Tony, could the 1802 be in LOAD mode (CLEAR low, WAIT low)? You'll
get DMA-IN cycles in that mode.
No, CLEAR and WAIT are both high.
However, I did some more testing this afternoon.
The SC1 = 1 pulse occurs during TPA and SC0 is low at that time. The
thing is duing an interrupt acknowledge cycle. And a bit more trsting
showed that there were pulses on the INT/ input, not going low enough to
be dectected as a sloid low by my logic analyser, but low enough to
trigger the 1802.
And those came from an RC circuit that's working correctly. The idea is
to generate an interrupt about 0.1s after the last one, thus giving some
kind of heartbeat clock. This is doing the right thing.
After doing some more tests, I have come to the following conclusions :
1) The CPU is working fine
2) All the small glue logic chips are fine
3) The CPU is executing instructions from ROM, cna read/write RAM, etc
So my current theories as to the likely fault are either bit-rot in the
EPROM (but a dump of it looks reasoanbly sane, there is readable text for
the setup messages, etc, so it's not totally dead) or garbage in the RAM
that needs to be initialised out.
In either case I need to find somebody who has one of these machines
(It's a Microwriter MW4/10). The ROM is socketed, but it's a 2564 which
has rahter different pinout to a 2764, and some programmers can't handle
it. Mine can, and I even have a few blanks in the junk box.
But my first though is that it's just an initialisation problem. I do
need to find a user manual...
-tony