Restoration of a PDP11/60

[pbirkel]

@Henk on your webpage you mention:

In our 11/60 the big fans suck the air in from the top of the card cages and blow it out on the bottom, like a vacuum cleaner.
Maybe this difference makes your 11/60 more noisy than ours ...

Are you sure that you didn't reinstall your fans upside down :-}?

 

[klapperp]

Are you sure that you didn't reinstall your fans upside down :-}?

I double checked it

 

[klapperp]

We have started to populate our 11/60 again with it's "brain" !

With the help of a useful tool, we were able to check our power supply again:

G7273 Diagnostic

Joomla! - dynamische Portal-Engine und Content-Management-System

retrocmp.com

The backplanes were connected accordingly and the Unibus cable for the tape drive was also laid out:


The CPU cage was then initially equipped with all components, but the bus was terminated there:

 

[klapperp]

As already suspected, unfortunately the CPU did not start completely right away:


But there is at least a sign of life from the CPU:


By the way, we have turned off the battery for our tests, therefor the indicator LED on the console panel is off.

 

[klapperp]

We then reduced the configuration in the CPU Cage to a minimum (with only 64KB RAM):


Now it's time to debug

To be continued ...

 

[klapperp]

After the console panel originally worked at the previous owner ...


... we will first of all check the wire-wrap connections on the backplane which lead to the connector from the console panel:


To be continued ...

 

[klapperp]

The KD11-K CPU backplane shows no visible errors:


During this inspection, we also reinstalled the missing NPG jumpers:


During the debugging process we noticed the following...

After power on, the following kept appearing on the console panel:


It turned out that this was related to the M9301-YH bootstrap terminator.

We then first removed everything except the 6 CPU cards from the KD11-K backplane:


Then the Proc-LED on the console panel went on:


Now we focused on the M9301-YH module, and fortunately we had a 2nd one availabe.

As soon as the other bootstrap module was installed, our 11/60 CPU could start:


With this, a first step is completed because our CPU can start into a state where the console panel works:


On the original bootstrap module that came with the machine, apparently incorrect or defective PROMs are installed.

To be continued …

 

[klapperp]

Would you mind posting some more photos of this unit so that we can see the innards, power connections, and such?

Unfortunately, ours has been modified:

 

[Henk]

Indeed "modified". Between the fdan and the terminal strip should be an H7xx power brick.
I was this afternoon in my "museum", and worked a little on my 11/60. Will take a picture next time.

I need to correct my webpage!
The large fan in the bottom sucks air out of the module cage, and blows the airstream out at the bottom towards the front side. So, the air flow is from the top of the box towards the bottom of the box.
When I switch on the 11/60, the displays shows .0.3.1.7.0.1 and when pressing CTRL HALT/SI the display changes to 000000 (without the dots). When I key in 1 2 3 4 5 6, the display shows 123456. So some part of the CPU seems to work. I vaguely remember that the CPU works fine when single stepping through code but when run it crashed. To be inverstigated further ...

 

[pbirkel]

Indeed "modified". Between the fdan and the terminal strip should be an H7xx power brick.

Apparently the complete brick; I found this photo which I think is yours:



Note the rear 2x3 Molex. According to EK-FP11E-UG-001_FP11-E_Users_Guide_Dec77.pdf there should be a 9-pin Molex connector at the other end of the attachment cable and that only +5V is supplied.


Note that what is being described is an H7412A:


The diagrammed power distribution board is consistent with an earlier photo from @klapperp, before the SU harnesses were attached. I wonder what that board looks like now with all cables attached?



It seems that the "modifications" included bypassing the double-ended Molex cable as well as the power distribution board. Six differently colored leads to the Fast-on clips seems like a recipe to NOT get the right connections made to the SU?

Going back to @Henk's photo judging by the PCB solder-side etch that's an H7441 (32A) rather than either an H744 (25A) or H7440 (25A). That's a lot of juice for just four boards. Interesting!

 

[pbirkel]

When I switch on the 11/60, the displays shows .0.3.1.7.0.1 and when pressing CTRL HALT/SI the display changes to 000000 (without the dots). When I key in 1 2 3 4 5 6, the display shows 123456. So some part of the CPU seems to work. I vaguely remember that the CPU works fine when single stepping through code but when run it crashed. To be inverstigated further ...

Are your FP11-E modules installed? How are you powering them?

 

[klapperp]

I wonder what that board looks like now with all cables attached?

With all the cables attached it looks like this:

 

[pbirkel]

With all the cables attached it looks like this:

Thank you. It seems that the diagram in EK-FP11E-UG-001_FP11-E_Users_Guide_Dec77.pdf is incorrect since it shows the (auxiliary) FP11-E power supply connected to J4 rather than J5?

 

[klapperp]

shows the (auxiliary) FP11-E power supply connected to J4 rather than J5?

This is what it looked like before we took it apart for cleaning:


The connectors on S.U. #6 which are now new, belong to the formatter board inside the TS03 tape drive. We added an extra DD11-D backplane in the expansion card cage which uses the connectors from S.U. #5 and one connector from S.U. #6 and therefore we connected the power supply for this single external Unibus board here

 

[pbirkel]

This is what it looked like before we took it apart for cleaning:
View attachment 1322986

The connectors on S.U. #6 which are now new, belong to the formatter board inside the TS03 tape drive. We added an extra DD11-D backplane in the expansion card cage which uses the connectors from S.U. #5 and one connector from S.U. #6 and therefore we connected the power supply for this single external Unibus board here

Sticking with the original choice of socket makes complete sense, but the referenced document clearly states:

8. Locate socket J4 on the power distribution board inside the BA11-PA box.
9. Connect the +5 Vdc cable from the FP11-E power supply to socket J4 (Figure 8-4).

So it's not just "poor artwork". I find this whole approach to connecting the FP11-E power cable a bit disconcerting as the cable harness FROM the various power supplies connects to the backside of the distribution board and then the plugs to the various SU connect to the front -- which is the standard practice on these distribution boards. So there must be some special-case wiring between J4 and J5 on the backside that allows J4/J5 to be swapped (power in vs. power out) without ill effect _OR_ the FP11-E User Guide is very wrong. I'm left assuming that the cable that you have in J4 uniquely/exclusively supplies the four FP11-E slots (8-11), so J4/J5 is just a back-to-back pass-through from one male Molex plug to the other.

I did find the following distribution board photo that agrees with your wiring and shows the FP11-E power supply cable connecting to J5:

 

[klapperp]

so J4/J5 is just a back-to-back pass-through from one male Molex plug to the other.

I will check this next time when I remove the cover from the card cage!

Since we are now making progress with our 11/60 project, we have started thinking about spare parts for the future.

 

[Henk]

Yes, that power supply with the blue-ish background is the one for the Floating Point option, in the back of my '"Frankenstein" 11/60 (in an H960 rack).
I took a picture of the FP power supply in the "official" 11/60 housing, but I did not have time to get it out to take pictures of the inside. Next time ...

 

[Henk]

Interesting, the (original DEC) PSU for the FP option picture with the blue-ish background of the "11/60 in H960" has a fan mounted in front of the heat sink of the power brick.
My other (completely original 11/60) FP option PSU, has just like yours, two fans. One fan near the heat sink and fan in the rear corner of the box.
The power brick in the box is an H7441.

 

[pbirkel]

The power brick in the box is an H7441.

Thank you for the confirmation. 32A seems extraordinary for (just) those four modules.

 

[Terri Kennedy]

Thank you for the confirmation. 32A seems extraordinary for (just) those four modules.

Allow me to introduce you to the VAX 8600 power system:
570A of 5/5.2V
200A of -2V