Acer Aspire 5740g run power issue

[sopik78]

Guys, I have troubles with Acer Aspire 5740g. When changed burned U72 TPS51123 for 3/5VALW I cant get 3D3V_S0 a 5V_S0, see picture. What could be wrong here?

[BlueMidnight]

Re: Acer Aspire 5740g run power issue

The S0 power rails will only be active after the power button has been pressed. After pressing the button, PM_SLP_S3# should go to 3V. Your picture shows you measured 0V. That will keep U70/U71 turned off.

But that 13V doesn't look right. That tells me current is flowing through that path even though the bottom FET in Q41 is supposedly turned off.

You measured 20.5V, 19.6V, and 13V. If the bottom FET in Q41 is off, you should measure 20.5V at all three of those places. If the bottom FET in Q41 is on, you should measure 20.5V, 4.8V, and 0V in those same locations. It is acting like Q41, and possibly even Q38, have shorts. Either that, or they are pulsing on and off quickly (which they shouldn't be) and 13V is just the average voltage.

A pulsing voltage can be checked for using the AC volt setting on your meter (or an oscilloscope). If none is present, then Q38/Q41 can be removed to check if all three voltage measurements rise to 20.5V. If so, one/both of those is bad.

If a problem with that circuit can be ruled out, and if pressing the power button does nothing, then you will have to find which prior step the motherboard is not doing properly.

Check the following until a problem is found:

1. Check that the EC (also known as the KBC or SIO) has set PM_RSMRST# to 3V to wake up the PCH. (Make sure you are measuring the signal with "PM" in the name). If not, then there might be an EC or EC firmware problem (or a bad capacitor, etc. on that circuit).

2. KBC_PWRBTN# should be at 3V, then go to 0V when button is pushed, then go back to 3V. If not, then there might be a problem with the power button or circuit.

3. PM_PWRBTN# should be at 3V, then go to 0V, then go back to 3V. If not, then there might be a problem with the EC or EC firmware (or a bad capacitor, etc. on that circuit).

4. PM_SLP_S4# should be at 3V. If not, then there may be a problem with the PCH (or a bad capacitor, etc. on that circuit).

5. If all the above steps show no problems, then there may be a problem with the PCH (or a bad capacitor, etc. on the PM_SLP_S3# circuit).

[sopik78]

Re: Acer Aspire 5740g run power issue

Many thanks to BlueMidnight for so detailed explanation! It helped me a lot.

I found the Q41 with some leakage and changed it, so I have 20V now on pin 5 and 6.
But I still cant obtain PM_SLP_S3 signal.

Due to instructions I found:

1. PM_RSMRST 0V

2. KBC_PWRBTN 3V to 0v to 3V

3. PM_PWRBTN 3V to 0V to 3V

4. PM_SLP_S4 0V

So I guess the problem might be with the PCH.

[piernov]

Re: Acer Aspire 5740g run power issue

Sounds like corrupt EC BIOS if you're not getting RSMRST, apparently that's not uncommon on those series (57xx). But first make sure you've VCC on the EC and that ECRST# is high.

[sopik78]

Re: Acer Aspire 5740g run power issue

Thanks Piernov.

I have already tried to flash EC and BIOS but with no success. VCC on EC 3.3V and ECRST 3.3V as well.

[BlueMidnight]

Re: Acer Aspire 5740g run power issue

You've made progress. Now we need to know why PM_RSMRST# is still 0V.

It is possible that the PCH is bad. I am worried because there appear to be 2 separate problems that have the PCH as the most direct thing they have in common. But I see some other things in the schematic that you can check.

The attached pictures gives an idea of what that whole circuit looks like.

You should test R534, R535, and D22 first, so you don't go crazy chasing other things. R534 should be 10k ohms and R535 should be 100k ohms. Measure between pins 1 and 3 of D22 with a multimeter in diode mode, it should have no continuity in one direction and around 200mV in the other direction.

The resistors R534 and R535 want to put about 3V on PM_RSMRST#. But the EC will hold RSMRST#_KBC at 0V until after it turns on the 3V/5V switching outputs. Due to schottky diode D22, PM_RSMRST# will be pulled to a little above 0V (the forward voltage drop of D22 should be around 200mV).

The fact that D22 exists may make the diagnosis a little easier. Pay close attention to the exact voltage on PM_RSMRST#. If you actually measure very near 0.00V then you know the cause is on the PCH side of the diode. So, R534 (or the traces connecting it) are bad, R535 is shorted (very rare for a resistor), D22 is shorted, the PCH is shorted, or some other object is shorting that trace.

On the other hand, if you measure at least 0.1V-0.2V on PM_RSMRST# but measure 0.00V on RSMRST#_KBC, then you know it is being pulled to ground on the EC side of the diode. The difference between the two voltages is because of the forward voltage drop of D22. In that case, either the EC is bad or the EC firmware is bad.

[sopik78]

Re: Acer Aspire 5740g run power issue

Thanks BlueMidnigth for guiding me again.

This going to be little crazy, because D22 and R534 are not present on the board. R535 near 100kohm, RSMRST#_KBC 3.3V, PM_RSMRST# 0V. I have no idea where PM_RSMRST# would come from.

[BlueMidnight]

Re: Acer Aspire 5740g run power issue

Is there a spot for them and they aren't there? or is there no spot labeled D22 or R534 at all?

What is the part number of the board? It should say something similar to 48.4GD01.0SB and 09285 on it. Let's make sure we have the right schematic.

[sopik78]

Re: Acer Aspire 5740g run power issue

Labeled on the board but not present.

Board number 48.4GD01.01M 09285

[BlueMidnight]

Re: Acer Aspire 5740g run power issue

Hmm alright. The only schematic that seems to be out there is for "48.4GD01.0SB". According to page 39... "SA", "SB", "1", and "1M" are PCB revisions.

I found some pictures of the "48.4GD01.01M" board and it looks like none of them have D22 or R534. Without an updated schematic, I'm not sure what changes they made.

It seems like the EC is doing what it's supposed to do (since RSMRST#_KBC is 3.3V). So, either the PCH is bad, or there's some unknown problem in between the EC and PCH. They must be able to talk to each other in some way.

You could do some continuity tests to find out what components RSMRST#_KBC and PM_RSMRST# are connected to. Then draw your own little schematic. It might not be too difficult because all the important components are probably around the EC or PCH on the board (but could be on the front or back side of the board).

[meritosari]

Re: Acer Aspire 5740g run power issue

Just type RSMRST# not PM_RSMRST# on your PDF reader.


link HD image:
https://drive.google.com/drive/folde...WJYTW1QSGZoZjQ

[BlueMidnight]

Re: Acer Aspire 5740g run power issue

Just type RSMRST# not PM_RSMRST# on your PDF reader.

According to the schematic, the only purpose of the net labeled simply "RSMRST#", despite the name, is to force S5_ENABLE and ECRST# low (and therefore, turn everything everything off except the linear 3V/5V regulators) when a thermal sensor trips while in the S0 power state. It does not affect the RSMRST# pin on the PCH in any direct way.

But the board definitely has some changes in components and/or labeling between the available schematic and the final PCB. Even RN46 and R408 in that circuit are completely missing from the board (not even a label).

The high quality images are super useful though.