820-00164 wont get past S5

[Atomrepair]

The board where I fixed the magsafe light earlier came back no longer starting up. Magsafe light still works. It has booted up once since I got it back, but mostly refuses to. It came in originally with corrosion, which cleaned up good and left the machine functional, but its not so hot now

PPBUS_G3H	8,6V
SMC_RESET_L	3,39V
SMC_BCACOK	3,39V
SMC_ADAPTER_EN	3.43V
PP5V_S5	5V
PP3V3_S5	3,34V
S5_PWRGD	3,4V
PP3V3_S3	0V
PP5V_S4RS3	0V
PP3V3_S3_EN	0V

 

[dukefawks]

The usual, U1900/U1950/U6100 and termination resistors, JTAG.

 

[Atomrepair]

U6100 looked fine, but in the middle of measuring its pins, it decided to start kind of working. It would power on, fan would gradually increase speed and eventually power off, and then on again.
I replaced U6100.
Now, it powers on normally, but it never finishes loading OS X. It gets through a slow progress bar (takes about 10 minutes), and reboots when finished to start that process anew.
I've tried booting from an usb installer stick, which behaves the same way. I know the drive is working fine, or at least it was three days ago.
I dont have time to work on this one any further today, but do I go back to measuring rails or would replacing U1900 and U1950 be the next logical step?

 

[dukefawks]

Of course you cannot just replace U6100 without fucking up the ME, put back original. I hope you still have the original as there is nothing wrong with it for 99% sure. Something else around U6100 got heated up so it works again, U1900, JTAG and the termination resistors/traces to U6100 are first suspects.

 

[Atomrepair]

U6100 is a CMOS chip huh.. fuck me, I thought the way those pins are named reminded me of something. Thought it would be a SOIC8 like every windows laptop has. Much shame for me. Will report back later.

 

[Atomrepair]

Alright, I've placed U6100 back and it boots up normally again.
I've measured all surrounding resistors, they all perfectly match their values as in the schematic. Traces around U6100 look fine.

JTAG and U1900 and surroundings also look perfect to me. I'll remove the JTAG and replace U1900 just in case.

see pics:

 

[Atomrepair]

JTAG is removed.
U1900 is replaced.
I verified that every pin has continuity to where its supposed to go on this chip and U6100. Macbook worked for about 15 minutes, then shut off. Its currently spinning up the fan, no gong,l no image. fan stays spinning for ~8 seconds, then stops. rinse, repeat.

U1950 and surrounds look just as clean as U1900. Should I replace that too, just in case?

 

[dukefawks]

Could be anything. It is corrosion somewhere. If Vcore is present then U1950 will be fine. Crap under SMC?

 

[Atomrepair]

I've first heated the SMC to pre-reflow temp, no change. I've reflowed, didn't help either. Today, I've replaced every IC that had corrosion prior to putting it through the UC, no change so far. I'll replace the SMC and report back.

 

[Atomrepair]

Replaced SMC - no change. Still getting 8 second fan spin 'sometimes'. Mose of the time nothing but a green and then orange light. I've checked the BIOS chip with my scope. While the fan is off, it is outputting a flat 3.3V on pin 1 through 8, save for GND of course. No data to be seen.

 

[Atomrepair]

I happened to come across this video: https://www.youtube.com/watch?v=uvRaC3VUYR0 - will see if SPI_MLBROM_CS_L makes it all the way to SPI_CS0_R_L tomorrow.

 

[Atomrepair]

So much for tomorrow. ..

SPU_CS0_R_L is fine at 3.34V.

Board is driving me nuts. Sometimes its no fan spin at all, sometimes theres an 8 second fan spin and then nothing, sometimes it will do the 8 seconds spin and then continuous spin afterwards, but I get no image.

One oddity I noticed - when hooking up my scope and attaching the ground lead, the magsafe light switches off, then goes on again. Same happens when I touch the board's ground with the ground lead on my multimeter. This is with nothing but the magsafe charger touching the board, and fan attached. I dont quite know what to do with that information..

 

[smiba]

Is it a pressure thing? Does it also happen if you poke at it with something like a wooden toothpick?
No idea what it means if its not a pressure issue, odd. Posting this as a comment by the way since I don't want to derail the thread

 

[Atomrepair]

Just did a little further testing to make sure I'm not being crazy or if its a random occurance. Its not. It will happen if I touch it with a screwdriver (though it seems to be harder with those, gotta touch it a few times), tweezers, MM probe, no matter. something metal. If its something with a wire, it happens more readily. It doesnt happen if I use something non conductive, like my finger or a piece of wood. I tried on the shielding for U2800, usb port housing, SSD connector housing, it doesnt matter as long as its ground. This is weird.

I made a lil vid to show whats going on:
https://youtu.be/rS0vkfBxhqo

 

[smiba]

Only thing I can think about is that by doing that you're introducing a piece of metal to the ground traces that has a different potential so you will be creating a microscopic spike or drop in voltage on ground. Because you're holding the metal piece or by using a long wire you might also be introducing noise. I don't understand how this is affecting the board in this way though, since ground should be very low impedance and thus not notice this at all. Tried a different charger and/or IO board? More interested in this then I probably should be haha

 

[Atomrepair]

I tried with a different I/O board, same thing. I only have one Magsafe 2 charger. Not really willing to cough up €90 for another one. Besides, this one a few months old and hasn't seen much use yet. I'm going to look over everything I've done another time, make sure I haven't got an IC placed crooked or something.