Same procedure to identify defective PSU also identifies
other defective components in that power supply 'system'.
Some caps will be part of the power supply controller located
on motherboard.

Again, if pin 9 is the purple wire (I don't remember), then
we have the culprit - or at least a usual suspect. If bulging
capacitors are connected to that power, then short probably
has been identified.

To properly identify that capacitor connects to pin 9, your
meter must be a newer design - so that its voltage does not
blow through and maybe even overstress IC P-N junctions. IOW
the meter's continuity tester must output no more than 0.7
volts and must limit current to single digit milliamps or
less. Some older meters output too much voltage and may
measure continuity through ICs where current is not expected
to flow. A low voltage continuity meter will only report
actual connections.

Using that meter, identify every electrolytic capacitor on
motherboard that connects to pin 9. Are they bulging? Those
capacitors can be replaced very cheaply - typically for about
$1 per cap. (BTW make note of which capacitor pin is ground
and which connects to +5 since electrolytic are polarized and
sometimes are installed backwards both during manufacturing &
repair.)

Above assumes no other peripheral is also shorting that 5VSB
(and I am assuming that pin 9 is the purple wire because the
question was not answered). IOW disconnect all peripherals
such as IO cards, keyboard, etc just to confirm they too are
not shorting that purple wire +5 volts always on voltage.

Don't remember if anyone mentioned it. But you must remove
power cord from wall before removing anything inside the
machine. If not, that five voltage should be powering
components as you remove them - bad for electronics not
designed to be hot popped.

Al wrote:
Sorry, Tom. I started a new thread asking about the bulging capacitors
in the computer. From all the replies about bad caps (and because a
known working ATX PSU showed the same behavior as the original PSU), I
quickly became convinced that the bad caps were the problem, not the
PSU.

Thank you for the help.

Appreciate that without removing anything, we can pretty much
isolate the problem down to one of three components in a power
supply 'system'. So which voltages rise on power up? Which
voltages do not? And what are numbers for other voltages
that, for example, must always be there?


as I understand it, pin 9 should always have 5v - but when I measured
between 3 (ground) and 9 I got the same as I always get anyhwere on the
back of the ATX connector: a short tiny movement of the analog meter.
I'd guessed that an analog meter's needle can't react quickly enough
to adequately display a momentary voltage, so I don't know what the
voltage really was. (It showed maybe .5 volt.)