MacBook Air M2 15 — power cycling

[max737]

Hi everyone,

I'm working on a MacBook Air M2 15" (A2941, board 820-03160) that is stuck in a continuous power-cycling loop. My measurements don't point to a classic short to ground, so I'd appreciate a second opinion on where to look next.

History

• Worked normally, then suddenly went black and would not power on again.
• No liquid damage — board is visually clean, no corrosion, indicators not tripped.
• A shop previously diagnosed it as a dead logic board.

Power-cycling behavior

• On both MagSafe and USB-C, the input voltage ramps from 5V up to ~19V and then repeats non-stop.
• Probing PPBUS_AON during the cycle: it builds up to ~12.3V, collapses to 0V, then builds up again, over and over. All rails rise and fall together in the same rhythm.
• The cycling is fast and continuous (several times per second).

Why I don't think it's a short

• I mapped the whole power section in diode mode to ground (red probe on GND). All values are within normal range — nothing reads 0.000 or noticeably lower than its neighbors.
• Every rail actually reaches its target voltage before the whole board resets together.

Based on this, it looks to me like a power-on sequence that keeps aborting and restarting (a boot loop) rather than a rail being pulled down by a short. I just can't tell what is aborting the sequence.

DFU

• Keyboard flex is connected and the power button works, but the board does not appear in Apple Configurator. I suspect the fast cycling prevents the DFU handshake from completing.
• No keyboard backlight and no click response from the trackpad.

Question Given this behavior, what would you check next to find what's aborting the power-on sequence? And is there a known force-DFU test point on the A2941 / 820-03160 that would let me force DFU independently of the cycling?

Thanks in advance — happy to provide additional measurements or photos.

 

[2informaticos]

First of all, welcome to the forum!

Did you test the board alone on the desk?

If PPBUS_AON still cycles so fast with just charger, then you can't access DFU mode.

 

[max737]

Thanks for the welcome!

Yes, bare board on the desk, battery disconnected.

With a 5V-only source (USB-A charger, DCP): completely stable. 5.2V / ~0.11A input, PPBUS_AON steady at 12.0V, runs indefinitely, no cycling.

With any PD charger (20W / 40W / 70W tested, via MagSafe and directly on a TB port): PD negotiation succeeds (9V / 15V / 20V), but the moment the higher voltage kicks in, input current drops to ~4mA, then VBUS goes to 0V (hard reset) and the loop repeats every ~2-3s.

PPBUS_AON never collapses during the loop — filmed both meters simultaneously: VBUS dies first, PPBUS stays at 12.3V. No short on PPBUS (diode mode ~0.52 across the rail, caps hold fine).