• Please review our updated Terms and Rules here

Restoring a DEC PDP-11/05

You could shorten the bus if you like by putting the M930 in the last DD11-B into the the CPU backplane where the M920 jumper goes.

But in any case there should be some activity when toggling switches. This usually means that there is something broken. The frontpanel is controled by the microcode, so if the microcode isn’t running there will be no front panel operation.
You write ”No activity”. Does this mean all LEDs are black?

I would start checking the data multiplexing circuit in the front panel itself. There is a clock that should tick and scan the front panel. Everything is well described in the manual. The schematics will be of great help.
 
There are also two DEC M930C terminator modules that were included. One of which was seated in the leftmost slot (if you are looking towards the rear of the unit) Is there a numbering scheme that applies to the slots for Unibus?
Slots are numbered 1-up starting with the one at the front. In some PDP-11 models there is labeling along one or both card-extraction rails indicating the slot number and the module number that should normally be present there. In your case you have the modules identified rather than the slot numbers. Those first nine slots are specially wired with dedicated module positions.

The module "fingers", and thus sections of an individual slot, are labeled F-E-D-C-B-A left-to-right looking at the set of slots from the front end. B-A are then the rightmost two "fingers" and contain the Unibus. So your terminator and jumpers always go into A-B (as you've been doing).

The individual 2-wide "blocks" of the backplane are known as System Units (SU). So, for example, you have two 4-slot DD11 SUs installed, each jumpered to the preceding SU. And in front you have a 9-slot specialty-SU.

SPC = Small Peripheral Controller; that would be a four-finger/tab module that goes into F-E-D-C in the DD11 and sometimes in particular slots in specialty SUs. These are "quad height". The front-panel controller is like this (quad-height) although it goes into a specially-wired slot.

Some modules, e.g., memory, use all six six tabs/fingers and are thus "hex height". As are the CPU modules.

A brief introduction to DEC-isms. More to come, I'm sure!
 
Are two terminators necessary/supported in the 11/05? I see two in the above photos. Perhaps as also previously suggested you cant try terminating at the end of the first backplane, that will eliminate the issue being in the other two backplanes.
 
You could shorten the bus if you like by putting the M930 in the last DD11-B into the the CPU backplane where the M920 jumper goes.

But in any case there should be some activity when toggling switches. This usually means that there is something broken. The frontpanel is controled by the microcode, so if the microcode isn’t running there will be no front panel operation.
You write ”No activity”. Does this mean all LEDs are black?

I would start checking the data multiplexing circuit in the front panel itself. There is a clock that should tick and scan the front panel. Everything is well described in the manual. The schematics will be of great help.
I'll give that a try. All of the LEDs are black (off) when toggling the switches. Ironically, when I had all of the modules removed from the system for PSU testing,
all of the console(?) LEDs were on, and toggling the switches wouldn't turn them off. For verifying the clock signal, should I start at the clock input on the display buffer/driver? I looked at the block diagram for the console and noted that the data input and function switches are 'independent' of the indicator LED circuitry.
 
Last edited:
The all LEDs on when boards out is normal. All black means that either the BLEG bus is all 0 or the 74150 on the datapath board is bad. Unless there is a clocking issue or counter issue on the front panel board. The front panel is rather simple and could be removed and debugged as the first step to rule out stupid things in there. Check the schematics on bitsavers.

When you checked the front panel for proper operation next step is to check the clock logic on the control board. If there is clocks check if there is any sign of life on the microaddress bus.
 
There should be a clock signal present on the front panel board making a couple of counters counting. The counters then drive which bit from the BLEG is selected using a 74150. Yes. The switches are completely independent of the LEDs.
 
The all LEDs on when boards out is normal. All black means that either the BLEG bus is all 0 or the 74150 on the datapath board is bad. Unless there is a clocking issue or counter issue on the front panel board. The front panel is rather simple and could be removed and debugged as the first step to rule out stupid things in there. Check the schematics on bitsavers.

When you checked the front panel for proper operation next step is to check the clock logic on the control board. If there is clocks check if there is any sign of life on the microaddress bus.
That sounds reasonable. Just to clarify, is the clock signal on the front panel the one I should be checking? I don't yet think I have a way of testing the 7260 'in situ'
without the module riser card.
I had to order two new probes for my oscilloscope, as the single one I had has seen better days. Until those arrive, I'm left with only a digital multimeter, which I've been using to probe around on the front panel to verify that the data input switches are receiving power (which they are). It has a built-in frequency counter which I haven't really tried to use for checking things like clock speeds, so I'm not sure if that will help.
I've been printing off all sorts of schematics to familiarise myself with. Pretty thankful that this machine is built using mostly discrete chips.
 
So I checked the clock output on the front panel (the 7413 IC shown in figure 7-2 of the manual). The graph isn't the best looking because I didn't have a probe to use, but it looks like the square wave the manual shows. Ch1 is on pin 5 (clock stage 2 output) and Ch2 is on pin 13 (stage 1 output). The clock is also present on pin 6 of the 8271 display buffer/drivers.



PDP11 74123 ch1_2.PNG
 
I thought some more about the 11 last night. The guy I bought it off of sent me a pic of it to verify it powered up. In that picture, the ‘HALT’ switch is down, all data entry switches are down, the ‘RUN’ light is lit, and the ‘14’ data/register LED is lit. I have not been able to reproduce that configuration. Is it possible that something came loose when I dismantled the unit for cleaning?
 
When the HALT is down, there should not be the RUN light. This is not OK. Look at the HALT key. Up is ENABLED , down HALT.
You will have to find out what's wrong.
Also when the machine is in RUNning mode, the DATA LEDs have no meaning. The frontpanel has very limited use while running. This is a little bit annoying with the 11/05.
 
When the HALT is down, there should not be the RUN light. This is not OK. Look at the HALT key. Up is ENABLED , down HALT.
You will have to find out what's wrong.
Also when the machine is in RUNning mode, the DATA LEDs have no meaning. The frontpanel has very limited use while running. This is a little bit annoying with the 11/05.

What part of the machine would most likely be causing that behavior? Here is the picture for reference:
 

Attachments

  • 0D9446FF-534E-4BB2-8DB2-1C905970EB6E.jpeg
    0D9446FF-534E-4BB2-8DB2-1C905970EB6E.jpeg
    71.8 KB · Views: 22
nDZRNUEl.jpg


What is interesting here is if there is a proper clock into E4 from pin CC on the connector. Then if that is OK that that 74193 counters are counting properly.

4VzGX1Pl.jpg

How does the signal DAK look like coming from pin PP?


RUN lit while HALT indicate major problem in the control section. Is the CPU clock running at all? But I still think it is a good idea to rule out any front panel issues since you will need to get a raiser card to do measurements on the CPU boards. There are a few signals that can be observed in the backplane. You’ll see which in the schematics. If I remember correctly the microcode address and clock are there.
 
nDZRNUEl.jpg


What is interesting here is if there is a proper clock into E4 from pin CC on the connector. Then if that is OK that that 74193 counters are counting properly.

4VzGX1Pl.jpg

How does the signal DAK look like coming from pin PP?


RUN lit while HALT indicate major problem in the control section. Is the CPU clock running at all? But I still think it is a good idea to rule out any front panel issues since you will need to get a raiser card to do measurements on the CPU boards. There are a few signals that can be observed in the backplane. You’ll see which in the schematics. If I remember correctly the microcode address and clock are there.
Here is what I have from pin PP on the front panel (with the ribbon cable removed):
 

Attachments

  • PDP11 Console pin P_P.PNG
    PDP11 Console pin P_P.PNG
    49.8 KB · Views: 12
Hasn't that particular listing been there since June 7th?

Looks like you gave up as the 11/05 is now on sale on Ebay. :)

Nope not mine! (well the one in the picture is...) The guy I bought it off of had two 11/05's and a PDP-8/I, he only took a picture of one of the 11/05s though, the one in the was the one I was sold (and nowhere near that price ;))... When I met up, he said that a few people outside eBay had contacted him as well, including some guy from Chicago willing to ship the thing. I actually found the 11/05 by accident when I saw he had some paper tape programs for various PDPs listed on FB marketplace. I was told the 8/I sold for $6000 IIRC.
 
You have to check all signals with the ribbon cable installed.
Here is what I have for pin PP with the ribbon cables installed between the front panel:

Pin P_P new.PNG

I was also thinking about the cables this weekend, it occurred to me that the 40-pin Berg plug does not appear to be 'polarized' and can be inserted both ways. Drawing BCO08R, which shows the pin orientation references markings on the shield side of the cable but does not specify if the shield should be on the side facing out the front or back. I didn't want to try reversing the cable for fear of messing up something, but it's just a little thought that popped into my head. (I'm notorious for spending a lot of time troubleshooting only to find the issue was something very simple)
 
Back
Top