Macbook Air 2020 (A2179 / 820-01958) PPBUS_G3H fluctuates from 0 to 12V

[gervaisc]

Hello, I am troubleshooting my daugther's MacBook Air 2020 (A2159), no liquid spills, not dropped. It was working well for the last 3 1/2 years and now refuse to power on.

Took of the board (nothing connected). Visual inspection with microscope looks ok. The USB-C voltage is 5.17V, current fluctuates from 0.05A to 0.120A. No sign of 20V.

I followed the (excellent) guide here: https://repair.wiki/w/MacBook_Air_A2...w_at_5V_repair

1- No shortsbetween PPBUS_G3H and ground. It shows more than 1Mohm
2- The PPDCIN_G3H shows a nice and steady 5.17V
3- PPBUS_G3H fluctuates a lot in a steady manner (I do not have a scope), multimeter shows from 4.5V to 12V when putting it in AC+DC mode. I confirmed with an analog multimeter, the needle goes from 4.5V to 12V, then drop to 4.5V then 12V.... It looks like "trying hard" to reach to 12V however not successful and returns to almost zero.
4- R7021 and R7022 show 23 and 47 ohms. I swear that they were reading around 1ohm yesterday...
5- R7061 and R7062 both show around 1 ohm.
6- The PP3V3_G3H_RTC and PPVIN_G3H_P3V3G3HRTC are also fluctuating - I think they are down the cascade line so I do not pay attention to this.
7- Other quick measurements I took around the U7000 ISL9240:
PPCHGR_VDDA = 5V
PPCHGR_VDDP = 5V
PPDCIN_G3H_CHGR_AMON = 5V
CHGR_CSI_N = 5V
CHGR_CSIR_N = 5V
CHGR_CSIR_P = 5V

After looking at many forums, I suspect the ISL9240 (U7000) is damaged. But before replacing it (I have never done this with such a small component...), could it be something else since it is "trying" hard to reach 12V?

Any other things to check?

many thanks!

[mon2]

Welcome. Before my Netflix binge...

1) Do not attempt the ISL9240 replacement till we can all agree that it is required. This is a very small WLCSP package and not an easy one to replace without damaging the pcb pads and/or nearby parts. This on top of soldering back a fresh one without shorts. Suggest some practice on a donor board only if this is even required. Another suggestion is to lean on a good neighborhood cell phone repair shop tech to perform the replacement. If they can replace tristar chips inside of iphones, they can handle this repair.

2)

R7021 and R7022 show 23 and 47 ohms. I swear that they were reading around 1ohm yesterday...

Stop. This is of concern. If you feel comfortable, consider to flux and carefully remove each off the board and measure their resistance again after the parts are cool and on your work table. They are tiny and very easy to lose. If you do lose them, they are available on Digikey / Mouser / Arrow but there are shipping fees associated with each. The parts are pennies but DK will charge +$8 for shipping (next business day) to Canada. Each resistor must match the resistance shown in the schematics.

3) You may wish to pickup a digital multimeter off Amazon. Most are inexpensive and can serve the purpose to confirm the resistance / voltage measurements. Be sure the meters are with fine needle like tips. Then short the probes to know the lowest (baseline) resistance of your meter. Then measure away.

4) Remove all power. Meter in resistance mode. Measure the resistance to ground of each and every inductor on the board. Each inductor is linked to a low switching buck regulator. If there is a lower than normal resistance, then that rail may be shorted. Start with the CPU voltage inductors.

5) Also, measure the resistance (without any power / no battery to the board) from the PPBUS_G3H junction (fuse is ok) AND then each inductor on the board. If you see a low resistance then this is of concern and means that the high side mosfet that mates with that inductor has leaked. With such a leak, the high voltage from PPBUS_G3H has travelled onto the low voltage rails.

Post the details from your hunting expedition.

[gervaisc]

Wow. thanks for the info! I'll have a look at #4 and #5 tomorrow morning with a fresh mind... For the multimeter, I am using two to make sure the reading are good, one is a Fluke 187 and the other one is an LCR meter. You have a good point, I used "clip on" needle adapter on my probes and I had way too much resistance on them when shorted. I used another set of probes with a bit bigger tip and measured 11 Ohm (R7021) and 91 Ohm (R7022). Why are you saying it is a concern? I know they are used to sense current, in my mind a resistor should not go up in resistance, shorting to zero but opposite is puzzling me. Stay tuned for the testing tomorrow. Thanks again.

[gervaisc]

Hello mon2 , I have completed the step #4. Nothing very conclusive except L7806 ? BTW I fixed my probe leads, shorting them together is now 0.31ohm. I (re)measured the problematic resistors, still high.
R7021 = 11ohm
R7022 = 111ohm

Note that I did not solder them off the board yet.

I also noticed that plugging USB-C before doing step #5 to measure PPBUS_G3H it is now gone!... 0V. The USB-C input current is now stable at 0.025A, I guess this is a bit worrisome...

Results from GND to inductor:

L1350 = n/a
L2800 = not tested, did not remove U2800/U2900 cover
L2900 = not tested, did not remove U2800/U2900 cover
L3100 = 1.7Mohm (4 pins)
L3200 = 1.7Mohm (4 pins)
L4460 = 2.2Kohm
L4590 = 2.2Kohm
L6400 = 2.2MOhm
L6401 = 2.2MOhm
L6520 = 1Mohm
L6700 = 2Mohm
L6701 = 1Mohm
L6702 = 1Mohm
L6703 = 4Mohm
L6960 = 1.5MOhm
L7030 = 2MOhm
L7210 = 107ohm
L7220 = 107ohm
L7230 = 107ohm
L7400 = 142ohm
L7670 = 700Kohm
L7690 = 700KOhm
L7700 = 300ohm
L7701 = 300ohm
L7806 = 6.8ohm (too low??)
L7807 = 1 side not accessible, other side same as L7806 = 6.8ohm
L7808 = 1 side not accessible, other side same as L7806 = 6.8ohm
L7809 = 1 side not accessible, other side same as L7806 = 6.8ohm
L7810 = 78ohm
L7811 = 20ohm
L7812 = 20ohm
L7813 = 2.5Mohm
L7814 = 2.3Kohm
L7815 = 2Kohm
L7816 = 30ohm
L7817 = 31ohm
L7819 = 30Kohm
L7820 = 30Kohm
L7821 = 860ohm
L7822 = 860ohm
L7823 = 220ohm
L7824= 1Mohm
L8100 = 270ohm
L8150 = 300ohm
L8410 = 1,5MOhm
L8420 = 160ohm
L8507 = more than 10Mohm (open, confirmed with C8584/C8585)
L8509 = more than 10Mohm (open, confirmed with C8580/C8581)
L8570 = 1Mohm
L8580 = more than 10Mohm (open, confirmed with C8580)
L8581 = more than 10Mohm (open, confirmed with C8581)
L8584 = more than 10Mohm (open, confirmed with C8584)
L8585 = more than 10Mohm (open, confirmed with C8585)
L9060 = 14KOhm

[mon2]

Remove all power. Measure the resistance to ground of PPBUS_G3H at the fuse.

(Friggen Canada just lost at the penalty kick)

[gervaisc]

From F7000 I assume - it is 2.55Mohm

[mon2]

Review this case study:

https://repair.wiki/w/MacBook_Air_A2...w_at_5V_repair

[gervaisc]

yes this is what I used before writing this post. Rechecked again:
- Between GND and PPBUS_G3H = open
- Voltage PPDCIN_G3H = 5.17V
- I get now 0V on PPBUS_G3H , leading to the ISL9240 bad chip according to the case study. However I do have those "bad" R7021 and R7022. What do you think?

[piernov]

No point in checking random stuff until the charging circuit is fixed and PPBUS_G3H comes back to normal. Replace the resistors for the charger current sensing path and if they burn again you know you have to replace the charger IC.

[gervaisc]

Hello mon2 piernov i changed R7021, R7022 and U7000 (ISL9240) from a donor board I got on eBay. I was quite excited (that was my first experience to change such a tiny component) because I got 20.07v on USBc - reassembled without connecting battery and saw twice the apple logo and heard the chime. But not booted completely, the progress bar was stuck and fan spinning to the maximum. Connected battery and saw amp went from .4 to 1.4 amp, always cycling every 5 secs or so. So after tried some random things such as trying to go DFU or SMC reset … and got after about30 mins 5v on USBc. Dissembled the board and my PPBUS_G3H is back to fluctuations from 0V to something. I am afraid that ISL9420 will die again… any thoughts???

[mon2]

Remove all power. Meter in resistance mode.

Measure the resistance to ground of PPBUS_G3H.

What is the resistance in ohms?

[gervaisc]

Thanks mon2 for your help. I had a bit more time to spend on my "project" today. Early in the morning I got a steady 20V at USB-C, and PPBUS_G3H was 12V, looks stable. Put the board back in the laptop (I had hopes), then USB-C input was back at... 5V. Back to square 1 and PPBUS_G3H is cycling (fluctuates again). Decided to remove R6961 (to open the circuit at U6960) then the PPBUS_G3H stable 12V. So the problem is no longer ISL9240. I am continuing my investigation. Feel free to chip in !

[gervaisc]

I noticed something weird - if I connect left rear USB-C port (Let's say), I get 12V on PPBUS_G3H. If I disconnect it (or change to left front), then reconnect after the PPBUS_G3H is gone. If I wait let's say 10 mins, it is back again. The R6961 is still open.

[mon2]

Inspect the quality of the USB Type C connectors that are on this logic board. Apple made these paddle boards modular (and hope they keep it this way). The replacement boards should be available through Mobile Sentrix (Toronto). We use them often and most of the time, they are more competitive than Aliexpress stores with next day delivery to us in Ontario.