engineering has done several experiments, which he reported in the SIG newsletters. One of his tests (on memory speeds) was in the March '87 HMS section. He says that most 11/23 boards can be boosted to 20MHz, making them faster (for most purposes) than an 11/73. I don't believe he has taken an 11/73 beyond 18 MHz. At one time, Chrislin (memory manufacturer) made an offer to upgrade 11/73 boards when one of their memory boards was purchased. In the September '89 issue of the RT-11 newsletter, Scott compared the instruction (including floating point) performance of several CPU boards. The big winner was the Mentec M901 (SBCM90 w/1MB memory) with a J11 running at 20 MHz. If you opt for the M904 (4MB memory on board), the clock drops to 18.5 MHz because they couldn't get sufficient quantities of the large memory chips that would meet the speed specs. Mentec is the only company I know of that sells CPU boards with the J11 chip. They are located in Ireland and are represented in the US by Andromeda Systems in Canoga Park, CA. ================================================================================ Note 499.2 Micro-11 Clock Crystal 2 of 2 EISNER::BRUCE "Barton F. Bruce - CCA/MA" 52 lines 25-OCT-1989 18:09 -< a few 11/73 speedup thoughts >- -------------------------------------------------------------------------------- > A few years ago, we modified some of our 11/73 processors and > increased the clock crystal from 15 MHz to 18 MHz. We haven't > seen any problems at that speed, and we get better throughput. A lot of this is posted here (somewhere), but here goes again. This is not exactly what you asked for, but IS speedup related. Were those the dual or the quad (-Bx) cards? Most crystal changes I heard of were on the dual wide cards, but those with the quad cards can do another type of hop-up. The ONLY difference between a QUAD 73 and 83 is the crystal, and maybe a FPP chip (that is almost true, but close enough for the following). The rest of the speed advantage the 83 has comes from PMI memory on the Q bus. This has additional signals on the C-D fingers, and MUST PRECEED the CPU and be in a Q-CD backplane. You can get all 4 megs of PMI memory on a single Clearpoint QED-2 card, and, since it can bank select through a variable size window, can support MORE than 4 meg. Before you have to have them blast custom PLAs (which they will gladly do), the 16 standard memory CSRs support 64 MEG! Of course you would need at LARGE Q-CD backplane. Clearpoints QED-2 can also be vanilla Q bus memory as there are jumpers that when yanked keep one from smoking in a Q-Q slot. It costs more than dumber Q only memory, but is FAST (ECCs 64 bits at a whack, and caches a couple such fetches so DATA fetches don't flush INST stream fetches and vica-versa). Using DEC PMI memory, DON'T use the MSV11-JB or -JC (would 'work', but lacking BLOCK-MODE-DMA should only be used on 11/84 Unibus boxes), but do use the newer -JD or -JE (good for 73/83/84s). > Has anybody else experimented with Micro-11 clock crystal, and if > so would you be willing to share your results? There is a DECUSERVE moderator/contributor running J11 chips in the UPPER 2x MHZ (or possibly higher) range BUT not on a DEC board. The Nissio Unibus cards are layed out differently and won't fail at those speeds. Some of his comments on this are somewhere here. Do a DIR *.*/aut=KENNEDY in HARDWARE. There is an Irish company that WON a suit and DEC has to supply them with 18MHZ J11 chips. Some of their Q bus boards make it to the North American market, and you could ask about pushing THEIR card. If you are going to Anaheim, there are some DECies worth asking that WILL be there, but whose names best NOT be posted here.  ================================================================================ Note 500.0 What is a VAX 8670 ??? 1 reply EISNER::DAINELIS 2 lines 25-OCT-1989 16:21 -------------------------------------------------------------------------------- I read in the DEC device driver manual about a VAX 8670 processor. Could anyone tell me what it is or where to look? ================================================================================ Note 500.1 What is a VAX 8670 ??? 1 of 1 EISNER::MAYHEW "Bill Mayhew" 1 line 25-OCT-1989 20:17 -< See topic 496 in this conference >- --------------------------------------------------------------------------------