820-3437. No green light - powers on.

ALB-Repairs

Member
Hey,
So this is a fun one. I have a 820-3437 clean of liquid damage.
It originally came in with no green light, i gave the SMC a reflow and got the light back. happy days, ultrasonically cleaned, sent out, all working.

It came back a few days later with the same fault., knowing SMC's are not flip chip or real heat sensitive, i think maybe there is some bad pad wetting or something. I remove and reball the SMC. No dice.
I replace the SMC with another pull, Same symptoms.

8.6V is present (first thing i checked)
PP3v42 is present
SMC_Reset_L is present
SMC_BC_ACOK is present
SYS_ONEWIRE is present
I/O connector and cable are clean- the board has been tested with new replacements.
DC board has 3.4V on the centre pin, suggesting that both SMC_BC_ACOK and SYS_ONEWIRE have made it to the DCIN board with no issues.

U7000 has been replaced for good measure.
Ive ran jumper wires to rebuild important traces to ensure there are no cracks in then traces. No symptom change at all.

So here's whats confusing, all the signals are being generated and pulled up just fine, however the SMC isnt communicating with the charger. I have a feeling it has to be a bad ceramic 0201 capacitor in the SMC support circuit but that in itself is a rabbit hole. Anyone got any ideas?
 

dukefawks

Administrator
I have a feeling there is crap under J9500. ONEWIRE carries data and any resistance/capacitance on this line will cause corruption of the signal, having 3.3V on there is not proof it is working.
 

ALB-Repairs

Member
I have a feeling there is crap under J9500. ONEWIRE carries data and any resistance/capacitance on this line will cause corruption of the signal, having 3.3V on there is not proof it is working.


Ive just replaced the connector, No dice. As stated it was clear from signs of liquid damage so i thought that connector failing would just be odd.

Im yet to ultrasonic clean the flux of the connector witch may just get me there (ive had issues with SIO chips and flux before) but i somewhat doubt that will solve the issue. Any further thoughts?
 

ALB-Repairs

Member
I have a feeling there is crap under J9500. ONEWIRE carries data and any resistance/capacitance on this line will cause corruption of the signal, having 3.3V on there is not proof it is working.


Ive just replaced the connector, No dice. As stated it was clear from signs of liquid damage so i thought that connector failing would just be odd.

Im yet to ultrasonic clean the flux of the connector witch may just get me there (ive had issues with SIO chips and flux before) but i somewhat doubt that will solve the issue. Any further thoughts?
 

dukefawks

Administrator
That pretty much only leaves an issue inside the PCB. What is diode mode and resistance to GND on the ONEWIRE bus?
 

ALB-Repairs

Member
Ive tested it. Diode mode reads open (on an agilent u1252B) and the resistance to ground measures exactly the same as a known good board.
 

dukefawks

Administrator
Diode mode should not read open with red probe to GND and black to ONEWIRE, compare with working board or use different multimeter.
 

ALB-Repairs

Member
sorry, im new to diode mode measurements for diagnosis purposes, until recently i was using a old fluke 8840A bench meter with no diode setting. i had the probes the wrong way round lol.

SMC_BC_ACOK measures 0.55 dead on a known good board.
it measure 0.5428 on the dead board
But on a second known good board we get 0.5364.

I would suggest the data is insubstantial.

SYS_ONEWIRE measure 0.764 on a known good board
0.763 on the bad board.
0.762 on a second known good board.
Again i dont think that is giving me any clues.
 

dukefawks

Administrator
I honestly have no more clues. You can look at the data signal on the scope and compare with a working board. This may very well be a PCB issue on the ONEWIRE bus.
 

Tony Tone

Member
Thanks Duke. I've been having a similar issue (except my diodes to ground were ok), no green light. I guess the DCIN's I was testing with were bad. Found an old one I swear is no good by looks of it and turns out to be good. Been at this one for days.
 
Top