Have you removed the PSU from the power for a good 5 seconds then retried?

Same board from your previous post - Gigabyte GA-6BXDS CPU support - was this ever a working system or just the parts you had available?

Could the board be the problem rather than the (faulty) PSU - did you test the PSU separately, else maybe if truly faulty it's damaged the board somehow.

If it was a working build once, what else is involved - cpus (1 or 2 / if 1 only then it should be in the CPU2 slot) / memory (what type & how much) / what AGP or PCI video?

Can't see potential problem with front panel PWR switch - standard 2 pin mb header for a matching 2 wire momentary switch (orientation doesn't matter). Is this not what the Aopen case has?

Not sure what switch you opened to check - momentary or some other - any pics / please detail.

Beyond that maybe try cleaning all contacts (board & cards) / new coin cell battery!

I got the system whole, it had been in use by a school before being recycled. The Aopen case has an AT-style switch but wired to a two position header, and the one time I got the system to run I actually had to power it off via the switch.

The board has two 400MHz Pentium II CPUs installed and the only thing I can think of that I did that would affect the system is swapping a PCI video card for an AGP one (an ATI Rage Pro that worked in a previous system). Could it be that somehow that AGP card is stopping the entire system from turning on?

EvieSigma wrote on 2021-11-22, 18:24:

I got the system whole, it had been in use by a school before being recycled. The Aopen case has an AT-style switch but wired to a two position header, and the one time I got the system to run I actually had to power it off via the switch.

The board has two 400MHz Pentium II CPUs installed and the only thing I can think of that I did that would affect the system is swapping a PCI video card for an AGP one (an ATI Rage Pro that worked in a previous system). Could it be that somehow that AGP card is stopping the entire system from turning on?

Can't see whay the Rage Pro would be the issue - it's a good late 90s contemporary to the motherboard (the manual lists another from that era in the performance tests - the GA-601 which is based on the 3Dlabs Permedia 2A chip). Suppose it's possible the AGP slot is faulty - did it come with the PCI video (which one) or is that something you added. If you haven't already, maybe worth trying PCI again.

I'm assuming the AT-style switch acts as momentary rather than latching, else the board would normally start then stop 4 seconds later! and as you've already failed to start the board by bridging the Soft PWR pins with a screwdriver I'm not even sure the PSU is faulty (unless there's some other issue you've not mentioned).

As always, best advice in these cases is bare essentials - strip the system right back by removing the board from the case and test only with one cpu (again this should be in the CPU2 slot), one stick of memory in the first slot (bank 0), AGP or PCI video, a fresh battery and a PSU. This makes it much easier to swap out individual items for subsequent re-test.

Other than that, can't offer much else atm.

I don't know what was wrong but I checked the power switch header, reseated the AGP card, and the system fired right up. Computers are weird.

They sense fear, never show fear.