A1708, 820-00875 20V 0.0185A

[Theking3737]

Hello everyone,

I have a A1708, 820-00875, in front of me that doesn't turn on and my ammeter says 20V and 0.0185A. The person I bought it from told me that it turns on occasionally but crashes and after that it's hard to get it to power on again. So far I couldn't get it to power on at all. I opened it up and removed the logic board. I measured 0.015V at PPBUS_G3H with a resistance to ground of 2.2 Ω. I then measured the resistance between PPBUS_G3H and L7420, L7410 and L7430 and those are basically all shorted at 0.4 Ω. The resistance for all the other inductors seems fine.

I suspect U7420, U7430 or U7420 could be dead. I think lifting L7420, L7410 and L7430 one by one an checking if the short is gone is the best course of action, am I right?

I do have a donor board that still has a bunch of these chips on them and I hope they still work. Although it being shorted to PPBUS_G3H kinda worries me.

[mon2]

Remove all power. Flux and remove the fuse @ F7000. Measure the resistance to ground on each side of the removed fuse using the PCB pads.

Which side is with the low resistance to ground of 2R2? If the downstream / consumer side, then power up again and check the voltage to ground of PPBUS_G3H as created by the ISL9239 controller. Is this voltage stable?

With no power, measure the resistance from the downstream side of the F7000 fuse to any inductor on the board. Keep testing each inductor for a low resistance path from PPBUS_G3H (F7000, pin 2). If you find a low resistance value then that high side mosfet is most likely leaky and may have caused damage to that power rail. Post all suspects. Hoping it is not the CPU rail.

[Theking3737]

Remove all power. Flux and remove the fuse @ F7000. Measure the resistance to ground on each side of the removed fuse using the PCB pads.

Which side is with the low resistance to ground of 2R2? If the downstream / consumer side, then power up again and check the voltage to ground of PPBUS_G3H as created by the ISL9239 controller. Is this voltage stable?

With no power, measure the resistance from the downstream side of the F7000 fuse to any inductor on the board. Keep testing each inductor for a low resistance path from PPBUS_G3H (F7000, pin 2). If you find a low resistance value then that high side mosfet is most likely leaky and may have caused damage to that power rail. Post all suspects. Hoping it is not the CPU rail.

I removed F7000 and noticed a very low resistance on L7420, L7410 and L7430 (all were at 0.4 Ω). The rest seemed fine. I then removed L7430 and L7410 (read about it on another post). After removing L7410 I noticed that the short was gone. I thought this could mean that U7410 was dead. So I decided to replace U7410 with a chip that I got from a donor board. After replacing it the short was gone!

I plugged in the ammeter and it showed that it was pulling a few hundred milliamps. PPBUS_G3H was also back at 13V. I decided to put it back into the case and test it out. I pressed the power button and the fan started to spin, the touchpad was clicking again and the ammeter showed more than 2.6A at some point. But sadly no display, also not on an external display using a dongle. I even tried a different case with a different display, but still the same result. The light on the caps lock button also doesn't seem to work and I never heard the MacBook chime.

[mon2]

Measure the voltage to ground of ALL_SYS_PWRGD.

[Theking3737]

Measure the voltage to ground of ALL_SYS_PWRGD.

I measured ALL_SYS_PWRGD (on R7806) at 3.19V. When measuring, the battery was not connected and I saw fan spin.

[mon2]

A logic '1' at this signal implies that all power rails are OK. Is the fan running at all times or only briefly? Is the fan spin 'normal speed' or crazy high to cool down burning cookies speed? If crazy high, then some sensor detection has failed which is often from the sensor(s) inside of the trackpad.

[Theking3737]

A logic '1' at this signal implies that all power rails are OK. Is the fan running at all times or only briefly? Is the fan spin 'normal speed' or crazy high to cool down burning cookies speed? If crazy high, then some sensor detection has failed which is often from the sensor(s) inside of the trackpad.

The fan spins briefly after pressing the button. If it spins, it spins quite slowly, I'd say it's inaudible. I also tried putting the board in a different case with a different screen, trackpad, keyboard, etc. It made no difference.

[mon2]

Do wait on other comments but suspicion is that the first post mentioned a low resistance to ground on PPBUS_G3H. Your investigation led you to the multiphase power supply linked to the GPU via U5410 / L7410. By replacing U5410, the short on PPBUS_G3H was removed.

This means that the high side of U5410 (internal mosfet) was leaking so the original ~13v travelled onto the GPU power rail and often this is not good news as this will kill the GPU. A dead GPU will result in no video.