Another 820-01700 - A2141 with 5v 0A

[cstout]

Hello all, this is my first post. I'm fairly new to micro-soldering and BoardView and learning as much as I can right now. I've been in the repair business for 18 years now and I'm working on this puzzle of a MacBook Pro and hope to find the cause of failure.

At this point, I'm not sure if I've suffered the more common failure that kills the NAND, but that's what I'm trying to confirm.

Here's what I've checked so far...

The values below are with the OEM 96W charger.

voltage on PPBUS_G3H = 0V
-checked via F7000

voltage on PP3v3_G3H_RTC_X = 0V
-via C3100
-4.75kΩ via C3100

voltage on PP3v3_G3H_SOCPMU = 0V
-via C7893 & 7869

With the bad news of those three tests, are there any other areas that I can verify before moving on to replacing components?

Thanks, everyone.

[mon2]

voltage on PPBUS_G3H = 0V
-checked via F7000

Remove all power (no adapter / no battery). Meter in resistance mode. Measure the resistance to ground @ F7000 (both sides). Post the measurement.

[cstout]

6.3Ω on both sides of F7000 with battery and AC removed.

[mon2]

Ouch. Carefully flux and remove F7000 fuse. Then check the resistance to ground again. Very likely you will see the low resistance on the downstream side of the fuse PCB pad = consumer side. The upstream (producer side; aka ISL9240 side) should be much higher but confirm it. Then power up without the fuse on the board. Do you see a valid voltage on the ISL side of the removed fuse PCB pad?

Would you have an adjustable power supply? At this time of writing, it appears that you have a shorted tantalum capacitor. Hope it is the case for you and with an adjustable power supply - you can find the bad boy and pluck it.

[cstout]

Fuse removed and F7000:2 PPBUS_G3H and F7000:1 PPVBAT_G3H_CHGR_REG both showing 4.3Ω between pad and ground. With AC plugged in, each pad reads 32mV. I do have an adjustable power supply on my bench if needed.

[cstout]

Additionally, F7000 removed has no continuity BUT the pads on the board do (not sure if this is a good thing). Neighboring fuse F7001 does have continuity so it does appear that F7000 is blown.

[mon2]

Interesting and somewhat rare. Remove F7001 as well off the board so that there are no fuses onboard. Then measure the resistance to ground again on each side of the fuse pads. Post each measurement.

Then power up and measure the voltage present on the F7000 or F7001 fuse pad pin # 1 (PPVBAT_G3H_CHGR_REG). What is the voltage to ground?

Fuses are available from Digikey / Mouser / Arrow Electronics / spare donor boards.

[cstout]

4.8Ω - F7001:2 PPBUS_G3H
29.5Ω - F7001:1 PPVBAT_G3H_CHGR_REG (keeps dropping the longer I measure it)

4.8Ω - F7000:2 PPBUS_G3H
29.5Ω - F7000:1 PPVBAT_G3H_CHGR_REG (keeps dropping the longer I measure it)

with voltage applied:
12.29V - F7001:1
12.29V - F7000:1

Fuse 7001 is removed now as well. They took an unusual amount of heat and time to remove and there is no longer continuity after removing it so I believe I cooked it during removal.

[cstout]

Possibly unrelated (seems to be a part of the GPU power circuit), but closer inspection on the underside of the board has revealed that RA602 is missing. Just scanning over it with my magnifying headset - I do not have a microscope.

[mon2]

Can you share a pic of the front face of your meter? Wish to confirm that you are indeed in the low ohms scale. Both sides of the fuse are very low in resistance to ground.

Suggest to practice on a donor board to get a feel for your hot air tools. Lots of flux and keep the heat away from the LCD screen which can discoloration or worse. Using flux and tight circles around the part should allow it to melt the solder underneath. Keep in mind that these are multilayer PCBs so lots of ground place which acts like a heat sink to dissipate the hot air being applied. Dial up the heat as deemed necessary but careful of nearby tiny parts and plastic components. Apple uses mobile phone style parts (ie. 0201 or smaller) on these models so it is a challenge to repair.

[cstout]

No problem, hopefully this makes sense

[mon2]

Very good. Make more sense with your shared pix. See the 2nd meter reading. There is a letter ‘M' which is for mega ohms and that is a lot of resistance. Like what my wife gives when I wish to buy a pricey toy. For logic boards this is good. So if you power up now, pin #1 to ground voltage should be stable and not cause any excessive heating because the heavy load of 5 ohms from the consumer side of this power rail is removed.

Most likely you have a shorted capacitor but is to be determined. It could be anywhere and we need to smoke it out by applying a voltage injection.

would you have an adjustable power supply? You will need one that can dial up 0v8 with support for 5A (amps) or more current. Amazon? Can return it if you don't need it.

with such a power supply, dial up 0v8 and apply onto pin #2 of the pcb pads. Start first with checking the voltage without the fuses on pin #1.

the power injection will cause the shorted part to heat up and eventually crack under pressure - like an interrogation.

ps: my analogy is bad here..but she is not reading this forum. Just know that mega ohms means no resistance to ground. Closer to 0 ohms is a short circuit and will draw heavy currents from the power supply isl9240's mosfets.

[piernov]

A shorted high-side MOSFET should be suspected first, or even SSD voltage regulator as is common on this board.

[mon2]

Thanks. Remove all power. Meter in lowest resistance scale. One meter probe on F7000 / F7001, pin # 2 (consumer side of the removed fuse) with other meter probe to any inductor on the board. Either side of the inductor is ok.

What is the measurement? If the measurement is with a low resistance then that local high side mosfet is probably defective and respectively, a high voltage from PPBUS_G3H has travelled onto the delicate low voltage circuits - causing damage. It will then be a no fix case.

Proceed to inspect each and every inductor on the board to hunt for a defective high side mosfet. Post the location(s) of low resistance in this checking, if found.

[cstout]

Thanks, all. I did a quick test from F7001:2 to L7030 (both sides) and got 6.3MΩ. Please forgive me if L7030 is not a proper inductor (or not one at all) to test with. It appears to be the closest one.

[cstout]

mon2 your analogy was both understood and very funny.

[cstout]

Here's all the inductor-style chips I could see. Board view lists them all with an "L" prefix.

L7030 - 6.3MΩ
L8410 - 9KΩ
L7701 - 429Ω
L8102 - 111Ω
L7660 - 2.45KΩ
L7600 - 3KΩ
LA950 - 45Ω
LA870 - 51Ω
LA860 - 51Ω
LA850 - 71Ω
LA750 - 0Ω
L7231 - 10Ω
L7330 - 34Ω
L7221 - 9.4Ω
L7211 - 9.4Ω
L7420 - 14Ω
L7410 - 14Ω

[mon2]

Wow. This may not have a happy ending. Can you post a pic of the meter face for the LA750 resistance measurement? So to confirm again, no power, meter in lowest resistance scale (if manual) with one meter probe on the F7000 or F7001 pin # 2 PCB pad (consumer side) and other probe on either side of LA750. The reported measurement implies that the high voltage from PPBUS_G3H has most likely travelled onto the PPSOC_S0_GPU power rail which has a typical voltage requirement of 0v9. If true, at the very least the GPU is dead. Many of the other measurements in the above list are similar.

A good mosfet that is not defective will have a very high resistance as in hundreds of kohms or more. So the same comments on the other components showing a low resistance.

For example, L7231 is linked to the regulator that powers the CPU. Such a low resistance means the same - the CPU is dead with such a transient.

[cstout]

Yeah, I'm not optimistic on the outcome for this one but my silver lining is the "knowing" of what specifically has failed instead of where I normally stop, which is - the main logic board is bad. It's intriguing to dive deeper and see - well, what about the logic board is bad?

Anyway, here's a shot of the meter for LA750 to F7000:2. Both sides of LA750 read the same.

[mon2]

Yes, indeed is not good news. The high side mosfet (internal or external - varying with the regulator) has shorted and would have leaked the high voltage onto this low voltage rail which would most definitely kill the target low voltage devices.