A1708 820-00840 liquid damage, no power

Doubleu82

New member
Hi All,

let me start with introducing myself, I'm an electronics engineer working in software now for ASML. I started repairing macbooks (actually Louis videos triggered me ;-) ) from a hobby perspective now and then but it has grown since to 1 or 2 a week. My repair side carrier made it possible to invest in proper tooling (trinocular microscoop, JBC soldering, Quick hotair, HP multimeter, power supply with dual 30v5A, i5 hackintosh workstation from Aliexpress :-D


I'm working on a case (read rabbit-hole) for data recovery of an A1708 board 820-00840, i'll try to be as elaborate as possible (most from memory)

when I received the board it was doing absolutely nothing. I took some pictures of the corrosion and ultrasonically cleaned the board. this revealed some problem areas.
a lot of PP3V3_GH circuit was broken/gone, replaced some resistors and removed some caps. It works now (doesn't look very nice since i only had 0603 resistors laying around)

Screenshot 2020-08-01 at 13.05.56.pngScreenshot 2020-08-01 at 13.08.53.png
the gone pads are TPAD testpoints/via's, didn't bother fixing those yet

PP3V3_GH, repaired the circuit and now 3,4v present (when powered via F7000)
still no power via USB-C (stays at 5v 0mA, other port 5v 50mA and getting hot)
first replaced one CD3215 (the one getting hot) , still not equal mA, replaced second CD3215, now both equal. PPDCIN_3GH present at Q7030 (usb 5v 10mA)
I did not trust the U7000 (ISL) (weird impedance at Aux_det) so replaced it
now Aux det is 1,2v and R7015, R7016 measure ok.
now I get 5v at charger and 40mA cycling at both USB ports, but ISL still creates no PPBUS (12v3)

when I supply power to the battery terminal, (7v5) it takes 15mA and I measure this voltage also at F7000 (tried 10v, no change), PPBUS is than equal to the supplied voltage

SMS_RESET_R_L = 3,4v at R7072
PM_EN_P3V3_GH = 5v at R6999
SMC_AUX_OK = 3,4v at R7071
PP3V3_3GH_UPC = 3v3 at C3100 and C3200

Any clues where to look further?
 

2informaticos

Administrator
Staff member
First of all, welcome to the forum!

3V3_G3H is the only corroded area on the board?

STOP applying external voltage at F7000, or battery terminals!!!

Are you sure the new CD3215 chips are good?
Many fake/remarked chinese crap out.
Check for 1V1 LDO voltage on the powered USB-C port.
Do you get BUF_SMC_RESET_L too?

"power supply with dual 30v5A"
Forget it and get a simple 24V/20A, or arm one with DPS5020.
Many times mentioned WHY in the forum...
 

Doubleu82

New member
Hi,

thanks for the fast reply,

The ISL area was a bit corroded, after cleaning with IPA and a brush it looked ok. but replaced the ISL (ebay source) after measuring some wierd levels

I ordered two batches of CD3215, first 5pcs from ali that arrived after 2months dangling in a plastic bag wrapped in bubble foil (awful), and a set from ebay, nicely packed in a cut reel (15eu a piece). What would be a trustworthy source (available in EU) for the cd3215.

I'll have a look into the signals mentioned and get back to it
 

Doubleu82

New member
First of all, welcome to the forum!

3V3_G3H is the only corroded area on the board?

STOP applying external voltage at F7000, or battery terminals!!!

Are you sure the new CD3215 chips are good?
Many fake/remarked chinese crap out.
Check for 1V1 LDO voltage on the powered USB-C port.
Do you get BUF_SMC_RESET_L too?

"power supply with dual 30v5A"
Forget it and get a simple 24V/20A, or arm one with DPS5020.
Many times mentioned WHY in the forum...


I checked the 1v1 LDO both are 0v
BUF_SMC_RESET_L = 3v42

I swapped out the aliexpress CD3215 for one I bought on ebay (the ones which where packed in a cut reel) I see also difference in the chips caps, the one from Ali are shiny-ish and the eBay ones have are more dull finish but all of them are reballed. the og chips are shiny (like the Ali batch)
 

2informaticos

Administrator
Staff member
Quote the text you want to make a refference.
I don't see the reason to quote an entire post...

You need to get 1V1 LDO on all CD3215 chips.
I'm still trying to find a serious provider, is not easy.
Is better to buy the "wings" board, like this
being original for sure.
Still exists the risk that initial problem was a CD3215 chip.
If you buy many, probably get 80-90% good ones...
 

Doubleu82

New member
When 1v1 is not present, is that a cd3215 issue ( broken) or could that be an issue elsewhere? Are there any additional checks i could do to verify a bad cd3215.

I use 425degrC and high air to (de)solder them could this be of any risk (i also do eg tristar with these settings and i'm successful with that)

when salvaging cd3215's I guess I need to reball, any suggestion for a stencil?
 

2informaticos

Administrator
Staff member


Use direct heat stencil, with solder paste.
Practice on scrap board, until get confiance.
1V1 LDO should appear on the powered USB port only.
 

Doubleu82

New member
i'll look into that, thanks for the advise.

Like mentioned earlier, this repair is for data recovery, is there another way of retrieving data from the ssd (encrypted), This model has a replacable SSD in a weird form-factor
 
Top