For sure. The back of my board is covered in a thick layer of flux and it looks
like the type that likes to eat traces over time and sometimes even becomes
conductive. Typical 80s flux. I wish I had a board washer at home because they
don't like us playing with the one at work. I'll just have to drown it in
alcohol. The board, I mean. Heh...Scott M.Sent from my T-Mobile 4G LTE Device
-------- Original message --------From: Jeffrey Birt <[email protected]>
Date: 12/17/20 1:15 PM (GMT-05:00) To: [email protected] Subject: Re:
[M100] Model 100 Repair - Keyboard not responding, LCD missing a column In the
past few weeks I have seen a couple of very odd problems with M100s. #1) One
tow of keys (from matrix point of view) showing up as the previous row (like
row 2 showing up as row 3). Looked at with a logic analyzer and you could see
the signal from each row. However, this was not a simple short it was a one-way
phenomenon. Pressing a row 3 key did not show a signal from both rows on the
LA. I pulled all components related to KB scanning and could still measure a
‘short’, which acted like a capacitance, between the two rows. This was caused
by some contamination on the PCB. Cleaned the PCB, reinstall components and it
was fine. #2) LCD corruption, serial loop back failure and 20ma of excessive
current draw. The LCD had missing blocks, areas of gibberish and some areas
were OK. Otherwise, it seemed to work fine. Trouble traced down to data bus
contention caused by M23. It was amazing it worked at all. #3) T25 in the
reset circuit died while I was testing something unrelated. The particular M100
failed to boot sometimes but I attributed it to my test set up. When it quit
booting all together then I was forced to look for the root cause. T25 measured
fine in circuit with the ‘diode test’ on the multimeter but was leaking enough
to have 0.6v on the base with 0V being applied my M28. This was not letting C78
to charge up, it was only getting to about 0.8V and the machine stayed in
reset.Point is very strange things can happen on these boards. My rule of thumb
on a vintage computer is to check voltages, the reset circuit, clock, and then
for bus activity. A logic probe or LA can mislead you with bus activity as it
will now show you issues with marginal voltage levels or bun contention.
FWIW,Jeff Birt (Hey Birt!) From: M100 <[email protected]> On
Behalf Of Stephen AdolphSent: Thursday, December 17, 2020 10:38 AMTo:
[email protected]: Re: [M100] Model 100 Repair - Keyboard not
responding, LCD missing a column Scott, if you have a scope that will be
extremely helpful. In my experience you can sorta tell when the CPU is
booting, and has access to RAM and ROM, by looking at the signals on the bus. I
typically check out the signals at the main ROM first, then go to the nearest
RAM. Those parts are at the "end of the line" for the address/data bus, so if
you have an open trace, you will see it there. If the Main ROM and 1st RAM
have good signals, it should boot. Also worth it to check that the CLK signal
on the 8085 is showing 2.45 MHz. I also check /RD, ./WR, IO/M etc. good luck,
and ask questions.Steve
