I gave up (for the moment) trying to find the problem with the
memory on my 8/L. It never returns anything but all-0's to the MB,
but all the control pulses and timing are correct. To make matters
worse, while scoping the output of the bit 5 sense amp I
discovered it would change every time when I changed bit 0 of the
switch register and hit DEP!
Meanwhile I decided to go after another problem (when you hit
START at 0000 it runs up to 0100 and "hangs" with the BREAK and
RUN lights on.) If it's executing 0000's (AND instructions) it
should just keep running. The only way to wake it back up is to
flip the SING STEP switch (which is NOT how it's supposed to work;
LOAD, EXAM, etc. should also clear the RUN flip flop). And nothing
was requesting a data break, either! So after some scoping and
extender card work, I found the backplane wire from the BRK SYNC
flip flop was broken in two. This made for a continuous break
request. I fixed it, now it starts and stops like it's supposed
to,...
And then I looked closely at the backplane and saw three more
broken wires! All of them are crossing the gap between the AB row
of sockets and the CD row. Guess 33 years of flexing and vibration
did them in. They are buried too deep to be mouse bites (besides,
the nest was in the back on the module side).
Unfortunately I ran out of wirewrap wire and the wife rang the
dinner bell. We'll see what ELSE is screwed up next week!
-Charles