"-Alby Hewlet" wrote in message ...

I've also learned that you can't trust mobo voltage reporting
at all and must check voltages with an accurate meter (digital)

Even then you may be up the creek. I just finished working on
a system where the problem was the power supply. Had 350 wt ps
in it. All voltages, measured at the molex connector, when it
was plugged into the motherboard and turned on, measured within
specs, yet it couldn't boot the system. A cheap 250 wt spare
I had in the parts bin booted it perfectly every, so you can't
tell just by measuring the voltages either. Well, you can if
you get a bad reading, but all good readings don't mean the ps
is good. If I have it figured out correctly, thanks to Ric in
another NG, the RC circuit regulating the power ready voltage
wasn't rising fast enough. With out a scope you can't detect
that. If it comes up too slowly, the system won't boot.

I had that problem with my old FIC PA-2007 mobo (VIA chipset). It
wouldn't work with my cheapo PSU unless I hit the reset button, but it
had no problems with an ancient IBM brand PSU that contained about
twice as many parts. The cheapo's power-ready circuity consisted of
just a transistor and an R-C network while the IBM's was on a separate
circuit board and had a comparator and some precision resistors and
measured not only the time delay but also all the voltages and even
checked for operation of the power oscillator. I didn't have an
oscilloscope, so I had to pay with R-C values in the cheapo PSU until
the mobo would boot reliably.