help needed: problem with UNIBUS access on my PDP-11/70

From: Gooijen H <GOOI_at_oce.nl>
Date: Mon Jul 15 03:19:00 2002

Thanks all for the many good replies.
This is what I have done so far.
At the end is what I did *not* do (yet).

At first, I kept the 3 MASSBUS interfaces installed. They were in
the machine, so I did not suspect them.
- I installed the H744 in the upper PSU.
  All +5, -15 and +15 V. all look OK, measured with a Fluke. I am not
  yet in the stage to pull out the scope and do "difficult" things.
- I will check the jumper on the M8132. Normally I make notes of steps
  I do on my machines, but Murphy is still alive. I cleaned the 11/70
  with remote console *completely*. Got the 11/70 with full console
  2 weeks later. Since I am keeping one and donate the other 11/70 to
  a museum, I did not want to clean the second 11/70 also. The guy at
  the museum may do his share of this bargain. So I swapped the consoles,
  and (AFAIK) the boards. That is the problem: I can not remember that
  last action. Perhaps I did not swap all CPU boards between the 2
  machines and only moved the FP option boards (and forgot/didn't know
  about the H744 to supply +5V to slots 2-3-4-5).
  As Tom wrote, the 11/40 indeed needs on 2 boards (IIRC) change of
  jumpers for FIS/EIS (or was it MMU)?
- I kept a G727 in the 4th MASSBUS slot at the correct position.
  Removed from all not used UNIBUS slots the G727 and replaced them
  with G7273 dual-height grant cards.
  Result is still NOK: hung on console address access (read/write).
- Removed the other 3 MASSBUS interfaces. Put a G727 instead of the
  M8153, according to FMPS. Hung remains.
- The "golden" tip from Tony. Removed the M9302.
  No more hung! XBUF status shows octal 200 (indicating TX buffer
  empty, ready to receive byte to transmit). Store octal 71 in the
  TX buffer and a "9" appears on the VT320!!
  Basically, I have a complete working machine, but ...
- Put an M930 as terminator in the 11/70. Still all is fine.
- Put an onther M9302 (from the working 11/34C) in the 11/70.
  The hung is back.
- Put the M930 back in again.
  LOADed address 17765744 (from the M9312 manual) and pressed START.
  All LEDs flicker for a few seconds, then all go off, except ADDRESS
  LED #2, so address 00000004. I guess this is a TRAP because the
  boot device was not accessible. Makes sense, there are no devices
  connected!

What I did not do.
- Remove the RX11 (non-DMA) RX01 interface and replace it with an
  RX211 (DMA) RX02 interface to see how that affects the NPR line.
- Did not check the ACLO and DCLO signals.
  Since I can run "a small program", memory interconnections, the
  memory and the CPU itself have no apparent problem.
- Did not try to read the ROM on the M9312 with the M9302 installed.
  Will do that this evening (probably). The weekend was shorter than
  I expected it to be. As usual ...
  I did read the CPU registers (IIRC), but check on that again.
- I will read the M9312 manual again. The numbers Johnny writes about
  sound familiar. However, it this moment I do not have any disks
  (RX,RL) connected.

Thanks again so far,

- Henk.
Received on Mon Jul 15 2002 - 03:19:00 BST