Pete Turnbull wrote:
Yes, Jules' machine has the first issue. It
sounds like he has almost
exactly the same version of machine as I do; mine has the same MPS6540
ROMs and MPS6550 RAMs. Except for the dead ones that I've replaced,
that is: many of those old MOS Technology chips are dying of old age,
and I've replaced three RAM chips in my PET in the last two years.
I had a bit of time to play with RAM chips today, but unfortunately only with
a pile that are also an unknown quantity. By swapping out various banks of RAM
in the dead machine, I could get four variations of behaviour:
Totally black screen on power-up
Garbage on power-up, clearing to black screen after < 1s
Garbage on power-up, *partially* clearing to black screen after < 1s
Garbage on power up, no change over time.
The partial screen clear's rather odd - the bottom 5 and a half or so lines of
the display would be cleared, but the rest would stay as garbage. Possibly the
screen-clear routine was running but crashing halfway though...
I've never seen a working 2001 series PET, but a totally black screen clear
and then apparent hang definitely isn't right behaviour :)
Anyway, Ethan's point about checking the board for problems seems sensible at
the moment. I'll try reading the ROMs out too.
I'd also like to test the RAM - I've got two choices there...
1) Somewhere we have a RAM tester that may well cope with the ICs used in the
PET - problem is that it got packed up when we were going to move
buildings/sites and laying my hands on it in the near future might be tricky.
2) I could maybe build something to hang off the PC parallel port and do the
job. It probably only needs a couple of latch ICs to latch the address, plus
control lines wired to the parallel port. Only downside is that I wouldn't be
able to drive the RAM at the frequencies found in the PET, so it might not
show up all errors.
Getting this sorted isn't time-critical at all, so I could wait for the RAM
tester to surface and take it from there - but the satisfaction of knowing
that it was done and all 100% working would be nice!
cheers
Jules
ps. intermittent video turned out to be corrosion on the pins at the back of
the CRT itself, so at least that problem's gone away.