🚀 go-pugleaf

mcgee@nic.csu.net writes:

>Last night was the primary application of power (Read: Smoke Test)
>for my PDP 11/44. Everything spun up, the room lights stayed lit,
>and, so far, no one complained about the noise the fans made. 
>However, I have a problem. The diagnostic rom says "?22 CP HUNG"
>and a T/E trace says "CONSOLE -A7".

Having just fixed one which had a similar symptom, and not knowing
your exact machine configuration, try the following.

1. Remove the bus jumper from the CPU backplane to the rest of the
machine.  Replace it with a terminator.

2. Power the machine up.  If you receive a console emulator prompt
the CPU is ok.  If not, you now need to start looking at things
on the CPU backplane. (see below)

3. If step 2 passed, reinsert the unibus jumper between the CPU 
and the rest of the system.  If you have more than one expansion
backplane, remove the jumper to it, replacing it with a terminator
as above.

4. Continue steps 2, 3 until the CPU again gives a ?22 CP HUNG
message.  At this point you have located the backplane where the
problem lies.

5. Now that you have isolated the defective backplane, you are down
to just a few potentially failed boards.  It is now trial and error
to locate which one is hanging the bus.  Do this by removing each
board individually, placing a bus grant card in position D of the
card slot, and cycling the power again.  You have to be careful,
though.  If the board uses NPR signals, you have to also either
jumper pins CA1, CB1 or use a G7227 grant board instead of a G727.
(I would recommend this one in any case just to be sure.)

You need to be careful of such things as RH11 controllers and so
on, you can't pull the boards in them like that, you have to be a
bit more careful of what you pull and replace with one of them or
other things such as DV11s...

In any case, the above procedure should at some time result in the
CPU finishing its power up test and giving you a normal console
prompt.  When it does, you have located the defective card.  Now
prepare yourself for the agony of the replacement cost.  Once you
have found the potentially failed board, carefully inspect the back-
plane on the wiring side to ensure that if the board is not an NPR
board, there is a jumper installed on CA1 -> CB1.  If not, put one
there, replace the board and try it again.  If all of the above steps
result in you standing there holding a board which is indicated to
be bad, replace it with a new or repaired one.

If the above procedure fails, send me an e-mail with some more specific
information about your system and I'll try to give you a bit more
information.  I'm quoting the above from memory - I'm at home and all
my 11/44 references are at work!


>Does anyone have any guess as to where I should start looking?
>-Rich

Hope this helps.

Chuck