Firebat T8 Plus T8PLUS_V12 damaged during probing

[spleenharvester]

Hi all

I added a Noctua fan to this dual NIC system at the DC jack, but this of course meant the fan was always on. I was probing for a 12V RUN point and my hand slipped. Now there are two problems - (1) the NIC closest to the power jack is now completely dead, no LEDs and not detected in device manager, and (2) the system is in a constant state of powered on - if the system is shut down it just turns on again a few seconds later. Apart from this the system and the second NIC are still working fine.

I've added a red arrow to where I was probing when my hand slipped, and the smaller blue arrow shows the direction in which my hand slipped. I definitely shorted the SMD electrolytic cap to the SOT23-6 IC but there was also a spark by the now-dead NIC and I'm not sure what I hit there. The SOT23-6 IC seems to read D5KE or D6KE with a 32 at the end, but that doesn't return any results so I'm not sure what it is. Have I probably just killed this IC or have I damaged something further downstream? I can live with the broken second NIC but the issue with the power being stuck on is really irritating. (It's occurred to me I could get an e-key RTL8111 NIC as I have a slot spare and either wire it separately in the case or somehow wire it to the existing dead port)

If anyone has a schematic for this board that would also be awesome, but I doubt there is one circulating.

Thanks

[spleenharvester]

Okay so I see there are various forms of RTL8111, if I am interpreting correctly it looks like some can take 1V and some can take 3V3? And I know power switches tend to involve shorting 3V3 to GND. Is it possible that the SOP-23-6 is a 12V to 3.3V step down that I have killed and as a result both the power button and nearest NIC have failed, but that the working NIC is powered from the 1V rail so is unaffected? I'm not sure anything else on this board takes 3V3 so it might explain why nothing else has stopped working?

Only SMD parts starting with D5 I can find that might be relevant are NPN/PNP transistors, possibly ROHM EMD5/UMD5N?

E: Taken another look, it's definitely a SOT23-6 so not the ROHM IC. I've attached continuity tests from that area, it looks like I have definitely killed that IC at minimum as 4 of the legs are shorted together but they are not to ground. Now just the fun of trying to identify it? I'm wary of taking any measurements while the board is running as my tremor is what caused me to slip to begin with and I am worried about causing further damage.

[spleenharvester]

Someone on another site has identified it as probably being an AO6405 MOSFET so I am going to replace it when I get a slow day!

I guess this explains the inability to switch off but the NIC is a different fault. Nothing obvious probing round the area, I assume I have killed the RTL8111 itself so might just have to accept it.