820-2262 - no power, no MagSafe light

[2informaticos]

First of all, welcome to the forum!

If PP18V5_DCIN fluctuates, very likely original charger detects some problem on the ACIN Detection, or Inrush Limiter circuit.
U6900 is the first suspect; Q6950 may have leakage too.

Check if get stable voltages using fake charger, or lab PSU.

 

[mikiotty]

Thank you!
Will do as soon as I get to the Mac!

 

[2informaticos]

Please, write directly on the Reply box, instead of clicking Reply button.
No reason to quote an entire post, visible few centimeters above.
We try to maintain forum aesthetic...

 

[mikiotty]

Right, sorry!
I got a steady green light with a fake charger. Replaced Q6950, got a green light with the Apple charger. Tested rails:

PP18V5_DCIN= 18.38V (at F6900 and Q6950 pin 1-3)
PP18V5_G3H_CHGR= 18.38V (at Q6950 pin 5-8)
PPDCIN_G3H= 17.59V (at C7800)
PP3V42_G3H= 3.33V (at L7810)
PPBUS_G3H= 0,31V (at C7155)

After a while, it got back to the initial status and now I have a short to ground on PPBUS_G3H. Q6950 gets hot (alcohol immediately evaporates on top of it) and charger LED pulses again even with fake charger.
Should I inject some low voltage on PPBUS_G3H and check for heat on components?

 

[2informaticos]

What is the exact ohm resistance to ground at PPBUS_G3H?
Compare the reading with the values from the big coils (to ground).

 

[mikiotty]

Ok, I found the short. It was not on PPBUS_G3H, found by disconnecting L7900 and isolating the short to the CHGR_PHASE side.
Q7902 had shorted itself internally to ground, replacing that allowed the board to turn on and have fanspin. The problem now is that the whole 18.5V from the charger is passed to PPBUS_G3H, so other than fan activity (at full blast), there is not much else.

 

[2informaticos]

Replace Q7901; is shorted too.

Be sure D7900 and R7903 are both OK.

 

[mikiotty]

Diode and resistor are ok. Replacing Q7901 (I swear it wasn't shorted before changing Q7902 ahah)

 

[mikiotty]

Ok now it doesn't turn on anymore (a brief little movement from the fan) and PPBUS_G3H is 11.92V
Original charger pulses again.

 

[2informaticos]

Do you get stable PPBUS_G3H with fake charger?
If yes, check 3V3/5V_S5 if present.

"I swear it wasn't shorted before changing Q7902 ahah"
If you think about short to ground, is true; wasn't.
But I didn't talk about short to ground; Q7901 was shorted D-S!!!

 

[mikiotty]

Not anymore, and the short on Q7902 has returned.
What am I doing wrong?

 

[2informaticos]

You must check/change both Q7901/02 at same time.
If one of R7903, or D7900 is bad, then Q7901 will burn again; after that, Q7902 will burn easy too.
Be sure to have good solders and traces.
If Gate pin of Q7901 remains "on air", it will get burnt.

 

[mikiotty]

I’ll check again tomorrow. I noticed that if I spray some alcohol on that zone of the PCB, the MagSafe light goes out and starts blinking… could this give you a hint on what’s happening?

 

[piernov]

Probably a good idea to replace U7900 at the same time as Q7901/Q7902

 

[mikiotty]

Believe it or not, that’s the absolute first thing I did even before writing on this forum. No difference.

R7903 is 2.8ohms vs the 2.2 it should be as per schematic (on my donor board it’s the same thing, 2.8 vs 2.2).
D7900 is .163 on one way in diode mode and 33Kohm in resistance mode the other way.

 

[mikiotty]

After resoldering U7900, I get a solid green light. No PPBUS_G3H and U7900 gets really hot (I cannot touch it with my finger, it will burn me).

 

[2informaticos]

Then this U7900 is bad.
Check diode mode to ground at its pins 25, 13, 26.

 

[mikiotty]

.652 on 25
.533 on 13
.531 on 26

I will buy some more ISL6257s because I don't have them anymore ahah