Grativo wrote:
What happens if you press the reset button on the computer when it
fails to boot with power being turned on? (Leave the power on for this
test)


I'll try this when I get home.


If the pulse from the PS does not occur fast enough
the MB logic will not reset and consequently not boot the computer.


What is the solution to this problem?

Grativo

For most people, it would be to buy a replacement PS or live with the
problem. I have not modified this circuit on ATX power supplies, but on
AT power supplies you could trace the PG lead and look for an R/C
circuit to modify after the comparitors. It was not unusual for the
cap's value to change enough to cause a timing problem. This is NOT
recommended for those with limited electronic knowledge.

One other thing you might explore if you have this computer on a UPS,
is plugging it (the PS) into wall and seeing if there is any
improvement. My UPS is NOT a pass through type and a lower input to the
PS if it was turned on at the same time "might" provide a similar
result. This is only a theory, but it seems to occur with my UPS. If I
defer turning on the computer until after the UPS has been on for a
couple of seconds, I see this problem less on my computer. It could be
coincidence, but it might cause the PS voltage to rise slower?