All,
Just catching up on the digest. I think next time I beg for advice I'll
take digest mode *off* first. The delay is frustrating. Tony was way ahead
of me last night, but I didn't get the benefit of his advice until this
morning.
Tony D. said:
I will bet that
Red = +5V
Black = Gnd
Orange = +12V
Blue = -12V...
You'd win, as far as I can tell with my meter, with the minor variations
I've noted.
Blue is one
wire, and goes to appx. 12 V. when the system is on.
I would think the last one is actually -12V ?
Correct, typo on my part.
More intersting would be to
use a 'scope to look for noise (particularly spikes at the PSU switching
frequency). Such spikes can make a supply rail look to be too high on
some meters.
Yeah. Sigh. What's a likely frequency for the PSU switching? Is my
(primitive analog) VOM likely to yield any useful info if I set it to the
"AC Voltage" ranges?
M9124
LM393N
QST
Aha. A dual voltage comparator. This could be used to check if power is
OK, and to generate the reset signal as appropriate.
I don't know the causality yet (failure causes comparator to switch, or
comparator switching causes failure), but the comparator (at least output
A) does switch when the failure appears. Output A is Pin 1 as you recalled
and Output B Pin 7, thanks to Toth for the link to the data sheet.
Now *I'll* bet - Tony wants to know what the comparator *inputs* are
connected to. Tonight's work is cut out for me. But in any case, I'm still
guessing the power supply is a likely culprit. It looks like the comparator
is properly protecting the mainboard from an overvoltage on +5V or +3.3V. I
don't know what's going on with +3.3 being referred to something other than
ground. In either case, a lower +5V supply (= appropriate repair to the
PSU) would solve the problem.
Is replacing all of the PSU output electrolytic capacitors just on
suspicion a good move? That's what I plan for the weekend. (Lucky for me
capacitors are not entitled to due process.. :-) )
- Mark