A1932 - 820-01521-A Fluctuation

ameasere

New member
I have an A1932 820-01521-A board that came in for no power. When tested, it did power but no chime/backlight/fan spin.

It was stuck at 5V and the meter kept cycling - 5V 0.2A then resets. Inspection showed U3000 had a hole blown in the middle of it, so it was replaced. No more cycling issues!
It then got stuck at 5V 0.23A, no cycling nothing. Many attempts to see if DFU mode was available were unsuccessful - keyboard combination and jumper methods yielded nothing.

Now, the board does get to 20V but has some strange behavior. It will go to 5V 0.23A for a few seconds, and then gets to 20V 0A with a very very slight fan spin then stops. After a few more seconds, it will push to 20V 0.25-0.3A and ridiculously loud fan spin with no chime/backlight/DFU. I can do what appears to be a power cycle using the power button - even when holding the DFU combination - the meter goes back to 0 and starts the process again if I hold the power button long enough. All of the tests were done both with and without battery connected, and with all cables connected and disconnected to ensure I wasn't just being silly.

Power rails show normal voltages:
  • PP5v_G3S: 5v
  • PP3v3_G3S: 3.3v
  • PP1v8_G3S: 1.8v
  • PP3v_G3H: 3v
  • PP1v8_AWAKE: 1.8v, Present
Other advice shows to replace U7800 - but before ordering BGA chips and going through that process, I thought to check here in case there was something I had missed.

I am unsure if 20v has gone right through somewhere it shouldn't have, like the CPU or the T2 chip - the T2 does get warm but nothing out of the ordinary under thermal camera (34 degrees Celsius maximum), CPU is stone cold.

Both USB-C ports exhibit the behavior above, the CD3215 looks visually fine and has no obvious signs of damage (and no holes, like U3000 had).
 
Last edited:

2informaticos

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

If U3000 got burnt, PCH/CPU can be damaged; being directly connected.
Compare diode mode to ground on all big coils with good board.
 

ameasere

New member
PPBUS_G3H = 12.60V
PPBUS_HS_CPU = 12.61V
PPDCIN_G3H = 19.51V
PP5V_S5_LDO = 5.13V
PP3V3_G3H_RTC = 3.3V
PP3V3_G3H = 3.3V
PP3V3_G3SSW_SNS = 3.3V
PP3V3_G3S = 3.3V
PP3V3_TBT_X_S0 = 3.3V
PP2V7_NAND = 2.516V
PPVCC_S0_CPU = 0.72V
PP1V8_SLPS2R = 1.8V and 1.8 in diode mode, stable

Additional voltages as followed by another similar post (here) with similar symptoms on the same board revision.

Edit: thank you for the warm welcome! Apologies for the previous error in the post, PP3V3_G3SSW_SNS shows 0.0V until the 20V 0.3A is negotiated which then shows 3.3V (misidentified, corrected now)
 
Last edited:

ameasere

New member
First of all, welcome to the forum!

If U3000 got burnt, PCH/CPU can be damaged; being directly connected.
Compare diode mode to ground on all big coils with good board.
Understood, I did consider PCH/CPU may be damaged from this. I will see if I can get a known good board ordered in, we do not have any on hand.
 

2informaticos

Administrator
Staff member
"PPVCC_S0_CPU = 0.72V"
Is this voltage, or diode mode?

Do not write "V" after a diode mode reading.

I asked you about diode mode readings.
Seems that you've posted voltages and many coils not listed; PCH, RAM, iGPU, etc...

If CPU core really gets 0.72V, other coils should get voltage too.
Check for USB activity; or Caps Lock LED reaction.
 

ameasere

New member
"PPVCC_S0_CPU = 0.72V"
Is this voltage, or diode mode?

Do not write "V" after a diode mode reading.

I asked you about diode mode readings.
Seems that you've posted voltages and many coils not listed; PCH, RAM, iGPU, etc...

If CPU core really gets 0.72V, other coils should get voltage too.
Check for USB activity; or Caps Lock LED reaction.
Those are voltages. I posted that reply before yours appeared, since I was already measuring them when your reply was posted. I only saw your response after submitting the reply.

I need to order a known good board to check the diode readings you said about, which should arrive next week. Said board also has a known good CPU, so I can do a BGA transfer if the PCH/CPU is indeed damaged.
 

ameasere

New member
I used the OpenBoardData to check diode readings, all readings except one match: L7030, CHGR_PHASE1 (and 2, the other side of the coil). OBD says it should be a diode reading of 47, with 276R. When measuring, I get a reading of 0.46 with 147kΩ of resistance to ground.

No USB activity, no Caps Lock LED. Interestingly, no more fan spin episodes and the USB meter shows 19.5V jumping from 0.05A to 0.473A, then goes back to 0.05A and repeats the cycle.
 

2informaticos

Administrator
Staff member
Not all values from that web are good.
"If CPU core really gets 0.72V, other coils should get voltage too"
I mean L7600/60, L8100/2, L7290.
 
Last edited:

ameasere

New member
L7600/60, L7270 and L7702 are not found on boardview, not sure if it is a typo or something not present on this board.
  • L8100/2: 0.22 diode reading
This is the only one I can find, unless I am misreading what you said.
 

2informaticos

Administrator
Staff member
Sorry, I was looking at wrong schematic; too many files opened in my acrobat reader.

L8100/2 both 0.22?

Check also L7670/90 and L7290.
 

2informaticos

Administrator
Staff member
I suppose these are diode mode readings.
I asked about voltages; because you've said to get 0.72V at L7210/30.
 

2informaticos

Administrator
Staff member
Can you what is CPU core voltage on working board, if available?
I'm not sure if 0.72V is the correct level.

This board should enter DFU; try again.
 

ameasere

New member
I don't have a working board available yet, it has been ordered and will arrive in the next few days.

I have tried DFU with keyboard combination and a DFU jumper wire, it doesn't appear in Apple Configurator at all. I'm not sure if I am doing it wrong or not.

I tried a wire jumper from R8010 to C7844 (as I understand it, FORCE_DFU must be pulled up to a 1.8V source) but the device never appears in Configurator even if I then cut the wire after boot. Could it be due to the fluctuations in current draw? In the current state, the board draws 20V 0.3A and suddenly drops to 20V 0A, slowly going back up - cycling repeatedly.
 

ameasere

New member
Strange update: the laptop now turns on completely, drawing 1.8A and I get to the login screen. I have no clue what happened, but the laptop does seem rather unstable - sometimes I get a panic during boot, sometimes it is fine.
 
Top