• Please review our updated Terms and Rules here

Random find: computer system with pdp 11/23 cpu

SEgamer

Experienced Member
Joined
Sep 18, 2011
Messages
97
Location
GA, USA
Seeing if anyone has any more info this system I recently got (mainly how to operate it). It's a Scientific Micro Systems MDX01170. I only found one post on a site where someone else was asking about it, but not much info on it other than the boards in it.

Pics of the front and back:
100_1289.jpg 100_1288.jpg

The only history I have on this one is that it came from a local air force base (they used to auction off a lot of cool items but now send items to another base to sell). The tags/labels on it had "Programming Station" on it, and after looking up the federal stock number, this is the description they have for it:

C/O PDP 11/23 CPU;256 KB RAM;8 IN. FLOPPY DISK DRIVE;VT-101 TERMINAL;160 CPS SERIAL PRINTER;SCAN SOFTWARE

I'd like to find a VT101 or VT100 terminal to use with it. There's probably one somewhere around here that was auctioned off, but it'll be tough finding a person who bought one.


Here's a list of the boards inside:

TZ0 - TZ3 are connected to M3106 which is a 4-line double-buffered async MUX
LP0, TT0-TT2 are connected to a Sigma Information Systems 400200 board, which I think handles serial communications
CPU board is a M8186-YC = 11/23 CPU without options, KDF11-AC DUAL HEIGHT CPU
Clearpoint QRAM-11B board with 256 kB memory.
Scientific Micro Systems 0004222 and 0003772 Disk Controller cards

It also has a maxtor xt-1085 HDD too. Highly unlikely it still works, but I'll find out when I can find a terminal to use with it.


I do have a couple of questions:
Can I use any terminal with it instead of a VT101 or VT100?
Is there something special with the LP0 connection? I'm wondering why it goes to the same board as the TT0-TT2 ports but is labeled different.
What sites have good information on this type of stuff? I guess I should be looking for info on the PDP-11 architecture?
 
Last edited:
I do have a couple of questions:
Can I use any terminal with it instead of a VT101 or VT100?
Is there something special with the LP0 connection? I'm wondering why it goes to the same board as the TT0-TT2 ports but is labeled different.
What sites have good information on this type of stuff? I guess I should be looking for info on the PDP-11 architecture?

What a nice system.

You would be best off using a PC with a terminal program that properly emulates a VT100 in case the functions keys are useful.

From the description you have received, the 25 pin LPO seems to be a serial printer output port.

It would be nice if the disk is still good. The system might boot up into UNIX. Hopefully it will not need a system password, otherwise you may need some special software to load using the 8" floppy drive.
 
Tried it today by hooking it to my PC through the serial port. I made sure the pinout on the db9 to db25 adapter was correct. Heard the hard drive spin and make two whines as it turned on, but the drive's light never turned on. I hooked the drive up to a 286 computer and the light came on as the computer tried to read it (got an initialization error). FDISK also gave the "error trying to read disk". I have no clue what might be causing it other than a bad drive, but if UNIX is on it I guess that could be causing the problem?

The only thing it returns in hyperterm is a space right when it turns on.

Another issue I see is with the RAM board. Its LED turned on probably 6 out of the 10+ times I turned it on. From what I read it means there's a parity error. Reseating the chips might fix it, but there's probably one or two chips on it giving the error.

Here's some photos of the boards incase anyone has a system with them:

RAM board
100_1285.jpg

I/O card
100_1284.jpg

Main disk controller
100_1286.jpg
 
Last edited:
  • Sounds like the disk is good. I wouldn't endanger it any more in a PC if you want to try to recover software from it.


  • I don't recognize that particular one, but it's a QUAD SLU, probably compatible with a DLVJ1 or DLV11-J. If this is so, then the output labled LP: is likely serial.


  • Good VT100 terminal or Terminal Emulator should be fine to replace a VT101.


  • The disk control I do not recognize at all. It appears to be more than that however.

Looks like a nice clutch of boards there. I should think it would be possible to get it running. I hope you took pictures before pulling any boards, to record how it was cabled and where they were in the backplane.

Nice find really.

The CPU will usually issue hints about it's attempts to boot on the console serial port. Did you have any terminal on it [TT0:] when you tried it?
 
If there were any docs to be found online I would expect them to be found here:

http://www.bitsavers.org/pdf/sms/qbus/

I don't see any docs there which look related to this system and controller card.

I have an SMS 1000 ( 3001076A_SMS1000oem_Feb87.pdf ) which has an integrated controller card with utilities built into the firmware for configuring and formatting the hard drive and floppy drive. Those utilities run locally on the 80188 on the controller card. If your controller card does not have its own serial port I assume the PDP-11 host needs to be functional to be able to access the firmware on the controller card. If you had access to an EPROM programmer it might be interesting to dump the firmware EPROMs on the controller card and see if anything interesting can be found there.

Did you try connecting your terminal emulator to all 4 ports attached to the Sigma card, and various baud rates including 9600? I would expect the system to be configured for one of the Sigma card ports as the console. Did you try sending a break from the terminal emulator, or try without the drive controller card installed? It might be configured to auto boot the firmware on the drive controller card. If you don't have the drive controller card installed or you send a break I would expect it to drop into the ODT prompt.

-Glen
 
Thanks for the replies! I'll definitely need all the help I can get since I know nothing about Q-bus systems.

I downloaded the latest hyperterminal program for my computer (Windows 7) and tried 300 and 1200 baud. The other settings were: 8 data bits, 1 stop bit, no flow control and no parity. There were other settings where I set the protocol as VT100 and also tried VT100J. I tried all of the serial ports on it and all I got was a British pound sign the first few times, then started getting the space character after I reseated the cards. All I did was randomly press keys and Return to see if a carriage return would do anything.

When it starts up, the Power and Run lights on the front are lit. One LED on the disk controller and another LED on the backpane light up at the same time for a few seconds, and then turn off. After this is when the parity LED on the RAM board will turn on if there's an error. If there is an error, then the Run light turns off.

I pulled the boards out one at a time to take the pics and write down the numbers so that I wouldn't get them mixed up. Now whether or not they were in the right spots when I got it, I'm not sure. The guy already had it opened to look up the boards and drives. I'm guessing that the boards have to be in certain slots for it to work correctly.

I'll try booting it without the controller cards in it and see if anything different happens.
 
Thanks for the replies! I'll definitely need all the help I can get since I know nothing about Q-bus systems.

I downloaded the latest hyperterminal program for my computer (Windows 7) and tried 300 and 1200 baud. The other settings were: 8 data bits, 1 stop bit, no flow control and no parity. There were other settings where I set the protocol as VT100 and also tried VT100J. I tried all of the serial ports on it and all I got was a British pound sign the first few times, then started getting the space character after I reseated the cards. All I did was randomly press keys and Return to see if a carriage return would do anything.

If you haven't tried 9600 baud I would try that. That is a more common console setting. Also, try sending a Break and see if that get any different response.

-Glen
 
I took the controller boards out and put the cards in various slots just to see if that would fix anything. I tried sending a break character (Ctrl + C), but that didn't do anything.

I managed to get it to send a prompt (indicated by @) and it echoed back characters with a question mark if it didn't recognize them. Seems that 'g' and 'p' return something, and 'r' is some sort of command. Or this was all coincidence because the RAM LED was always on and the Run light never came on.

term.jpg
 
I managed to get it to send a prompt (indicated by @) and it echoed back characters with a question mark if it didn't recognize them. Seems that 'g' and 'p' return something, and 'r' is some sort of command.

That is some progress. The '@' character is the ODT prompt. Make a note of which serial port on the system you had the terminal connected to and what the serial port settings were so you can get back to the ODT prompt on the console port again.

If you haven't done so already you really need to download and take a look at the KDF11-AA Users's Guide:

http://www.bitsavers.org/pdf/dec/pdp11/1123/EK-KDF11-UG-PR2_Mar79.pdf

Chapter 3 describes the ODT command interface.

You should also take a look at Chapter 2 Table 2-1 and Figure 2-1 and make a note of how the jumpers W5 - W15 are configured on your M8186. If W5 is Removed and W6 is Installed it is configured for Power-Up Mode 2, and then if jumper W8 is installed it is configured to power up bootstrap at address 173000.

If you get to the ODT prompt with the disk controller card installed and at the '@' prompt you type "17773000/" what does it respond with?

If you type "17773000g" what happens?
 
I took the controller boards out and put the cards in various slots just to see if that would fix anything...

This is alarming.

Please do not re-arrange cards from their original positions. This is a PDP-11, not a PC and not only can this make it "not work", but it can cause FLAMES and permanent unhappiness.

Do you know the original card positions? Maybe post a pic. We need to determine what type of backplane the box has.
 
Well, he could have noted where they were and put them back in the original configuration again. Or with a photo might be able to if he hasn't yet.
 
There are three main types of LSI-11 bus backplanes:

(1) Serpentine LSI-11 bus.
(2) Straight LSI-11 / C-D bus.
(3) Mix of straight 22-bit LSI-11 / C-D bus and serpentine 22-bit LSI-11 bus.

Also, the LSI-11 bus may be either 18-bit or 22-bit.

My guess is that this system has a serpentine LSI-11 bus, in which case moving around the boards that came with the system should most likely not cause any harm, but may cause things to not work due to issues such as missing grant continuity.

The real bad things can happen if the backplane has C-D slots and you plug a dual width card into the C-D half of a C-D slot, or if you have a quad width card which uses the C-D bus and you plug it into a slot which is not a C-D slot.

With notes on how the boards were originally configured we could most likely correctly guess the backplane type. For example, if the M8186 CPU was originally in the upper left slot and the QRAM memory board was next to it in the upper right slot the assumption would be that it was a serpentine LSI-11 bus backplane with no C-D slots.


Here are some references on the DEC LSI-11 backplanes that are likely to be encountered in PDP-11 systems:

http://www.bitsavers.org/pdf/dec/qbus/EB-23144-18_QbusIntrfs_1983.pdf

H9270 - 4 x 4 backplane (four rows of four slots)
Serpentine LSI-11 bus.

H9273-A 9 x 4 backplane (nine rows of four slots)
Straight LSI-11 / C-D bus.

H9275-A 9 X 4 backplane (nine rows of four slots)
Serpentine 22-bit LSI-11 bus.

H9276 9 X 4 backplane (nine rows of four slots)
Straight 22-bit LSI-11 / C-D bus.

H9281-AA/BA - 4 X 2 backplane (four rows of two slots)
H9281-AB/BB - 8 X 2 backplane (eight rows of two slots)
H9281-AC/BC - 12 X 2 backplane (eight rows of two slots)
Straight LSI-11 bus.

http://www.bitsavers.org/pdf/dec/qbus/EK-OLCP5-TM-002_microPDP11tm_Jan84.pdf

H9278-A 8 X 4 backplane (eight rows of four slots)
Three rows of straight 22-bit LSI-11 / C-D bus.
Five rows of serpentine 22-bit LSI-11 bus.

-Glen
 
Here's photos of the slot labels (what is printed on the backplane) with the layout of the cards when I got it and when the prompt came up. Not sure if the secondary card hooked to the disk controller is considered a disk interface (hence the question mark). I didn't put any of the cards in the first row (J0) and mostly moved the RAM around. It does the same thing that it did before I moved the cards when I put the cards back into their original locations.

Edit: just read your last post, gslick, and from what the labels are it is 1 row with ABCD, then serpentine.

QBUS3.jpg
 
Last edited:
Those pictures help. But they are upside down from normal convention and would be less confusing if you flipped them 180 degress.

It's interesting that the disk controller cards were in the first slots above (if the picture was flipped) the CPU card. I wouldn't have expected that.

Normally the CPU is the first card in the blackplane (one exception is an M8190 KDJ11-B CPU with M8637 MSV11-J PMI memory located above it in the backplane).
 
Fixed the photo. After starting it up a few more times, I realized that it went to the prompt yesterday because it would randomly hit a parity error because it was set for bootstrap mode. I put the jumper on W5 for ODT mode and now the prompt comes up every time. I'm able to see and set data in the addresses and registers by using the r and / commands, so it appears to be working as far as CPU, RAM, and serial card goes.

I think the disk controller is faulty though. The pins on some of the socketed ICs are oxidized, so I'll re-seat them and see if that works. If not, then I'll check for cold solder joints on the backplane.
 
Boy this is a "Wild west Thread"... I hope we can help you before any damage is done.

I'm goiing to just throw out questions and thoughts...


  • Glen, this doesn't look like a DEC manufactured backplane to me, especially since I see the nomenclature on those diagrams he posted.
  • Incidentally - where did those come from? I suspect more photos will show an interior chassis label that will help us understand.
  • In general, it is very risky to install an ABCD [QUAD wide] card in any slot other than an ABCD slot.
  • Dual wide cards must be installed only in AB slots. This can be the first two [left most as labeled] of an ABCD slot.
  • Your "Diagram" would indicate to me that the first slot ("0") is indeed an ABCD. For your experiments, either leave this one empty or install only the QUAD disk controller in it. [that rule will help my nightmares - thank you.] Do not put the QUAD card anywhere else. Leave it empty for now.
  • Another rule of thumb is that the "Chain" of AB slots should be unbroken to the CPU. [no empties]
  • Based on your diagrams I would take the chain to be 1A-2B-3A-4B and so on. If correct, makes it NOT a DEC "S" pattern. Still, it only makes a real difference for "Tail End Charlie".
  • Correct, you got an ODT prompt "@". [that means the speed settings were correct] It appears to be responding in 16-bit mode. Please pull your CPU and photo the card for us. In particular, try to make the printing readable and we'll look it up. Some were 22-bit capable, some not.

If you haven't smelled any magic blue smoke yet, you've probably avoided a disaster. Just go slow from here and we'll try to keep you safe. You have the makings of a nice system there. I don't see your bootstrap there. Glen, is it customarily on those Disk Controls? [like in the DSD controls?]
 
Last edited:
I didn't mean to imply that I thought this was a DEC manufactured backplane. I was just listing some of the various DEC manufactured backplanes configurations and where to find some documentation on those as a reference to find out what some the various backplane options you might find. Or course it would easier to refer to the proper 3rd party documentation for this system, but that doens't appear to be available.

3rd party manufactured backplanes will likely be similar to one of the DEC manufactured backplanes, but they could also do something strange. I would say this SMS manufactured backplane is probably slightly unusual if the disk controller is normally located in the first slot and a half.

I don’t see cables going to the drives from the controller cards. Were drive cables removed from the controller cards to make the pictures more clear? Or are the drive cables attached somewhere else on the backplane? If the drive cables attach to the backplane then almost certainly the drive controller needs to be in a specific slot so that the non-standard signal routing works between the drive controller and the drive cable connectors on the backplane. (Some DEC backplanes do some routing like that for the DSSI bus from the KA640 or KA660 VAX CPU in the backplane to the DSSI drives, I forget the exact details on that).

A lot of 3rd manufactured disk controllers have a boot PDP-11 ROMs. I would assume that the SMS controller here has boot ROMs configured at the standard 173000 address if that is how the M8186 CPU was configured to boot. If he can get to the ODT prompt again with the disk controller installed then if typing "17773000/" returns data then that is probably the disk controller boot ROMs.

Also, if "17772150/" returns data then the disk controller is an MSCP controller at the default address.

-Glen
 
I'll try to get to the backplane tomorrow and will post the manufacturer and fabrication/model number if there is one. You're right about the order of the slots, RSXM11+. I don't know why I said serpentine in that one post. The cables were removed to show the cards better.

Update: Tested 17773000/ and it returned "005000" , and also tested 17772150/ and it returned "?"

I guess I had dumb luck by deciding not put a dual wide card in the first row, so none of the cards should be damaged.

Here's a pic of the CPU board (KDF11-A right?):

100_1292.jpg
 
Last edited:
Looking again at your photos, especially the one entitled "Original Layout" I believe a card was removed from your system before you got it. [from J4B]

Please include a photo of the card originally in J2B so we can identify it.
 
Back
Top