MacBook Pro a2442 - 20v restart every second

[obelix24]

I just got one MacBook Pro a2442 (M1 Pro) there was small liquid damage.

Board starts normally at 5V and then jumps to 20V but after few seconds USB-C meter restarts (on every port) - sometime 1 or 2 times it's start and doesn't restart.

There was hotspot on U5200 (but it looks like on PIN1 (CHGR_GATE_Q4) but I've check in diode mode and it was ok reading (but I've replaced it anyway but still the same ...

There was corrosion around UR660, UR630, UR600 but I've checked all CAPS and it looks ok ...

When I connect USB-C meter and check with thermal camera I can see light on RC282, RC283 (only blink when USB-C meter turn off and again on) but diode mode is ok for both sides.

Also U9500 and U9400 get title warm - but I guess it's ok as there was no corrosion around this

PPBUS_AON on F5200 and F5201 is 12,32V and then it goes to 0V (when USB-C meter turns off and on

What should I check to find what is causing this ? I guess it must be short somewhere but I'm unable to find it ...

[mon2]

What is the resistance to ground of PPBUS_AON?

Flux and remove the 2 fuses on PPBUS_AON and test again to see if the power cycling stops.

[obelix24]

Resistance on fuses to ground is 5M ohms
when i've removed both fuses usb-c meter stays on but only on 5V.

[obelix24]

should i try to inject voltage to PPBUS_AON and monitor what is getting hot ? (12V and monitor board or this isn't good idea ? i'm just thinking because PPBUS_AON should handle 12-13V and as there is no short probably it should be visible with injection or no ?)

[mon2]

No. Never inject such a high voltage because this high voltage may be directly leaking onto a lower power rail. This PPBUS_AON voltage is used as the highest voltage to power many buck regulators like the CPU rail. If the high side mosfet on any of these downstream power rails is defective and leaking, then that leg of the power rail will be killed by the high injection voltage. Very possible it is already dead but proceed always with caution.

You will not see anything heat up because the resistance to ground on the removed fuses is too high. Only if the resistance was very low, then you can see the part(s) heating up under a thermal camera or IPA will evaporate faster due to the generated heat.

At this time, suspecting the CD321x controllers. Leave the fuses OFF the board. Confirm the checklist created by piernov and others on how 20V is selected by these PD controllers.

Confirm the presence of each LDO rail on each CD321x controller.

https://logi.wiki/index.php/CD3215_Bootup_Sequence

[obelix24]

but when fuses were there i got 20V isn't this proof that all CD3217xx chips are ok ? I was in this that when i get 20V PD controllers (CD321xxx IC are ok).

also other question i should get 20V even if fuses are removed ?

[obelix24]

I've checked
PP3V3_UPC0_LDO - 3,4V
PP3V3_UPC1_LDO - 3,4V
PP3V3_UPC2_LDO - 3,4V
PP3V3_UPC5_LDO - 3,4V

PP1V5_UPC0_LDO - 1,5V
PP1V5_UPC1_LDO - 1,5V
PP1V5_UPC2_LDO - 1,5V
PP1V5_UPC5_LDO - 1,5V

so they all looks ok

[mon2]

There are more rails to check for the BMC interface @ 1v1

[obelix24]

I'm still not sure about the 2 resistors RC282 and RC283 they just blink (when fuse is there)
but both lines PP5V_S2_MAIN and P5VS2_DSCHG looks ok (I get no short) - see attached video

[obelix24]

I just noticed one strange thing. When i keep USB-C amp meter connected it keeps restarting (starts at 5v reach 20v approx 0,2a and restarts) but after 4 restarts (one time it was after 6, anothertime after 2 or 3) it stops and stay on 20v 0a (it-s not restarting ampmeter)

does this help to indentify what this can be causing ? Some mosfet that needs to heat up or something ?

[obelix24]

Also to add when it stays on 20v 0a PPBUS_AON is 12,04V