EVGA 430W PSU Correct voltages but computer won't turn on

[objecttothis]

I had a desktop form factor mini-server die on me a few weeks ago. Through my initial diagnostics, I was convinced it was the motherboard because when I shorted the Green wire on the ATX connecter, the multimeter gave me all the correct voltages on the various pins of the ATX connector and 5VSTB was also correct. To my surprise the replacement motherboard wouldn't turn on with the case power button either. I threw in another PSU and it turned on right away. This is an 80 Plus EVGA 430W PSU with PN:100-W1-0430. I pulled the cover off and don't see any immediately obvious blown caps or scorch marks on the PCB. I'm knowledgeable enough to know that I don't want to go rooting around in this thing without the main cap being first discharged, but the most involved board repair I've ever done was replacing an HDMI port on a RTX 3070. All that said, it's not an expensive part to replace so I don't feel like I must repair it, but I want to try to repair it if there is an obvious part in the PSU that would cause shorting green to ground manually to turn the PSU on but it would not work to turn on the PSU with the case power button (which I assume shorts green to ground) when the ATX connector is plugged in (attempted on two different motherboards). Again, I believe it's something on the PSU because a different PSU works fine.

[stj]

check the power-good signal on the atx connector
it tells the motherboard when the power is stable for the board to start up

[momaka]

1st suspect here would be the 5VSB rail. Can you load it with something like 1 Amp of load and will it still be within spec? ... i.e. 5V +/-5%, (4.75 to 5.25V)
If you have some 12V auto bulbs, that should be easy enough. IIRC, I'm using a 20W 12V bulb, and it draws right around 1 Amp of current at 5V. If the 5VSB circuit is faulty, this should show it.

Now if the 5VSB is tested and found good above...
my 2nd suspect would be that something might have happened to the voltage level of the PSU's PS_ON signal, where perhaps your server mini-board is not pulling it to a low enough level that it recognizes to know that it's supposed to turn on. In other words, if your server mini-board is only pulling the PS_ON signal down to something like 0.5V, but the PSU is expecting to see it lower that, say, 0.3-0.4V for example.
To check for this, connect the PSU to your mini-server board, but do NOT connect it to AC power yet. Next, take a multimeter and connect the black probe (-) to a ground wire and connect the red probe (+) to the PSU's PS_ON signal wire (just shove it in the ATX connector as it's plugged into the motherboard.) Now turn on the multimeter and connect AC power to the PSU, watching what the PS_ON signal line does. Hopefully you should see it go up and stay above 3.3V. Now press the power button on the computer (or motherboard) and see what value does the PS_ON go down to. Also look if the PSU fan spins up briefly and stops or not, as that may indicate a different problem not related to the PS_ON line.

And that's about all I can think of right now.
If you can, maybe post some pictures of the insides of the PSU as well. Sometimes it's good to see what you're seeing, in case we see something that you don't.