• Please review our updated Terms and Rules here

HP 1000/A900, with 7937 HDD

roberttx

Veteran Member
Joined
Aug 6, 2015
Messages
1,133
Location
Texas
I recently acquired two of each - when they were launched in 1982, the feds had a list of 16 countries that they could be sold to. Anywhere else and it was The Inquisition. I must start by giving credit to the HP Computer Museum. It has good write ups, excellent "Cliff's Notes" on testing and a comprehensive archive of documentation. That last turned out to be very important, as I shall relate.

It all came in an HP 19" rack. Space is tight and I've learned the folly of placing racks in my storage container, which is on the other side of town from my workshop and has no power. So, I deemed it best to remove the boxes from the rack on site.

The computers were not a problem. The front panels didn't want to pop off easily, but a little judicious leverage with a padded screwdriver and off they came. Four screws and they slid right out and were easily lifted onto some wheeled Metro shelving. This was looking easy - except I was wrong.

Turning my attention to the first of the HDDs, I removed screws and slid it out on its rails (it had been fitted using the 19" adapter kit). I then started looking for the little buttons, levers or whatever to disengage part of the rail from the rest - but couldn't find them. So, I brought a ridiculously powerful LED light into play, as an aide to my old eyes. Still nothing.

In desperation, I turned to the HP Computer Museum site and downloaded the Installation and Operation Manual, to my phone. I had already looked at it, at home, to look for anything about locking heads for transit. The good news is that, not only were the heads already locked for transit but whomever retired the drives put a little post-it note on each reading "heads locked". So, I can be reasonably confident that the heads have been locked since it was last powered down.

The bad news was that I failed to read the rest of the manual. It turns out that the slides for the rails don't separate. Instead, HP sold a special crane, that clamped to the top of the rack. To install a drive you first installed the adapter kit, then installed the rails. then you used the crane to winch up the drive, fully extended the rails, carefully winched the drive down and, when it was lined up, screwed the rails to the drive. Reverse the process to remove.

I didn't have the special crane, of course. Fortunately, the place that I was getting them from had an abundance of wheeled transports and lumber. With some carts, some 2x4s, some 4x4s and, in one case, some very robust cardboard, with a bit of fettling, I was able to place suitable wheeled platforms under each of the drives, enabling me to remove the screws fastening them to the rails and slide them out without them dropping and suffering impact shock. But, man, those things were heavy!

As things stand, the computers are on the back seat of my truck and the drives are still on site. I'll get them next week. The computers have all of the base cards, plus some I/O (HPIB, Serial, one of them has an Ethernet card) so seeing whether they pass a self test shouldn't be too onerous. But it might take a while. I've been wanting to get back to my pdp11/34, which will need space to work in. And much of that space is occupied by other projects (like a PET and an IBM 5110 that really should have first call on my attention). So, I may have to move backwards, in order to move forwards.

I will try to post some photos, though, as and when I can.
 
Last edited:
I have two HP 1000 A900 boxes myself. I acquired them in bits and pieces, not as completely configured systems.

I don't have any HPIB interface tape drives. Eventually I'll get around to trying to install RTE-A from an HPDrive emulated tape drive. There should be complete installation tape images on:

www.bitsavers.org/bits/HP/HP_1000_software_collection/

I have some 12005 async boards, 12009 HPIB boards, 12040 async mux boards, a couple of 12076 Ethernet boards, and even a couple of 12016 SCSI boards.

I have a couple of 3MB boards, an 8MB board, and a 3rd party 4MB board that looks like it can be upgraded to an 8MB board but don't have any documentation for that board. I only have one single memory board frontplane connector (12222-60001) and one two memory board frontplane (12222-60002) which limits the memory configuration options. If anyone has a source of A900 12222-6000x memory frontplane connectors I'd like to know.

I can power up to the VCP console and run self tests on both boxes. That is a far as I have ever gotten with them so far.
 
I think that both of mine have 3 board frontplanes, but I'll double check later.

It's going to be a while before I get to playing with them, but I figure it can't hurt to make a start on figuring out the procedure.

I guess the first step is to document the locations of all the boards, then attempt powering it up with a minimal board set. If it passes the self test, reinstall the boards one at a time and try again. Once I'm back to the current configuration, I'll need to get into it, somehow. I think that both have serial boards and one also has ethernet. In terms of cabling, I got one edge connector to HPIB, one edge connector to AUI, marked ethernet and one edge connector to an 8 port serial breakout box, along with some plain old HPIB cables. All of them had been disconnected, so I have no idea what the original cabling setup looked like.

It seems likely that, if the serial boards are mux ones, access will be via the breakout box with a terminal emulator. I foresee some trial and error there. If I can get past that, It may be time to look at the hard drives. It seems that the first step there is to unlock the heads, power it on and see what happens with the self test. If it passes, connect it via HPIB and power on the drive first, then the computer. I'll have to do some reading to see what comes after that.

I don't know which computer was connected to which drive, so it might be a good idea to try each drive that I test with both, in case a bootstrap loader does anything interesting. It will be easier to swap the computers, which are much lighter than the drives. I have no idea how to address the ethernet board.

I'll try to get some photos, this morning, when I unload my truck. It's going to be a long time before I do anything with the drives, at least. Due to their size and weight I'm going to need space and a very robust cart. That will mean clearing the decks of all the other projects that are in the way.

I had been thinking of abandoning the rack, but now I think I might save it. There's space in my storage container and I can always abandon it later, if necessary.
 
For years I worked at a Chemical Manufacturing plant in their Quality Control Lab doing chemical analysis. HP was always big in Engineering and analytical systems.

We had an HP1000\A600 for a good while.

The system ran on RTE-1000 which is very Unix like. The application it ran was a data acquisition and reporting system called LAS (Laboratory Automation System).
 
Here's a few pics. The two A900s have consecutive serial #s and have the same boards in the same order, with the exception of the LAN board, which only the top machine has. From the right, they are:

HP IB, Sequencer, Data Path, Cache Control, Memory Controller, Memory 3MB, Memory 3MB, HP IB, ISO I/O (or 1SO I/O, I couldn't make it out), 8 Chan Mux, HP IB, LAN.

The faceplates are in decent shape and the breakout box is an HP 28658-60005 RS-232C Panel.

a900a.jpg

a900b.jpg

a900c.jpg

a900d.jpg

a900e.jpg
 
Thanks! Probably just a terminal emulator to start with. I have an HP 2393A, but no keyboard or monitor and I think it's one that doesn't play nice with regular monitors.

As part of my pdp11/34 project, I plan on putting a terminal emulator on some flavor of pi and incorporating it in one of those 1U keyboard and LCD monitor deals. I have around three of them, so might end up making another for the HP, if I end up keeping the rack. Or, the HP might end up going in the rack with the DEC, who knows?

I'm not looking forward to moving the 7937 hard drives, though! I might need to get some help...
 
There are keyed pins on the connectors of the four board set that makes up the CPU. The pins are half circles and are rotated differently between each board so they should only mate with the correct backplane slot. They also go into the slots from right to left in their board number sequence:

12201 Sequencer Card
12202 Data Path Card
12203 Cache Control Card
12204 Memory Controller Card

The first I/O card needs to go in the rightmost slot next to the 12201 Sequencer Card. The only exception is if the optional 12205 Control Store Card is present. If the 12005 Control Store Card is present it needs to be installed next to the 12201 Sequencer Card and connected to it by a 12205-60002 frontplane connector.

Good thing you have the 28658-63005 RS-232C MUX Panel Cable and the 28658-60005 RS-232C Panel for the 12040 async mux board, and also a 12076-63002 Ethernet Stub cable for the 12076 Ethernet board. The cables are usually more difficult to find and can be more expensive than the boards.

Do you have two 12040 async mux board, cable, and panel sets? If you only have one then you either need another one of those for the console for the second system, or a 12005 async board.

Either a 12005 async board or a 12040 async mux board can be used for the VCP console interface. If the 12005 U1 Switch 1 is open the 12005 will not operate as the VCP interface. If the 12005 U1 Switch 1 is closed the 12005 will operate as the VCP interface. If the 12040 U1 Switch 1 is open Port 0 of the 12040 will not operate as the VCP interface. If the 12040 U1 Switch 1 is closed Port 0 of the 12040 will operate as the VCP interface.

There were several revisions of the 12040 async mux board layout and firmware versions. I have boards with these firmware versions as labeled on the 12040 board EPROM stickers:
5180-1970 12040B
5180-7228 12040C
5180-7268 12040D, Revision 5.02
The last version I see listed in the manual is 5180-7300 or 5181-8663 (same firmware, different EPROM vendor) 12040D, Revision 5.20.

There were also several revisions of the VCP firmware, which is located in EPROMs on the 12203 Cache Control Card. I have these firmware versions as labeled on the 12203 board EPROM stickers:
12203-80009 / 12203-80010 Revision 4004
5180-4271 / 5180-4272 Revision 4020
5181-8680 / 5181-8681 Revision 4025

I don't know if anyone has bothered to archive images of the different versions of the firmware EPROMs. If no one has already done that it would be nice to do so. What firmware versions do you have on you 12040 and 12203 boards?

If you haven't already found these manuals you need to grab copies of them to start with. These might also be on the hpmuseum.net site as well.
www.bitsavers.org/pdf/hp/1000/RTE-A/A-Series_CE_Handbook.pdf
www.bitsavers.org/pdf/hp/1000/A-series/02139-90001_A900ref_Oct86.pdf
www.bitsavers.org/pdf/hp/1000/A-series/02139-90003_A900_ERD_Apr86.pdf
www.bitsavers.org/pdf/hp/1000/A-series/02196-90002_1000_26_27_29_Inst_May85.pdf

There is also some interesting information buried in these documents:
www.bitsavers.org/pdf/hp/communicator/1000/5955-3257_Mar-1984.pdf
www.bitsavers.org/pdf/hp/communicator/1000/5961-6201_Dec-1992.pdf
 
Thanks! Probably just a terminal emulator to start with. I have an HP 2393A, but no keyboard or monitor and I think it's one that doesn't play nice with regular monitors.

You might want to try the QCTerm 700/92 Emulator. I haven't gotten around to trying that myself yet.

www.hpmuseum.net/display_item.php?sw=585

I have the color version of the HP 2393A, the HP 2397A. I forget if I was able to get any of the LCD or CRT displays I have to work with it. I was able to get an Epiphan VGA2USB LR configured to sync with it for video capture. (Actual capture is better than the compressed version here).

SelfTest.jpg

For future reference, this is the video timing I worked out for the HP 2397A

Code:
Oscillator frequency:       35.19MHz
Pixel clock frequency:      23.46MHz (Oscillator frequency / 1.5) / 0.426uS

Horizontal Resolution:      640 pixels / 27.2805us
Horizontal Front Porch:     82 pixels / 3.495us
Horizontal Sync Pulse:      72 pixels / 3.069uS
Horizontal Back Porch:      126 pixels / 5.371us
Horizontal Total:           920 pixels / 39.216us / 25.5KHz

Vertical Resolution:        400 lines / 15.686mS
Vertical Front Porch:       3 lines / mS
Vertical Sync Pulse:        3 lines / mS
Vertical Back Porch:        19 lines / mS
Vertical Total:             425 lines / 16.667ms / 60Hz
 
Last edited:
Thanks, gslick. I had some of those pdfs, but the hp museum isn't as comprehensive as I'd thought, it seems. I don't know why I didn't go to Bitsavers fist. I can only guess that it was because I was on site, under time pressure and the HP museum came up first in search.

The first I/O card needs to go in the rightmost slot next to the 12201 Sequencer Card.

Does that have to be an HP IB card, as at present or could it (temporarily) be a 12040 async mux board

Good thing you have the 28658-63005 RS-232C MUX Panel Cable and the 28658-60005 RS-232C Panel for the 12040 async mux board, and also a 12076-63002 Ethernet Stub cable for the 12076 Ethernet board. The cables are usually more difficult to find and can be more expensive than the boards.

Do you have two 12040 async mux board, cable, and panel sets? If you only have one then you either need another one of those for the console for the second system, or a 12005 async board.

I have two 12040 async mux boards, but only one each of the the 28658-63005 RS-232C MUX Panel Cable, the 28658-60005 RS-232C Panel and the 12076-63002 Ethernet Stub cable. But that's OK because, while I'll test both machines, I only plan on keeping one. Probably the one with the 12076 Ethernet board - perhaps I can get it to talk to my pdp11/34? It also came with Ethernet kit.

I don't know if anyone has bothered to archive images of the different versions of the firmware EPROMs. If no one has already done that it would be nice to do so. What firmware versions do you have on you 12040 and 12203 boards?

I'll report back on that when I have a chance to pull the boards and look. Not sure what I have in the way of EPROM kit, right now but if all else fails I recently scored a Commodore Colt that might support the ISA card/ programmer combo that I got from another member last year. Worst case, I've been meaning to drag one of the 5150s out of the (dry, don't worry) basement and fire it up. They've been on the back burner for far too long - I may start another thread about the reasons why.

It will be some time before I can start any testing - as mentioned, I need to make space and that means completing at least two of three projects - a pdp11/34, a PET 4032 with 8050 FDD and an IBM 5110 with 5114 FDD and 5103 Printer. I've been acquiring stuff faster than I can play with it, this past year or so.

Speaking of which, does anybody know whether an HP 86B can access an HP 7945 HDD? I recently scored an 86B, already had two 9130 FDDs and while in my storage I noticed a previously forgotten 7945 HDD in a 19" rack adapter. It seems to me that that lot, along with a suitable monitor, would fit nicely in the upper part of the HP rack, if I keep it (which I think I'm talking myself into as I type this). The 1000 A900 and 7937 HDD will definitely fit in the bottom part. Who knows, perhaps I could use the 86B as the terminal for the 1000 A900? They made an 82939A Serial Interface and there's at least one terminal emulator for it - TERM/80.

When I stripped my pdp11/34 out of the double wide rack that I definitely didn't have room for, there were a couple of little 1U pull out shelves, presumably for placing manuals and the like on. I took them, because why not and if I cut a couple of slots in one for a strap, it would be perfect for the 86B.

Seems to me that that would be an excellent way to use all of the space in the rack. Two systems for the footprint of one, as I'd still be able to use the 86B as a standalone machine as well as a terminal for the 1000 A900. Damn, but that has me enthused! Time to pull my thumb out and make some progress!
 
Does that have to be an HP IB card, as at present or could it (temporarily) be a 12040 async mux board

On the earlier HP 1000 series the slot placement of I/O cards in the I/O backplane determined the I/O card select code, in addition to the card priority. On the A-series the I/O card select code is determined by switches on the card so the card priority is the only real consideration.

One of the manual says: "Interrupts from a higher priority device inhibit lower prority interrupts by breaking the priority chain. From the standpoint of system response time, it is more efficient to assign the higher priorities to high-speed peripheral devices."

For initial testing with just a VCP console connection it would be fine to install the 12040 async mux in the first I/O card slot next to the 12201 Sequencer Card. The important thing is to not have any empty I/O slots between the first I/O card in use and the last I/O card in use.

Speaking of which, does anybody know whether an HP 86B can access an HP 7945 HDD? I recently scored an 86B, already had two 9130 FDDs and while in my storage I noticed a previously forgotten 7945 HDD in a 19" rack adapter.

Do you actually have an 86A, not an 86B? The 9130A floppies attached to an 86A through an interface built into the 86A, which was not HPIB. The 86B had a built in HPIB interface instead of the 9130A floppy interface.

The 80-series Mass Storage ROM supports Amigo protocol storage devices. The Extended Mass Storage ROM is needed to support SS/80 protocol storage devices. CS/80 protocol storage devices are not supported. The 7945 is a CS/80 drive so it wouldn't be supported.

More info here:
vintagecomputers.site90.net/hp85/parts.htm#drives
 
On the earlier HP 1000 series the slot placement of I/O cards in the I/O backplane determined the I/O card select code, in addition to the card priority. On the A-series the I/O card select code is determined by switches on the card so the card priority is the only real consideration.

One of the manual says: "Interrupts from a higher priority device inhibit lower prority interrupts by breaking the priority chain. From the standpoint of system response time, it is more efficient to assign the higher priorities to high-speed peripheral devices."

For initial testing with just a VCP console connection it would be fine to install the 12040 async mux in the first I/O card slot next to the 12201 Sequencer Card. The important thing is to not have any empty I/O slots between the first I/O card in use and the last I/O card in use.

Thanks, that's helpful.

Do you actually have an 86A, not an 86B? The 9130A floppies attached to an 86A through an interface built into the 86A, which was not HPIB. The 86B had a built in HPIB interface instead of the 9130A floppy interface.

The 80-series Mass Storage ROM supports Amigo protocol storage devices. The Extended Mass Storage ROM is needed to support SS/80 protocol storage devices. CS/80 protocol storage devices are not supported. The 7945 is a CS/80 drive so it wouldn't be supported.

More info here:
vintagecomputers.site90.net/hp85/parts.htm#drives

This is why I shouldn't post late at night when the kit isn't in front of me. I have an 86B, with HPIB, two 9130As, with proprietary interface for an 86A and a 7945 which is, as you have kindly explained CS/80. None of these things are compatible with each other. So much for that idea.
 
I have the opportunity to obtain an A900, it has no peripherals (disk, tape) or any cables but has many cards installed, not sure what yet.. Has any work been done to use standard non HP 1000 hard disk drives with these?
It is a 115V version, is it possible by changing jumpers/wiring etc for it to work on 240V or will I need a step down transformer?
Thanks.
 
Last edited:
I have the opportunity to obtain an A900, it has no peripherals (disk, tape) or any cables but has many cards installed, not sure what yet.. Has any work been done to use standard non HP 1000 hard disk drives with these?
It is a 115V version, is it possible by changing jumpers/wiring etc for it to work on 240V or will I need a step down transformer?
Thanks.

Take a look at the Chapter 11, Power Supply section of this manual:
www.bitsavers.org/pdf/hp/1000/A-series/02139-90003_A900_ERD_Apr86.pdf

It appears that both the 440-watt supply for the 20-slot backplane and the 300-watt supply for the 16-slot backplane can operate from either 115 VAC or 230 VAC.

However, the fans run on a fixed AC voltage and the power cable to the fans needs to be connected to either the 115V or the 230V connector on the supply which corresponds to the line supply voltage.

If the A900 system you might acquire has any I/O cards installed I would be surprised if it doesn't have a least one 12009A HPIB I/O card installed. That is the standard I/O card for HPIB hard drives and tape drives. If you don't have any HPIB hard drives you can use a PC to emulate them. See the HPDrive Project: www.hp9845.net/9845/projects/hpdrive/

There is also the 12016A SCSI interface. Those are rare and $$$. I have two of those, but only one cable. Even the cable is hard to find and expensive. I have never gotten around to trying out the 12016A SCSI interface yet.
 
Or, if you have your heart set on a period correct HDD, the 7945s come up from time to time and are a lot smaller (both in capacity and physically) than the 7937s.

I would encourage you to buy the machine, then see what boards are in it.
 
s-l1600.jpgs-l1600a.jpgs-l1600b.jpgThanks, both you posts are encouraging and informative. I'll update the post if I acquire it.. I've attached some jpg's of spare cards that are with the system. There are 7 cards installed in the A900 from right to left. I haven't physically seen the machine to get part numbers.
 
Last edited:
In those photos the cards with four 40-pin DIP chips look like 12007B HDLC Modem Interface cards. Not really useful unless you connect multiple HP-1000 systems together in DS/1000-IV networks.

12044A HDLC Interface / 12007B Modem Interface (two similar but different cards)
www.hpmuseum.net/display_item.php?hw=1006
www.hpmuseum.net/exhibit.php?hwimg=1006

The card with a single 48-pin DIP chip is a 12009A HPIB Interface.
www.hpmuseum.net/display_item.php?hw=662

The memory card might be either a 12220A 768KB or a 12221A 3MB Memory Array card. Same cards with either 64K x 1 or 256K x 1 DRAM chips. The memory cards need to be installed next to the 12204A Memory Controller card and all connected to it via a 12222 frontplane board. The frontplane attaches to the black connector running between the two card eject handles. The frontplane boards can be more difficult to find than the 12204A Memory Controller card or the Memory Array cards. If you pick up the system make sure you get the memory array frontplane. There are various sizes of frontplanes depending on how many memory array cards are installed.

In addition to a 12204A Memory Controller card and at least one Memory Array card and frontplane, an A900 also needs a 12201A Sequencer card, a 12202A Data Path card, and a 12203A Cache Control card. A 12205A Control Store card is optional. Any additional cards would be various I/O cards.
 
Last edited:
Beyond the core A900 CPU and memory cards, you also need either a 12005A serial interface card and cable, or a 12040 8-channel async mux card and breakout box to use as the system console.

12005 Serial Interface
www.hpmuseum.net/display_item.php?hw=705

12040D MUX Interface
www.hpmuseum.net/display_item.php?hw=664

Along with the 12009A HPIB interface card, the 12005A and 12040D (or earlier revisions) cards seem to be the most common I/O cards.

A900 core CPU cards 12201, 12202, 12203, 12204
www.hpmuseum.net/exhibit.php?hwimg=594

A900 768KB card 12220A
www.hpmuseum.net/display_item.php?hw=706

A900 3MB card 12221A
www.hpmuseum.net/display_item.php?hw=729

A900 8MB card 12221B
www.hpmuseum.net/display_item.php?hw=1159
 
Back
Top