A picture of the mainboard´s flipside (PSU connector area) would be helpful.

This was a direct short for the PSU itself, the board itself should still be ok.
EDIT: I didn't notice there was a component near the burned area. This can be much more severe, although most likely still fixable.

You should remove the SMD component right underneath the block spot on the AUX connector before applying power again. And check underneath for similar black/bad components.

Then check for a short from 5v to GND, and 12v to GND. If no short, connect power and try again.

If you need the exact SMD components to replace, the SMD under the black spot on the AUX connector is likely the same value as the one next to it. If you get a clearer photo you could work out the value, or just post the photo.

MikeSG wrote on 2024-08-21, 09:44:

You should remove the SMD component right underneath the block spot on the AUX connector before applying power again. And check underneath for similar black/bad components.

Then check for a short from 5v to GND, and 12v to GND. If no short, connect power and try again.

If you need the exact SMD components to replace, the SMD under the black spot on the AUX connector is likely the same value as the one next to it. If you get a clearer photo you could work out the value, or just post the photo.

Failed component is most likely a resistor and should be easy to replace if one can find the correct value for it. If it is a resistor and it has released the magic smoke out, that circuit is pretty much 100% failed open and system most likely won’t work at least correctly as it may be missing some power line alltogether.

Additional photos as requested

That looks like an internal short in the board, you will have to remove as much of the burned board as possible and make sure there are no shorts on the inner layers. The carbonized epoxy can be quite conductive and it will take some effort to fully get rid of it.

Did the standoff pierce the board ? I'm not seeing any marks on the "PS-SUPPORT" pins, the area where the burn is has no exposed copper other than a via right in the middle of it which I guess that standoff managed to connect with ?

Tiido wrote on 2024-08-21, 13:26:

That looks like an internal short in the board, you will have to remove as much of the burned board as possible and make sure there are no shorts on the inner layers. The carbonized epoxy can be quite conductive and it will take some effort to fully get rid of it.

Did the standoff pierce the board ? I'm not seeing any marks on the "PS-SUPPORT" pins, the area where the burn is has no exposed copper other than a via right in the middle of it which I guess that standoff managed to connect with ?

The standoff is blackened but did not leave a clean "hole" in the board--the black charring on the top of the board looks like it was starting to burn through but not completely.

So, the $500 question -- is there someone who does this kind of repair work, or do I buy a new board? As we've demonstrated, I can (mostly) plug things into the board, but I can't repair the electronics on the board itself

If you are in the UK, send it to Sorin Electronics, AdamantIT or Tony359
If you are in the Canary Islands, send it to Richard from LER
If you are in the US, send it to Alix from Northridgefix

As a bonus, we will all have some nice video in youtube to show us how it's done. 😀

(note: some of these are businesses, but I'm not endorsing or promoting them, these are just people I know from youtube.)

Did you test the board after the incident?

ux-3 wrote on 2024-08-21, 20:08:

Did you test the board after the incident?

I was thinking the same thing. With a bit of luck, it's only a connection between the main power connector and the auxiliary power connector.

On the other hand; it may be an expensive gamble ...

Yes I did give it another quick go--I removed the errant standoff and plugged it back in and hit the power button. More smoke. Pulled the plug, haven't plugged it in since.

That strongly suggests the carbonisation is shorting some inner layers. The burned area will need to be excavated and cleaned up until all of the burned bits are gone. Anything that is left can continue to cause a short, which in turn causes more surrounding parts to carbonize and feed the fire so to speak. I have done this once in past and it was quite a pain, I found a dremel type tool with a round shape bit is most useful for this. One can use alcohol soaked cotton swab to see when things are clean, once it stops staining black, the area is good to go. Magnification is helpful to make sure the copper in the inner layers doesn't touch each other, the separations are quite tiny...

The question is: what happened at all?
It doesn't look like there was any short at the PS-Support connector nor does this SMD component on the top side really look destroyed. It's more like the via next to it took all the load.
The standoff in the case is somewhat at GND potential, so to generate a short this via must be (have been) at some other potential, like 5V or 12V.
Shorting it to GND must have generated enough heat to damage the PCB material around it potentially up to the point that now there's an inner short between the layers of the PCB.
If that's the case you should be able to read some very low resistance when measuring between GND and the other potentials, for example directly at the ATX power socket. Make a note of what's shorted while making the measurements.

For repair I'd suggest first to try to remove that black mess (mostly molten epoxy) with some small tool like very fine tweezers or a sharp tip of a knive/cutter but without messing around with that SMD resistor next to it. You should be able to get a clearer image of the damage. Clean with a small brush and some isopropylalcohol.
If you can isolate it to the area just around that via it should be possible, following Tiidos idea, to drill that out just large enough so to get all damaged part from the board, thereby removing the assumed short within the PCB. Use a sharp/new drill. Up to 5mm or 6mm seem to be needed but start with a smaller one, 2mm or so. Test resistance between potentials between drill steps. The area around isn't densely populated so there shouldn't be any tracks within the board. But it won't hurt to check before by holding the PCB against a strong light source. If there are tracks shining trough repair will get difficult, if not (indicating a closed copper plane) drilling there should be safe. Repeat from both sides!
Check the drilled area with magnifying glasses/microscope/smartphone cam zoomed in. It should be possible to see the separate copper layers within the PCB.

After that worked out good test the resistor value. Or maybe it can still be read after some cleaning!? The missing link from that resistor to the now removed via can easily be patched with some wire. You'd need to know the correct potential it was connected before though. Maybe it was the same as the resistor next to it is connected too. But that's hard to say from the pictures.

If that all went ok you can only hope there wasn't any 12V shorted to may be 5V or others so that components on the lower rails could get damaged by a too high voltage.

All in all not an absoulte beginners work but also no rocket science either 😉
Good luck!

ChrisK wrote on 2024-08-22, 06:31:

The question is: what happened at all? It doesn't look like there was any short at the PS-Support connector nor does this SMD co […]
Show full quote

The question is: what happened at all?
It doesn't look like there was any short at the PS-Support connector nor does this SMD component on the top side really look destroyed. It's more like the via next to it took all the load.
The standoff in the case is somewhat at GND potential, so to generate a short this via must be (have been) at some other potential, like 5V or 12V.
Shorting it to GND must have generated enough heat to damage the PCB material around it potentially up to the point that now there's an inner short between the layers of the PCB.
If that's the case you should be able to read some very low resistance when measuring between GND and the other potentials, for example directly at the ATX power socket. Make a note of what's shorted while making the measurements.

I think the stand-off was GND, and the via was 5 or 12v (from main ATX supply), through a 120-ohm resistor to the unconnected PS-Support header. If the line goes to anything else other than the header it needs to be checked...

The via at the least needs to be unwelded from GND-power, and the resistor removed.

The board may be 4-6 layers with the GND and power on the inner most layers.... If it was my board I'd have a go with an engraver dremel attachment (ball with fine diamond dust). A repair shop would do the same but charge hundreds of $$$.

After the via is separated form GND-power, and it only goes to the unused PS-Support Header then nothing else needs to be done. If it goes to a circuit then you need to find out the voltage (5 or 12) and run a new wire to a new resistor (not through the via).

are you sure that this board are using standard ATX PSU layout and not some-obscured-non-standard psu pin?
//My bad, the board work fine in bench and i guess the standoff create a short which destroy the trace and cause the crater