(347) 552-2258 [email protected]

820-00138-A Shuts itself down before booting to system.

Beto_MFC

Beto_MFC

New member
The machine had coffee all over it's keyboard but it dind't reach the board itself. Same behaviour with keyboard, trackpad and ssd removed. The problem with this board is that there's no boardview available. But as the board dind't had liquid on it, maybe the damage was caused by a short on the keyboard. Where do I start measuring?
 
D

dukefawks

Administrator
Usually happens when there is corrosion on the trackpad/cable/connector. Inspect all for corrosion and try known good trackpad+cable.
 
Beto_MFC

Beto_MFC

New member
No signal of corrosion and both keyboard and trackpad are disconnected. Board show no signs of damage at firs glance. Will look more carefully.
 
Beto_MFC

Beto_MFC

New member
PPV5_S4 is power cycling. Now after further inspection it looks like C7372 area might have liquid contact. This cap is on the S5_HS_COMPUTING line.
 
Beto_MFC

Beto_MFC

New member
Maybe i wasn't clear. Working on the board itself, nothing else plugged. It turns on for some seconds and shoots itse;f off.
 
D

dukefawks

Administrator
It does power on with image for a while?
Power cycling on 5V_S4 can be anything and I mean anything. Inspect for corrosion and damage.
 
Beto_MFC

Beto_MFC

New member
dukefawks said:
It does power on with image for a while?
Power cycling on 5V_S4 can be anything and I mean anything. Inspect for corrosion and damage.
Click to expand...

Yes, tryes to boot and before entering on MacOS shuts off. Same behaviour when working only with the board. No signal of corrosion anywhere. Not having the boardview makes it harder, i'm using 820-3662 brd but is not 100% similar.
 
Beto_MFC

Beto_MFC

New member
I think one of the phases of the cpu buck converter got spilled cofee. But it was very hard to pinpoint that. Will order a new buck converter and replace. Seems like that may be the issue. One of the phases a little funky would make the machine turn off.
 
D

dukefawks

Administrator
It would not even be a bad chip it would be corrosion under it or a corroded trace. Inspect CPU Vcore regulator and FETs carefully.
 
Beto_MFC

Beto_MFC

New member
Changed CPU buck regulator which has coffe contact but no dice. Seems that the fin stack sensor got coffee too. Can this sensor reporting wrong value causing the issue?
 
D

dukefawks

Administrator
In a very remote chance it could be possible. Just pulling it completely would solve a shut down issue. Machine will run slow though.
 
Beto_MFC

Beto_MFC

New member
No change removing finstack sensor. Even if no apparent liquid contact over U7200 gonna have a check on it.
 
smiba

smiba

New member
Like duke said could be a lot of things

Keep an eye on the important voltages (S5 and G3H voltages) when the system fails and powers off. Might be something that goes below its supposed range.

I also have seen this issue before on a damaged ISL, it might look good and the system even tries to boot. But fails randomly or at higher load. You most likely see G3H drop for a split second if this happens but no idea if most multimeters can catch this. This is very rare though, just guessing.
Does a diagnostic test give out any details or does it fully pass?

Try to boot the system in verbose mode, maybe you can catch a certain driver initializing before it fails

Also try to inspect the S4 loop a little more (Does it fully loop into S0 or just S5-->S4-->S5?)

Otherwise no idea
 
Last edited:
D

dukefawks

Administrator
There is no magic solution for this. You saw where the corrosion was and have to asses what trace/resistor/chip could be the problem. Random shutdowns like this are impossible to trouble shoot over the internet and even with hands on it is very difficult.
Shut downs on boot when entering OSX is usually something in CPU power supply, but as I said it could be 100 other things.
 
Beto_MFC

Beto_MFC

New member
Will mos def check all that again. Just reminding that this board doesn't have the boardview and i'm using one from a similar board. Will post here as soon as proceed with further testing. Thanks for the input!
 
Beto_MFC

Beto_MFC

New member
Forgot to mention that after found abnormal 5v behaviour replaced 5.5/3/3v supply but still no dice. Something ahead of it is causing the problems.
 
You must log in or register to reply here.
Top