PDP8a CPU fixed! (was Re: DEC DELQA - seems not to work. Anyone got a spare?)

[Brian Walenz]

On Mon, Nov 7, 2016 at 11:44 AM, Ian S. King <isking at uw.edu> wrote:

> I love reading stories of component-level repair.
>

Assuming my notes and sketchy memory make any sense:

One of my PDP 8a CPU boards (the M8315) passed all the self tests I could
toggle in, EXCEPT, it would ignore HLT instructions.  It'd just blow right
through them.  That'll get you questioning your sanity real quick.

Several hours of squinting at schematics said I'm looking for the STOPL
signal - I found three or four places, one of which involved a ROM (at
least my notes say there is a ROM involved).  Uh oh.  Page 3-17 of the
microprocessor user manual lists when STOPL is asserted.  Basically, front
panel or HLT.  Page 4-39 has the logic for the front panel, and that was
enough to narrow it down to one instance of STOPL in the schematics, in the
middle of page H-9.  E39 (an 8881 aka 7439) or E33 (a 7402).  I swapped
those two out, repaired the trace that I busted, and viola!  HLT now
works.  I vaguely remember it was the 8881 at fault.

BTW, I'm open to suggestions as to how to even begin debugging an HP 1000 E
Series.  There seems to be a case holding a power supply in the way of any
access to the motherboard.  The machine fails to exit the 'counting' self
test right after power up.  I plan on writing up a better description once
it gets cold and snowy out, so don't feel bad if you don't see this plea
for help.

Cheers!

b