rasz_pl wrote on 2022-02-09, 23:29:
Problem with that cap theory is its position, it would have to leak against gravity.
It looks like MC34063 is driving Q17 Q18? and pin 1 where that resistor seems to connect also provides main power to the transistors.
Geofand26 did you by any chance forced CPU in wrong orientation and powered the board?
I was wondering something like that, but I can't figure out why that trace would get hot. I had it wrong before, and thought that trace was an output from the switch controller. But I couldn't make it make sense with the datasheet. Now I think that trace is a 12V supply to the switch controller (maybe routed from the nearby fan header), to give it the drive strength to turn the FETs on and off quickly. But it should be protected from a short on the CPU by those FETs, I thought it would only be driving their gates. If it started switch too fast then maybe it'd draw enough current. But wouldn't a CPU short just cause the high side FET to be on all the time, so the area around the FET would would burn, but not the 34063? There's a timing capacitor on pin 3 that controls the switching speed, but that look ok.
I think the circuit is probably similar to page 7 on the datasheet: https://www.onsemi.com/pdf/datasheet/mc34063a-d.pdf, with the external PNP option replaced by a nFET (Q18, using 12V to switch it, so Vgs is large), and what I think is a big diode at Q17.