I appreciate the feedback from all.

Yes Steve I have checked the MB for this setting. The "Keyboard Power
On" setting is/was disabled.

By the way the motherboard is/was an ECS - K7AMA (1.5) with a AMD
Athlon Thunderbird 1.4GHz.

In pursuit of a remedy, needed ASAP, today I placed a new motherboard
in the system and brought it back online. I did not have another ECS
so I am using a Shuttle AK39N.

There is much discussion about the AC being "dirty" so as suggested I
may only be treating a symptom. The surge protector that the problem
computer plugs into is shared by a monitor and speakers. Would it be
unusual for only the PC to be effected by the dirty AC?

I will have to double check but I believe the circuit that feeds this
system is surge protected at the panel. Would that prevent the AC
from being dirty?

jbr

I would like to hear JBR say "Yes, I checked the keyboard power
jumper." It may not seem a logical cause for repeated power supply
failure, but that is what actually happens. Until I discovered this
one computer would lose its power supply within two weeks, in one case
the supply lasted about a minute. I finally read the manual,
discovered the jumper, and the system has lasted since 2000. In
another case a store-built clone was losing supplies, when the second
one failed he brought it to me, sure enough keyboard power was
enabled, changed the jumper, and no more failures. Yes using better
supplies would help, but only to provide enough standby current.
That's why I'm chipping in with this suggestion, it doesn't appear
that others in this thread have encountered this issue, or have, but
don't know it.

Step one, check for this jumper, if it exists even, and assuming it's
a PIII, THEN move on to incoming AC. Logical?

Also I don't find that the supplies "degrade" usually it's the zener
shorting, a resistor overheating, or some such.

. STeve ..