No power on G512LI

[mon2]

Re: No power on G512LI

I would replace the resistors (both of these) and confirm the voltage divider functions as per the calculator. Then add the charger IC again or a fresh one.

[vrasp]

Re: No power on G512LI

What about the PD9002 diode and the PC8905 capacitor? I m curious to know why they aren't considered as potential culprits.

[mon2]

Re: No power on G512LI

Technically, everyone on this party line is a suspect.

You can measure the resistance to ground at the ACDET pin. If the measurement is ~20k then likely the other parts are ok.

Certainly can quickly remove the others to remove them from the equation. Start with the capacitor.

[vrasp]

Re: No power on G512LI

Sounds good.

I can't find the same MOSFETs (PKCH2BB and QM3058M6AC) in online stores. When I filter out the results for similar MOSFETs with low Rds < 1 milliOhm (PKCH2BB) and < 30 milliOhm (QM3058M6AC) and a Vds limit of > 30 V, nothing with the same package/case shows up. Can I get something from a different package or will it not fit?

Also, do you have any advice/tips on how not to lose resistors/capacitors or other small components when you desolder them? I keep a white piece of paper next to me and place them on there but it's so easy to lose them. How do you even store something so small?


Thank you!

[mon2]

Re: No power on G512LI

If you are not in a rush, there is Aliexpress:

https://www.aliexpress.com/w/wholesa...home.1000002.0

https://www.aliexpress.com/item/3289...300005886053_2

NB: The same vendor should be able to source both mosfets.

Alternatively there is Digikey / Mouser / Arrow for equivalent devices from TI or similar. Donor boards are another option.

To store parts, use small antistatic bags with tweezers. You could even use sandwich / ziplock bags since passive resistors / caps are not static sensitive.

[mon2]

Re: No power on G512LI

Review this one for an option to replace PKCH2BB / QM3058M6AC (suitable for both):

https://www.digikey.ca/en/products/d...ON6500/3340454

[vrasp]

Re: No power on G512LI

Super helpful! I m placing an order for everything on that line. Thanks!

[vrasp]

Re: No power on G512LI

Hey,

I've been trying to figure it out by myself but got stuck.

I replaced the two MOSFETs as well as PR8904, PR8905, PC8905, PR9024, and PD9002. I was still getting about 1.7 V at ACDET (with battery IC removed) I proceeded to desolder pin 2 of PD9002 to isolate the circuit from AC_BAT_SYS and got a nice 3.2 V at ACDET (the 20KOhm from the voltage divider was replaced with 24.3 KOhm so that makes sense). It appears the issues is coming from the AC_BAT_SYS line.

There is continuity with ground on the AC_BAT_SYS line. I removed PQ8903, PC8901, PC8902, and PC8903 but the short is still there. What should I do next?

Thanks!

[vrasp]

Re: No power on G512LI

I also visually checked the rest of the AC_BAT_SYS line and didn't see anything odd. There are so many capacitors and resistors connected to this line... Is this the kind of scenario where I would inject 1 V then 2 V and so on and check what gets hot?

[mcplslg123]

Re: No power on G512LI

AC_BAT_SYS is the main source line for powering almost every rail-so problem could be with one of those rails. It's difficult to find which rail is pulling AC_BAT_SYS line to GND. Best option is to inject 1V(not more then this)/5A on this shorted line and see which area heats a bit. Perhaps you wont feel any burning component but very slightly warm area of the board will give you an idea of possible shorted component. You may leave 1V/5A injected for 10-15 mins if you dont find anything hot.

[vrasp]

Re: No power on G512LI

I tried this. First, I noticed that at 1 V, the amp draw is 0.87 A. My understanding is that it's not fully shorted to ground or it would draw up to the limit I set on my bench PSU which was set to 2 A. I found that U0301 was getting warm (>30 C). This is the Intel Cometlake-H processor. Same with all the capacitors on the other side of the board where U0301 is. Is this normal?

In case this helps, the resistance to ground on the AC_BAT_SYS line is 2.6 Ohms.

[Sephir0th]

Re: No power on G512LI

That's a typical so called high-side short, means for this case, the Mosfet of the buck converter between the Main power rail and the CPU Core supply is internal shorted and has probably send 19V into the CPU already. Usually That's a dead-end.

However, in some rare cases the CPU survives that might work again if you replace the whole buck converter (all semiconductors) related to the CPU Core supply.

Since we (also myself) were confused with ACDET line in addition, we actually missed to do a simple resistance measurement of the main power rail. Otherwise the issue would have been clear from the beginning.

[vrasp]

Re: No power on G512LI

Thanks! I might as well try that. I'm not sure how to find the buck converter for the CPU core supply. Can you tell me where it is on the schematic or what the abbreviation is?

[Sephir0th]

Re: No power on G512LI

Thanks! I might as well try that. I'm not sure how to find the buck converter for the CPU core supply. Can you tell me where it is on the schematic or what the abbreviation is?

If you want to stay safe, everything which is circled here. However, I suggest to confirm the diagnose by lifting either each coil or each of the dual-mosfets, till the short of the Main power rail (AC_BAT_SYS) is gone