[chbot] 6809 Processors

[Volker Kuhlmann]

On Mon 10 Feb 2014 09:02:19 NZDT +1300, Marshland Engineering wrote:

> That made me wonder if I have a RAM problem. The RAM is 6116 and
> 6264 chips. I suppose I have nothing  to loose in changing these and
> they do appear to still be available.

I would check the most likely and the least expensive possibilities
first. That is your power supply. The symptoms of gradual worsening
do not support the theory of an induced noise problem (unless you made
other changes that make the whole system more sensitive to noise), but
do support failing electrolytics. Yes microchips do fail over time, but
PSUs do more so. Those RAM chips are sensitive to supply drops and the
symptom is memory corruption.

You could try to measure the steadiness of the supply near the chips(!),
but it's always painful because you can't assume a useful ground
reference. The osci probes being earthed always severely interferes with
your measurement. Try a differential measurement with both probe grounds
connected, then hooked up to circuit ground at the supply with e.g. a
croc cable, but differential measurements even on very expensive
digilent digital scopes are often impossible because of osci system
limits. Analog rocks... ;-)

Perhaps it's easier (but more epensive) to bang a few new 'lytics around
first? If you can get the whole setup to fail reliably and quickly you
have a good starting point from which you should see big improvements
when you get closer to the problem.

Volker

-- 
Volker Kuhlmann
http://volker.top.geek.nz/	Please do not CC list postings to me.