No 20V at VBUS anymore in combination with pulsing ALW voltages most likely means that
a) that the replacement PD-controller was not soldered properly, two times. Do we have pictures of proof for correct soldering from different angles? This assumes too, that the statement of "auto-programming" applies here. Was the EC rom checked for PD firmware or is there another proof that this is the case here for sure (100%)?
And b) either a second underlying issue further down the sequence (not on the 3VALW/5VALW but later) or it has something to do with the PD-Controller...
Sephir0th
Badcaps Legend
Last Activity: Yesterday, 07:15 AM
Joined: 10-17-2020
Location: Apen, Lower Saxony
-
I would say its worth trying to replace PRT402 with the one from the donor and report. Be careful, these thermistors are indeed somewhat heat sensitive afaikLeave a comment:
-
It's part of the iGPU, so i would not remove it since we don't know how this all is connected together.
What i am more interested in now is this one. A typo or really in Ohms? Additionally with NTC thermistors the resistance should decrease and not increase when the temperature rises. So the previous measurements of the four thermistors are somewhat confusing.
Maybe you can measure again after the board was running for a while after 10 - 15 mins or so
It would be great to finally find a faulty one, but even four faulty thermistors doesn't sound right to ...Leave a comment:
-
Doesn't sound promising. Did you test from charger only or from battery too?Leave a comment:
-
Best bet is to replace the charging controller i would say since nothing else is off at this point.
Fingers are crossedLeave a comment:
-
It could mean that a even more critical issue was overlooked.
So what's working and what not right now, in detail? 50mA is a really parasitic consumption. A dead chipset should be ruled out.
And of course issues caused while soldering the replacement EC (bridges/shorts/cold joints/etc.)Leave a comment:
-
So PROCHOT# and everything which is connected to it definitely isn't the source of the limit, so we need to think about other options which could result into the current behaviour.
The charger is OEM, known good and of correct wattage?
I would check PR103/104/113/114 for sanity.
Ultima ratio would be to conclude the CPU being internal fried in this way after everything else was ruled out.
And history please! What was initially wrong with the 1.2V supply (for RAM?)? The statement alone is a strong indicator for something wrong with the CPU it...Leave a comment:
-
EC_SMB_DA1 and EC_SMB_CK1 missing implicates that not only the +3VLP is missing but +19VB too. In your initial post you wrote it is present.
Of course such mistakes can lead the approach into the wrong direction. But it's fine since it was advised to check the charging controller anyway. Correct DCIN, ACIN, VDD and VDDP voltage in combination with no ASGATE to drive the Mosfets normally means a dead charging controller... except the Mosfets were replaced with wrong parts.
other subtle faults are possible too, like PR304, PR305 failed or low resistance at VDD/VDDP. Yes,...Leave a comment:
-
Add at least the Pincount or even better, add the correct designations from the schematics. To use the boardview as reference is not sufficient here.Leave a comment:
-
I think you could finally share the voltage of PROCHOT# respectively VR_HOT# to get a idea whether the issue really originates from this signal.
This all is digging in the mist so far.
Edit: 1.05V foe VR_HOT# is okay. Hownis it on the other side, VR_HOT_P# ?Last edited by Sephir0th; 07-26-2025, 09:21 AM.Leave a comment:
-
I tried to be polite as possible.
It is normal that the first, the second, the third and even more repair attempts are not succesful. This is part of the process and the typical reason is that the issue was made worse at some point where it is not (economically or without special tools) repairable anymore. We all had to go through this.
But as I said we need something like a overview. I can't follow anymore without investing hours of time examining the thread.Leave a comment:
-
What about a simple good old resistance to GND measurement. What do we have on the 3VALW?
Per default the Chipset is always a suspect when the 3VALW is not alright.Leave a comment:
-
I admire your persistence but at some point someone should realize that it can not go on like this after filling this thread with 11 pages for almost 3 months. Half of it irrelevant and misleading and without any structure if m1ch43lzm wouldn't assist anymore.
What it really needs now is a structured overview of the original issue/history, everything what was done so far, and what has changed after that (or not). Only this way someone will be possibly able to make sense of all this.
You decide which way you want to go, but it definitely can't go on like this....Leave a comment:
-
There is no point to inject anything into the 2.5V rail without indication, so it's hard to understand this advice. It was already clarified that the 0 Ohms short on the P3V3AL originated from U60400 and the resistance is still not healthy.
The real solution is as simple as obvious: find the other defective components on the P3V3AL track and remove them until the resistance to GND reads healthy again. Chipset fried too? Possible.
It's a common sort of failure. More often with the 5VALW, than with the 3VALW but still common. In these cases its the best to know something...Leave a comment:
-
To share a overview about the whole board situation (voltages and resistances) and Not only of those areas what you think might be important.
The key in such cases to share as much informations as possible, since the given information obviously doesn't lead to the solution.Leave a comment:
-
It's most likely a capacitor between B+ and GND, possibly with same specs as the "neighbour". However a missing cap there would not prevent the board from turning on. Its purpuse is just to flatten the voltage. So it would be helpful when you finally share what's really going on with the device, with full history and all important observations stated.
To rely on the cap only is for sure not the right approach here.Leave a comment:
-
It's a dead GPU, isn't it? Search in google for "DIS to UMA conversion"Leave a comment:
-
This is normal.
We're discussing here about completely normal behaviour of a PWM-Controller. It sets itself into protection. You need to wait until the circuit is discharged. This can take a while. Then you can measure the "real" resistance to GND of the rails.
Additionally the EN pin doesn't control the buck converters, but it is responsible only for basic function and the LDOs.
The Step-Down voltages are enabled from ENTRIP1 and ENTRIP2 instead which can be found out while reading the datasheet of the IC. The circuit can be watched in the schematics...Leave a comment:
-
A unprogrammed PD-Controller will not result in no-P.O.S.T. as long as the device can be powered from the proprietary Jack and/or from battery. IF the PD-Controller is indeed the origin herr (which is not very likely) then it's caused by technician fault (soldering failure).
So is strongly assume you're on the path here.Leave a comment:
-
O.L. is fine but only when assuming you made proper contact with your probes. The value in diode mode proves that the Mosfets have not failed open beside diode mode is not suitable for normal resistance measurements.
Since a shorted Mosfet normally wouldn't prevent the voltage from passing the 1st Mosfet from Drain to Source, it's time to inspect the charging controller. Most likely the ALW are not coming up, the charging controller doesn't inform the EC via ACOK.
And i strongly suggest to inspect other possible related spots for previous work.Leave a comment:
No activity results to display
Leave a comment: