820-3115-B powers on briefly, corrosion near ISL chip?

[Hardwood T/S]

Hey! This time I actually have a 2012 13" board, 3115-B. Similar problem to last time, except this guy gets no power even with battery plugged in. When AC adapter is attached, sleed LED lights and fan spins for a quarter-second, then rests. Board does power on in SMC bypass mode.

Board had corrosion near the ISL chip, specifically northern part of chip and above components--cleaned with IPA and resoldered to remove corrosion on pads--nothing. 3v42 is stable, however rails such as SMBUS_SMC_5_G3_SCL and SMBUS_SMC_5_G3_SDA phase up to ~3.475 and then down to ~3.445, seesawing back and forth. Should I try replacing ISL?

Measured the resistance of the northern components that had some corrosion against a good board, all checked fine.

 

[JohnB8812]

Check resistance between pins 17/18 and 27/28 of the ISL6259. Sounds like a current sensing issue

 

[2informaticos]

If it starts in SMC bypass mode, then is sensor issue for sure.
Check voltage at pins 9 & 15 of ISL6259.
Depending of values, you will know if problem is related with R7020, or R7050.
Or measure directly between pins, as above mentioned.

 

[Hardwood T/S]

Checked the pins, 9 was at ~30.5mv, which was about ~3mv lower than a working board, and 15 was at .1mv, which was .2mv lower than my working board. Can't get a great reading on pins 17/18, but pins 27/28 are normal resistance. R7052 and R7051 read correctly, as does C7050. Thinking 17/18 may still be an issue tho, any ideas?

 

[2informaticos]

"Can't get a great reading on pins 17/18"
Why great?
You must read aprox 003 there...

Check pins 9 & 15 when board is on in SMC bypass.
Also check all sensors output from page 50 and 49 too.

 

[JohnB8812]

Post your measurement in ohms between 17/18 and 27/28

 

[Hardwood T/S]

27/28 is 20ohms, which I know is a good reading. 17/18 I'm not getting anything, but C7050, R7051 and R7052 all read correctly (3ohms) for their resistance. Diode mode shows very low readings for all three, however, I'm thinking of replacing that capacitor. At first I thought the traces could be corroded between ISL and C7050 due to the corrosion there, but when I jumped wires the issue got worse instead of better (dim blinking amber magsafe instead of solid green but no power), so I removed them.

I've noticed my pin 8 is also showing very low, usually it's supposed to be around 2v but it's showing something much lower, I believe around ~70mv. Again, the associated capacitors and resistors to that pin are measuring correctly against a good board. I've noticed a few of my boards have low voltage on pin 8--I'm guessing a related cause? Replacing the chip didn't fix anything.

 

[2informaticos]

Forget C7050, is not critical: board works without it.
R7051/52 can be replaced directly with tiny wires.
If both resistors are in place and have correct value, you must get 000 before them (to R7050 side) and 003 after (to ISL6259 side).
If you replace both resistors with wires, you must get 000 between pins 17/18.

"when I jumped wires the issue got worse"
Did you check resistance to ground at F7040 after that?

 

[Hardwood T/S]

Now I'm only getting a dim blinking amber light when magsafe is plugged in--C7050 is off the board, R7051 doesn't seem to be resisting properly now as well, measuring only around .6ohms instead of 2.5. Pin 8 is still very low, not sure exactly where to proceed.
What do you mean replaced directly with tiny wires? You mean just solder a small copper wire between the pads where R7051 would be?

Also, another board I'm working on has a similar problem, with the light just being dim green instead. All voltages are present and normal around ISL except for R7001, which is showing nowhere near 5v. It's resisting properly around 5ohms, but the capacitors to the north of it (C7000/1) are not showing any ohms in resistance, and 0.00 in diode mode.

 

[2informaticos]

I hope you didn't measure continuity (diode mode, or resistance) on a powered board.

"Did you check resistance to ground at F7040 after that?"
And still waiting...

"You mean just solder a small copper wire between the pads where R7051 would be?"
Same for R7052 too.

R7001 should have 5V on both pins.

 

[Hardwood T/S]

ofc not while powered. I'll solder a small wire between R7051 pads today and check F7040. My hot air station is pretty barbaric, which I think may be causing some troubles. I'll get better equip soon.

What do you think of low voltage on R7001, what would be cause? 3v42 is present, all other power rails are present.

 

[2informaticos]

"(C7000/1) are not showing any ohms in resistance, and 0.00 in diode mode"
Can you post exact values in ohm scale and diode mode too?

 

[Hardwood T/S]

Also, on another board, R5280 pin 2 is low, around 1.6/1.7v instead of 3.4v, with no detection on the battery but everything else works. I've seen this a few times on these boards, common cause?

I soldered a wire between the pads for R7051, resistance is showing 12mohms between pins 17 and 18--should i resolder jumper wires between the resistors and those pins again? This board is confusing me. W/wire on R7501, F7040 shows rapidly decreasing resistance from 25mohms, slows down around 5mohms but still decreases, diode mode shows ~.420 volts but very very slowly increasing.

For C7001, I'm getting ~5.6ohms in resistance now, and C7000 about ~.8ohms. Diode mode shows 0.000 on each pin of C7000, and 0.003 on pin 1 of C7001.

 

[2informaticos]

Replace U7000 if you get short to ground on both pins of R7001.

"resistance is showing 12mohms between pins 17 and 18"
I already wrote the correct value; just pay more atention and don't waste time with other things...

 

[Hardwood T/S]

I know the correct value, I'm asking you if the correct action in this case would be resoldering jumper wires from R7501/2 to pins 17/18 instead of replacing U7000 or not, since I'm already jumping the pads of R7051 and previous attempts at the jumper wires yielded varying results.

and what about R5280? Not sure what could be creating a lower voltage on that SDA line, U7000? I can only assume, since U7000 is the first start on that line when battery isn't plugged in--pin 10 creates that dataline from G3H, yes?

 

[Hardwood T/S]

Replaced U7000 on the board with C7001 showing 0.000 on diode mode, now R7051 is showing wacky resistance--why does this guy fail so often?

 

[2informaticos]

Remove C70001, if still shows short.
Eliminate any short to ground before power up the board again.
And "make" resistance between 17/18 to have correct value!

 

[Hardwood T/S]

Will do! I'll probably try tomorrow, I'm soldering on my knees and in the dark here... getting some microscope lenses in tomorrow hopefully!
C7001 doesn't show short anymore, but the board isn't powering on and dim light on charger still... I'll check the power rails and some more ISL stuff to see what the issue could possibly be.

 

[Hardwood T/S]

Turns out my C7001 board had a misaligned jack connector... well, that one is fixed. I hate people who strip screws.

Alright, so that leaves my 17/18 board in the dark still, I honestly think I probably toasted it somewhere and I'm running out of known good ISL chips to use.

The other board I have that won't detect the battery, the one with low voltage on R5280 SDA line, around ~1.5v, you think replace ISL? Not sure where else to go from there, maybe replace D6950, tho the SCL line is fine... which is why I think ISL.

 

[Hardwood T/S]

Replaced ISL from donor board, no fan spin--pretty sure that ISL was bad... put back og ISL, got fan spin again. I really need more good 6259AHRTZ chips.
They're compatible between all 13" unibody models if the chip starts with 6259AHRTZ, correct? The stuff afterwards doesn't matter?

Anyway, back to square one. :[] []: