820-01949 20v 0.06-0.35a

[Isolated]

Got this horribly liquid damaged, corrosion basically everywhere but somehow T2 and CPU seemed to have dodged most of it. Looks like it flowed around them rather than through. Replaced U7000, U6903, and U7100 where some of the worst of the corrosion was at which fixed stuck at 5v. Ran the board through my ultrasonic cleaner and currently get 20v and current slowly rises from 0.06a to ~0.35a then drops back down to 0.08a. I am still pretty new to board level repair so I apologize if I don't understand some of the jargon initially.

ALL_SYS_PWRGD: 0V

PVDDQ_PGOOD: 0V​

P1V1S3_PGOOD: 0V​

CPUVR_PGOOD: 0V​

PPVCCIN_S0_CPU: 0V

I have tried with a good battery, it is not seen in Configurator 2 but it does seem to be charging the battery at 19.5v 2.3a but still does the same thing, no power, PCH gets that warm dot.

I hope this is just something to do with ALL_SYS_PWRGD because I am starting to believe the PCH is dead as you'll see in the attached photo. When the current starts rising, various IC's will begin warming up some along with T2, then PCH gets the warm dot and amperage goes down to 0.08a.

Thank you!

 

[2informaticos]

First of all, welcome to the forum!

Post diode mode to ground readings on the big coils.

Looks like board tries to boot.
You should get voltage on some coils; which ones?

 

[Isolated]

L7400: 115 / fluctuates anywhere between 0.40-0.70v
L7230: 055 / 0v
L7220: 055 / 0v
L7210: 055 / 0v
L8100: 002 / 0v
L3500_VWR: OL, resistance measures ~16.5k / 0v
L7660: OL, resistance measures ~100k / 3.36v
L7030: OL, resistance measures ~630k / 12.55v
L3500_VXT: OL, resistance measures ~16k / 0v

Tested voltages when it attempts to boot as well, same results.

 

[2informaticos]

"L8100: 002"
Seems that PP1V1_S3 power rail is shorted to ground.
Lift L8100 one side and check where the short remains.

About OL readings, test again in diode mode; red probe connected to ground!

 

[Isolated]

Looks like the short is on pin 2 of L8100, removed R8118 and the short is on PP1v1S3 like you said. Injecting 1v all RAM modules change color slightly but are almost not even noticeable on thermals. Since voltage injection isn't really an option as it won't create enough heat, should I just start lifting components on the 1v1s3 line or do you recommend a better method?

Red probe ground
L3500_VWR: 495
L7660: 435
L7030: 516
L3500_VXT: 493

 

[2informaticos]

0.3 ohm is really pure short; 1V should heat something.
Use powerfull cables to connect lab PSU; you need to heat the board, not the cables!
You can also increase voltage until 1.5V there.

At least one RAM chip is fried I suppose...

 

[Isolated]

The only thing that looks to get any sort of thermal signature is all 4 ram modules. Also, if I try to increase voltage past 1.3v current drops from 0.17a to 0.04a. Not sure if that will give you any extra info but I guess it's time to start pulling RAM chips for now.

Thanks for the help, I'll post an update if anything changes or I am successful.

 

[Isolated]

Removed all 4 RAM modules, and looks like the short is gone... after I removed the last one (U2300_1). Thinking about it now, I should have taken that one off first as its closest to the 12v MOSFET for PPBUS_HS_CPU so I guess it had a higher chance of damage. There is a rainbow looking mess under it so hopefully, the module is still good and the balls were just shorting.

Will continue with further updates

 

[Isolated]

Confirmed RAM module fried, reballed and resoldered just the one to board and short came back. Any way to just replace a single module or do they need to come as a set of 4? Looking online they seem to be hard to find, guess I could always buy a donor board.

Thanks for the help, unfortunately, this is looking like a dead end.

 

[Isolated]

Forgot to attach photo of something I noticed on the underside of the bad RAM module. The black mark is inside the chip and connects with the pin that attaches to PP1v1_S3. I don't see that black 'burn' spot on any of the other 3 RAM modules.

Just posting for anyone who is curious or for future repairs people may have.

 

[2informaticos]

You may have all RAM chips bad, not just one.
Check in diode mode between ground and power pads of the chip.
Use boardview to identify corresponding pads.

 

[Isolated]

Look like you are right, all RAM chips are fried. They all have 15ohm to ground but at least the short is gone on the board so maybe I can source some replacements somewhere.

Thanks for all your help, this is pretty much resolved.