9 May
2006
9 May
'06
12:22 a.m.
Tony wrote...
It might well be a backplane problem -- a signal getting lost because of
a bad wire or something. But the easiest way to find that is to work
through the circuitry that handles the vector, both the 'path' the vector
takes and the control lines that control (!) that.
I think the problem is on one of the processor
cards. To find it is going
to involve diging into the logic
Oh Joy. Supposedly this set of processor cards
booted up Unix just before it
was shipped to me, but I guess you never know. I'd be more inclined to
suspect that someone/thing before me messed up something on the backplane. -- do you have the necessary test gear
(logic probe, 'scope, etc)?
Very likely. HP logic probe, pulser, current
tracer, Tek 2246 scope, HP 1631
logic analyzer, HP freq. counter, HP bench DVM... probably most things you'd
call out :) Do you feel happy getting amongst a machine
with over 1000 ICs in it (yes, I know the feeling. That's how I felt when
I got my 11/45. I spent weeks just reading the printset).
Happy? Absolutely! Feel
like I have a clue what I'm doing? No! But I'm willing to learn. I have no extender cards for the
machine though, so
hopefully much can be done just from the backplane side.
Hmm... We may have to solder short bits of wire to the back of some of
the boards, put the board back into the backplane, then probe the ends of
the wires. That's how I get by without an extender on most of my machines.
OK, I'll look at the manuals to see what should be going on.
-tony