820-3536-A with a half-shorted PP3V3_S0

Fanfwe

Member
Hi,
I've got this board and it's giving me trouble.
It will not start because PP3V3_S0 only has 1.8V
When I measure resistance to ground, I get about 130 Ohms, while a good board has approx 1KOhm
Thing is 130 Ohms is too much to be able to get the bad guy to get hot enough to find it.

I've tried to inject 3.3V anyway, and didn't indeed find anything especially hot.
But after I did that and let the voltage run for a few minutes, I was able to see the board was sometimes booting and sometimes not. I connected a speaker to be certain and I did hear a chime.
Now, the next day, the board doesn't boot anymore.

I'm not quite sure what to do from here. I can't really consider removing all the stuff on the 3v3_s0 rail until the 130Ohms to ground goes away, there's just too much stuff.
So any suggestion would be more than welcome :)
 

2informaticos

Administrator
Staff member
Any history available, liquid signs?
Can be a low value caused by accumulated crap under some BGA chip.
Also inspect all big ceramic caps on that line; remove the ones which look ugly/strange.
 

Fanfwe

Member
I don't really know the history. I got this board from a joblot off ebay. At first it was totally missing PP3V3_S0, with a dead short to ground. There was clear signs of liquid spill in the corner near the headphones jack, and especially, the microphone ribbon cable connector looked really nasty.
So I removed that connector and cleaned the pads. Then I got what I described above, which is no longer a dead short but still not quite a normal resistance for that rail.

I've already looked at all the caps on that rail, and none of them really look bad, unfortunately.

If there's crap under a BGA chip, I guess it could be the audio codec, since it's pretty close to the area where the dead short was.

But I noticed one thing, though. PP3V3_S0 is created from PP3V3_S5 through a mosfet switch. I do see 1.8V on the S0 rail, but the S5 rail has a clean 3.3V.
Shouldn't the short also drive the S5 rail down to 1.8V ? This makes me think I should maybe replace the mosfet for the S0 rail. Would you agree ?
 
Last edited:

2informaticos

Administrator
Staff member
Yes, is a good start.

BTW, do not quote an entire post visible few centimeters above.
I don't see any real reason...
 

Fanfwe

Member
Yeah sorry about the quote, I don't usually do that. Not sure why I did it here, my mistake.
So, I have removed the mosfet, and I now see about 1kohm to ground on PP3V3_S0, which looks much better.
I don't have a replacement mosfet at hand, so I'll have to get some from aliexpress and it will take a while, but it really looks like the problem did come from there.
 

2informaticos

Administrator
Staff member
If you deal with Apple stuff, buy scrap boards from Aliexpress.
You can get good price, if buy bulk and directly from the seller.
Then you get SMCs and lot of rare components...
 

Fanfwe

Member
Yeah, I do have a few scrap boards, but that mosfet is a raw die BGA. I haven't had much success trying to reball those in the past, unlike regular BGA packages. Not quite sure why.
 

Fanfwe

Member
Yeah, I tried what you said, and I was indeed able to solder it directly.
And replacing that mosfet fixed the board. PP3V3_S0 now runs at 3.3V and the board gives me a fanspin and a chime.
Thanks !
 
Top