820-01521 20v, no boot

[captainmac]

Need a little help, if you don't mind.
This board was liquid damaged, but very minorly: the audio board looks like it took a hit on it's audio amp chip.
the main board looks pretty good, with just some minor spots around U7290 and Q7291, replaced the whole area, controller, buck and affected resistors and caps around it.
There were no shorts on ppbus or any coils, so i decided to try and turn it on:

20v, 0,089Amp draw,
Looks like it makes it through the poweronsequence as far as SSD_PMU_RESET_L, and nothing after that (no SSD voltages at all)
PPBUS is 12.6V by the way.

I have tried to get it into DFU by jumpering FORCE_DFU pads to 1V8_Awake, but the machine won't be revived, fails with some error about transitioning between DFU states.
By the way, whats weird is that i have to keep the jumper on FORCEDFU pads present, if i snap the wire, a couple of seconds after i see the machine pop up in AC2, it disappears...

I have measured for shorts on the SSD rails, but found none.

Any hints what to do next?
Thanks in advance!

 

[2informaticos]

Very likely damaged T2, because of leakage from audio board; PPBUS_G3H to data lines.
Check diode mode at J6100; on the data pins nearest to PPBUS_G3H.

Tried forcing SOC_FORCE_DFU to 1V8_SLPS2R, not AWAKE?

 

[captainmac]

diode mode readings on the entire J6100 all look perfectly normal (i used openboarddata for reference, all datalines match up with those readings)
Tying FORCE_DFU to 1V8_SLPS2R didn't change anything by the way.

I realised i didnt completely finish diagnosing correctly yesterday, probably was tired:
I found that all voltages up to PP3V3_S5 are coming on, for a few seconds (why i missed it yesterday probably wasnt fast enough).
PP3V3_S5 is shorted with a 0.4 Ohms resistance to ground.

I decided to inject some voltage because of such low resistance:
Can't feel anything getting hot, some general areas are getting warmer,mostly around U2800, which surprises me, because i can't see the link between pp3v3_s5 and U2800?

 

[2informaticos]

What voltage did you inject?
You should get sufficient current to heat something, having almost pure short there.
Don't forget to set amp limit to max on your lab PSU.
BTW, 5A PSU may be insufficient for this case...

 

[captainmac]

Jesus.. im sorry, my bench-mat was still hot from work i did earlier, so the heat around U2800 was coming from that.
I was using shit-wires before, changed those, I injected about 2.8v maximum, it was taking about 1A and only the PCH-side of the CPU is getting hot..

 

[2informaticos]

2.8V into less than 1 ohm (reported 0.4 ohm) should get near 3A there.
If PCH gets hot with 1A, is enough to point as dead PCH however.

 

[captainmac]

Yeah I’m pretty sure my resistance measurement was correct and was surprised it took that high voltage to get to about 1amps..
My psu might not be very accurate.

damn I don’t like no-fixes