(347) 552-2258 [email protected]

A2779 (820-02841-A) Gets 20V, 5V->20V handshake loops and does not boot. Can enter DFU. Liquid Damaged.

P

PaulPCGuy

New member
Hello, this MacBook is acting strangely. USB-C and Magsafe are able to get 20V for about a second, before the board crashes and it renegotiates. This is done on an infinite loop. i found evidence of liquid damage, namely some corroded components and some trails on the board. Notably, the SPMU (U7700) had a sticky looking circle on top of it, might be related might not. 3 of the inductors around it had dark brown spots on top of them. I replaced them and the corroded components but got nowhere. PPBUS_AON is not shorted to ground and I can't seem to find shorts anywhere else. I did notice a brown spot on the WiFi IC, but it came off with alcohol so I'm inclined to believe that was from the physical spill, and not from the WiFi IC shorting out and getting hot. There is otherwise no corrosion in that area as well. NAND's are also good, and under the backplate is completely clean. All replacements came from a donor board. Here's what I've replaced:

RP850 (PPLUXE_INPUT_R_V csense) - corrosion
QP800 (Connected to RP850) - corrosion
R6030 (20K sitting between P3V8AON_ISNS1_P and U6070) - Original corroded off the board completely, it came right off with no heat. Trace repair was not needed.
Reflowed U6070, not replaced - Was potentially lightly corroded, though hard to tell
L7810, L7842, L7852 - See aforementioned dark spots above


Starting at 9V, all the way up to 12.8V (where I stopped), the board will constantly draw 0.005A. When prompting boot, it will sputter about before crashing to 0A, then 0.005A, then repeating the cycle. I was able to force it into DFU with MacVDMTool while it is attempting to boot. It does successfully boot into DFU, and Apple Configurator is able to see it. Attempting to restore will yield the same power loop as before, and eventually it will stop with the error "Gave up waiting for device to go from DFU state to DFU state". Out of curiosity, I tried forcing DFU again once the power cycling starts, and that just caused AC2 to hang indefinitely (tested for 30 minutes, progress bar did not move at all).
 
P

PaulPCGuy

New member
Update 2: I can't find the short. It's a 0 ohm short, and no amount of amperage (6.1A at 1.2V) produces heat on a thermal camera. I've pulled off just about every IC and MOSFET that's either directly connected to PP3V8_AON, or indirectly past a csense or an inductor. This includes the WiFi IC. The only IC's I haven't pulled off are the 6 SPKRAMP IC's. This is because there was never any corrosion or signs of heat near them, and I've never seen them fail before.
 
2informaticos

2informaticos

Administrator
Staff member
"(6.1A at 1.2V)"
This means 7W aprox, something should go really hot.
Be sure to inject voltage in the correct points.
Use powerful cables; you want to heat something on the board, not the cables.

BTW, set PSU amp limit to max first.
 
P

PaulPCGuy

New member
I found the short, and fixed it. I reassembled the board, but the short came back. Narrowed it down to UR500 (audio codec?), and removed it. The only three IC's not present at this point are that, UH904 (HDMI MUX?) because it shorted out and I assume it isn't needed to boot, and the WiFi IC as I don't have a stencil for it at the moment. At this point, I attempted to power the board, and now i don't even get 20V at all. It can enter DFU still, restore doesn't work.

PPBUS_AON is 12.33V
PP3V8 is 3.8V
PP1V8_AON is 1.8V
PP1V8_S2, PP1V8_AWAKE, PP1V8_S1_CLVR_VDD jitter between 0.3V-1.8V. No short. Cold voltage injection yields nothing.

I tried removing UN400 but the behavior did not change. So either I never fixed the original issue, or something I have removed is necessary for boot (excluding UN400 I know that one is needed) I will replace everything (except for WiFi) tomorrow. But in the meantime, is there anything specific I should look at? Keep in mind that I have reballed the SPMU (U7700).
 
2informaticos

2informaticos

Administrator
Staff member
If WiFi chip is dead, can't be replaced; it is married with CPU.
In such case, think if is worth to continue with this board.

Possibly problems appeared reballing U7700, or heat applied on the board affected something else.
 
P

PaulPCGuy

New member
There are AM wifi IC's, just need to be bound to the CPU.

So you're saying it's more likely to be problems with generating those rails than some kind of short?
 
P

PaulPCGuy

New member
It was PP3V8_AON. Specifically, the subrails for WiFi and for HDMI. It's unclear if WiFi was caused by my, but HDMI was definitely from the prior damage.

I will note that U7700 had a sticky circle on top of it near one of the edges when I first got the board. Looking at the boardview, this edge holds PP1V8_S2 and PP1V_AWAKE.


I removed U7700 again and tested the behavior on PP1V8_S2, PP1V8_AWAKE, and PP1V8_S1_CLVR_VDDH. Since S2 and AWAKE are generated by it, they were expectedly low. PP1V8_S1_CLVR_VDDH was also low though. Previously, if I tried to inject voltage into any of these 3 rails (specifically while attempting to boot), my PSU would show the same jittering behavior. Now that U7700 is not present, none of these 3 rails exhibit the jitter. Though I will also note that they act similarly to if I tried to power them without USB power, so this isn't really a conclusive test.


I'm thinking about replacing U7700, is that a reasonable next step?
 
P

PaulPCGuy

New member
3V8 stays at a stable 3.8V. So if there was an overvoltage spike, it quickly burned out whatever made the connection. I'll check what IC's and FET's have both 3.8V and 1.8V connections. I think I'm also going to replace U7700 anyways just to be safe. It'll be a little bit before I work on this again, as I have cleared my bench for other MacBooks to work on.
 
2informaticos

2informaticos

Administrator
Staff member
Think if is worth to spend more time on it.
Will customer accept the machine without embedded WiFi?
Needs to connect external adapter...
 
P

PaulPCGuy

New member
I don't know for sure that the WiFi IC is dead. And if it is dead, I will use the Luban programmer as explained on repair.wiki to replace it.

The customer is myself, I do refurbishing. So, yes someone will buy it if I can't restore WiFi, just at a steep discount.
 
P

PaulPCGuy

New member
I believe I have found the issue. Or at least, the current one. The original SPMU (U7700) did not have a hotspot, but would not generate 1V8S2/1V8AWAKE properly. Presumably other rails were also affected, but this is the only one I actually looked at.

Much earlier in the diagnostic, I tried swapping it with a donor from a liquid damaged A2442. When that one came back with a very hot hotspot, I assumed it was just dead from liquid damage. Now my replacement from Sentrix has come in and it has the same behavior. I pulled it off and started probing. I found that what I assume is the main power input for the SPMU is shorted. Tomorrow I will figure out what on VSS_ANA_SPMU is shorted, and hopefully with some luck this MacBook will be booting tomorrow.
 
2informaticos

2informaticos

Administrator
Staff member
Ooops, VSS is a ground point!
Looking at page 39 you can see a lot of VSS pins; where they go???

In general, in electronics, VDD and VCC are labels for positive power supply designation.
VSS and VEE are used for ground, or negative power supply.
 
P

PaulPCGuy

New member
Thanks for that. I found that (at least) PPVDD2H0_S2SW and PPVDDQ0_S1 are shorted. I pulled off 6 caps which were suspect, but it didn't go away. I pulled off the CPU, and it still hasn't gone away. At least now I can safely drive more voltage into it but still this is getting insane lol
 
P

PaulPCGuy

New member
Already beat you to that, we're good in the short department. All that's left for me right now is to reassemble the board minus wifi and cpu, wait for stencils to come in, and pray I didn't kill wifi or cpu with heat.
 
You must log in or register to reply here.
Top