820-00165 PP5V_S5 Cycling

[brunomotter]

Hello. Got this board that was not powering on. After inspection I found little corrosion around SMC chip. I then reballed the SMC to remove any corrosion under it and replaced C7536 as it had corrosion on it also. No other spots of corrosion found.

PPBUS_G3H stable at 8.57V
PP3V42_G3H stable on 3.44V
SMC_RST_L stable on 3.38V
PPVRTC_G3H a little unstable, but variating between 3.3V and 3.45V
PP5V_S5 oscillating like roller coaster, same to PP3V3_S5

Tried replacing U7501 and C7505 just in case. Same thing. I compared diode mode between this board and a working one and found this board is giving lower value on L7520(this one gives 0.393 and working on gives 0.409) but I'm not sure if this is a real issue as the values are close. Any ideas on what to check? I tried already tu put flux around CPU and heat it from under side on 250 degrees for some time until flux got liquid and flowed under the chip, but same result.

 

[2informaticos]

0.393 to 0.409 is not significant difference.
You cannot expect exactly the same value between different boards.

Try to start in SMC bypass mode first.
Did you test the board alone on the desk?

 

[brunomotter]

0.393 to 0.409 is not significant difference.
You cannot expect exactly the same value between different boards.

Try to start ub SMC bypass mode first.
Did you test the board alone on the desk?

cannot start on SMC bypass also. Yes, tried it on the desk same result

 

[2informaticos]

Apart of 3V3_S5, do you note pulsing other coils of the board?
Check for short on all coils.
Also check behind Power Fets, page 58.

Please, do not quote an entire post, visible few centimeters above...

 

[brunomotter]

S0 rails except for VCORE pulse. L7320 and L7310 have 2OHM short to GND. Compared to a good board and it should give around 30 do 40OHM... Suspecting bad CPU

 

[2informaticos]

Looks like shorted CPU.
You can lift both coils (one pin) and confirm where the short remains.