MacBook Air 2018 A1932 820-01521 stuck at 5V, but succesfully does the DFU and boots into OS installer

[mitchw]

I have a 820-01521 that is stuck at 5V, and drawing just 230mA.
I did some troubleshooting, and managed to charge the battery up (was at 0%) using a different logic board.
Also I ran DFU and it did it successfully. Had to use a PMU_FORCE_DFU jumper to do it.
At this point the laptop started up and it successfully boots into the OS Install.

The charger is now at 20V, but drawing 0mA, so the laptop is essentially running off the battery. It doesn't detect the charger...

I think there could be a problem with one of the CD3215s, but all the LDOs are there...

[mon2]

MacBook Air A1932 Not turning on, pulling 0.03-0.06A at 20V repair - Repair Wiki

[mitchw]

I'll check it out, thanks.

Anyways, if I have battery connected, it gets to 20V, but stays at 0mA.
With battery disconnected, it only gets 5V, but draws 230-240mA.

[mon2]

Review the attached mosfets that are between the CD3215 and the PPBUS_G3H rail. Remove all power. Meter in resistance mode.

Check the resistance of each mosfet -> check source & drain; source & gate; gate & drain for each of the dual mosfets. That is, you have 2 mosfets in a single package so each is a suspect.

[mitchw]

Both measure the same.

Found next (possible problems):
- PPBUS_G3H is 12.3V (should be 12.6V I think)
- PP3V3_G3H is 3.3V
- PP5V_G3S is 0V
- PP1V8_G3S is 0V
No shorts on those two, which makes sense as the laptop worked fine on battery when booted to OS install after DFU.
On battery of course there are all voltages present...

I also did another quick check of the LDOs of the CD3215s.
All of them are fine, except there is a bit strange behavior of the PP1V1_UPC_XA_LDO_BMC and PP1V1_UPC_XB_LDO_BMC voltages:

If the charger is connected to the U3200 USB-C ports (the closest to the screen)
- PP1V1_UPC_XA_LDO_BMC is 0V
- PP1V1_UPC_XB_LDO_BMC is 1.1V

But if I connect the charger to the USB-C port that is connected to U3100 (the closest to the trakcpad), then I get the following:
- PP1V1_UPC_XA_LDO_BMC is 1.1V
- PP1V1_UPC_XB_LDO_BMC is 0.98V

Is that correct? If not, then which CD3125 is misbehaving? I sadly don't have a known good board to verify this

[mon2]

When you measured the bmc_ldo rails, did you rotate the usb type C connector?

review this guide on how the cd3215 devices work to select the 20v power delivery contract.

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

[mitchw]

I did rotate them

[mon2]

PP1V1_UPC_XA_LDO_BMC should also be 1v1 when used as the active port.

check each ldo in diode mode. Remove all power. Meter in diode mode. Red meter probe, yes red to ground. Black meter probe to the ldo rail under test.

post each measurement of each ldo rail on each cd3215.

[mitchw]

Here you go:

PP1V1_UPC_XA_LDO_BMC: 0.550V
PP1V1_UPC_XB_LDO_BMC: 0.550V

PP1V8_UPC_XA_LDOD: 0.512V
PP1V8_UPC_XB_LDOD: 0.511V

PP1V8_UPC_XA_LDOA: 0.554V
PP1V8_UPC_XB_LDOA: 0.553V

PP3V3_UPC_XA_LDO: 0.525V
PP3V3_UPC_XB_LDO 0.525V

Both read exactly the same...

... on the logi.wiki it says: PP1V1_UPC_XB_LDO_BMC (On active port after reading TBT ROM). So that means if I get a 0.98V on inactive port, and 1.1V on active, that the inactive is dead? Connected other way around, I do get 1.1V on active and 0V on inactive...

[mitchw]

Anyways, I replaced the U3200 (the most likely suspect) and the board is working now