820-2915 green light blink..PP3V42_G3H/PPBUS_G3H cycling.. F6905 diode mode 750

[protocold]

observation:
1) gmux jtag has flux and solder sign
2) L6995 come loose but solder is tight and continuity good

where should I start the witch hunt?

 

[2informaticos]

F6905 voltage fluctuates too?
If yes, try fake charger, or lab PSU.
You can have leakage on input stage; usually Q7080 issue.

 

[protocold]

yes voltage cycling at F6905.

Using a fake charger:
1) 16.46V at R7080 pin 1
2) 12.5v at F7040
3) PP3V42_G3H is 3.45v (measured at L6995)

no Fan spin.

 

[2informaticos]

Then change Q7080...

 

[protocold]

you are right. Q7080 is bad, got it replaced, now green light stay, 12.5v at F7040.

But the fan now start and then stop. No ram.

PP5V_S5 5v
PP5V_S3 missing, diode mode 316
PP3V3_S4 3.3v
PP3V3_S5 3.3v
PP3V3_S3 missing, diode mode 254

 

[2informaticos]

No S0 voltages?
Check all big coils and pages 70/71.

 

[protocold]

sorry my bad... i think because i didnt put in the ram in there and so it shutdown?

Not sure what but the board seems to be ok now.

 

[protocold]

can you educate me what had happend?

1) why a bad Q7080 would cause a green light blink?
2) why using a fake charger would help identify the issue?

 

[2informaticos]

Original Apple charger is smart enough to detect leakage on the board.
In fact it has an embedded processor to control output voltage; only delivers 6V in the first 2s.
Do not imagine that chinese will place such a processor in their f***g chargers.