Diagnosing 2012 13' Unibody Pro

[sangolossyfon]

Have a pro 2012, board 820-3115. No boot or sign of life except the green charger light. No visible board damage.
G3H shows 13V and PP5V_S5 shows 5.3V, when board plugged in. PP3V3_S5 shows 3.47V.
However, the S3 and S0 lines are dead for 5V and 3.3V.
The only short I have found is a dead short on the PP5V_S3 line. It is a dead short, however, it seemingly 'goes away' about 20 seconds after I unplug the board. So if I unplug the board and immediately measure, the capacitance measure shows 001, and it also shows this if I measure while the board is plugged in. However, after about 20 seconds, it slowly rises to 470, no longer a short. The 5V_S0, 3V3_S3, 3V3_S0 lines do not have any shorts even while the board is powered on.

The fact that it goes away makes it hard to pull capacitors off and test the line for a short one at a time. I don't want to power the board on without some capacitors for risk of causing further damage.

I don't have a ton of experience repairing boards but I've only seen shorts that are always present, even if the board has no power. I don't know why it is going away after about 20 seconds, capacitor shorts I have seen in the past don't go away like that. Thanks everyone for any help.

edit: just realized I have two identical boards with this exact same issue.

 

[2informaticos]

Once power is removed, always allow 10-15s before measuring in ohm scale, or diode mode.
This gives time the capacitors to discharge remanent voltages.

What is exact PPBUS_G3H voltage with charger alone (no battery)?
Did you try to start in SMC bypass mode?

 

[sangolossyfon]

Sorry for the late reply- have been out of town for a few weeks but getting back to fixing these boards.
I actually have three 2012 820-3115-B boards with no boot issues including the two I linked in the above post.
Board #1 fan runs max speed when plugged in no matter what. But no backlight or display, no activity when USB is plugged in, even in SMC bypass. All S5-S0 5V and 3.3V rails are good so suspecting CPU failure. This one had no sign of life at all until I replaced C1768 which had broken loose from the board and now it just spins the fan. G3H voltage 13.22V, PP5VS0 5.23V, P3V3S0 3.51V. Draws 18W from wall.
I would just like to know how to diagnose this one further especially with respect to the CPU being good or bad.
Board #2 No sign of life. No shorts on any 5V-3V3 S5, S3, S0 rails. No fan spin even when attempted SMC bypass. G3H voltage 13.23V. Draws about 1 watt from wall in regular or SMC bypass mode.
Board #3 No sign of life, but does spin fan max speed in SMC bypass mode. But even in this mode it has the same symptoms as board #1- no display etc. No shorts on any 5V/3.3V rails. G3H voltage 13.33V. Draws 0.6 watts in when plugged in (no fan) and 6W when plugged in and in SMC bypass mode (full fan).

No boards have water damage signs and were tested in a working enclosure.
Thanks for any help.

 

[2informaticos]

Please open separate thread with each board.
Their failures have probably different solutions.

 

[sangolossyfon]

Thanks I will do that. I was wondering though so that I could rule out the CPU being shorted on any board- What is the best power rail to check for a short that would indicate a fried CPU?

 

[2informaticos]

CPU coils on these boards have very low resistance to ground.
But not only short means bad CPU.
It may have various internal failures...