820-3662 No Green Light

[hambook]

Hi all, I have an 820-3662 that came in with the notorious crashing issue, swapped out the CPU MOSFETs with good quality FDMF6708's from Mouser and 5mins after the repair tested the board - it powered on w/green light and reached 700-900mA so that part was successful. This was using genuine 85w cable. I ran the board through an ultrasonic after this to remove the flux and gunk from the repair (my machine isn't frequency sweepable but I move the board around every 10-15s and I've never had this machine itself kill a board before) and after the clean and dry I get no green light and the machine pulls 45mA from the charger. Checked PPBUS and found it to be 0.45v, also checked and it seems to be shorted to ground. CPU coils have 12ohms resistance, so is it safe to assume the CPU is still ok and I can go hunting for this short using voltage injection?

 

[2informaticos]

Don't you think to check first if OTHER coil (not ONLY CPU), has the same short?
Why ultrasonic, if only changed the chips?
Toothbrush and IPA is more than sufficient.

What is exact ohm resistance to ground at PPBUS_G3H?

 

[hambook]

Sorry, will make a note to check the other coils too next time and stick to the alcohol/toothbrush method.

Exact resistance of PPBUS_G3H is 1.1ohms

 

[2informaticos]

If no big coil has the same value to ground, then is safe to inject voltage.
However, start with 1V only...

 

[hambook]

I'm finding that L7560 has the same resistance as PPBUS on the side going to U7501. Does this mean U7501 is likely dead?

 

[2informaticos]

I don't understand how can you get different reading on the other pin, if the coil is soldered on the board!!!

Do you get 0 ohm between L7560 and PPBUS_G3H?
If yes, lift the coil one side and check where the short remains.
If is pad 2, remove Q7560 and check again.