Clevo P150HM dead NVidia 560M and maybe more

Hello,

I am very new on the community and this is my first post.

Actually I found this forum because some days ago my Clevo P150Hm died on me. I was playing when it suddenly shutdown. At that point I noticed that the PSU was overheated and seemly was not working. Once I unplugged the cable from the notebook the PSU came back to life. So I understood that the problem was the computer itself. After disassembling some parts I realized that the problem was being caused by the NVidia 560M that had apparently died, what I confirmed later with a multimeter. The GPU is short-circuited.

Now, I am thinking in buying a new graphics card and trying to revive the my notebook (I don't want to give up on it). But, after some experiments and based on posts that I read here in the forum, it seems that when the GPU fried it carried other parts of the computer with it. I say this because when I try to turn on the notebook without the GPU nothing happens: no lights, no fans, no bips. However, even off, it is still charging the battery correctly and the indicating charging LEDs turn on.

Trying to figure out if something else was dead, I experimented removing different components until it remained with the motherboard only. But, in every case, the computer does not turn on or show any visible sign of life. I am thinking that the MoBo is also dead. So, other than buying and trying a new MoBo, is there a way to verify if the MoBo is really dead? Is it normal that the computer does not turn on or show any sign when I hit the power on button?

Thank you for any insight and sorry for the long post.

 

A correctly working motherboard should power up without a GPU but I do not expect it to boot, even with a blank screen. If LEDs turn on and the CPU starts warming up the motherboard is working as expected.

UPDATE: Actually, checking your motherboard schematic you are not missing the GPU detect resistor like the EM series, so you should boot to a blank screen with the GPU pulled.

 

Typically if it does NOTHING that is a pretty strong indicator that the MOBO is done.

 

It might not have been the GPU but a surge from the PSU and most of the power stages between it and the GPU are dead such as the CPU VRM section. The whole board is likely to be a write off.

 

Hi Meaker@Sager,

Thanks for commenting.

It seems that also the GPU is dead. I tested the resistances with a multimeter as described in a post I found here, and all the readings indicated that it is in short-circuit. Besides, when I remove the GPU and plug the PSU to the notebook the battery starts to be charged correctly and the PSU LEDs stay on and stable. On the other hand, when I insert the GPU back, the PSU LEDs goes off and it start to do a strange noise and overheat.

What do you think?

 

Hi Khenglish,

Thank you for checking the schematics and the information.

Thinking on what you said about the EM series, I am now in doubt. Actually I cannot say with 100% sure that the hardware in my notebook is really that of the HM series. Let me explain why, I bought it in Brazil from a builder that employs the Sager/Clevo hardware, and of everything in the computer, the only thing that says it is a P150HM is a small adhesive tag in the bottom of the chassis with a bar code and the code "P150HM". In fact, it was with this code that some years ago I found out that it was a Clevo.

So, now I have new questions. How I verify if my MoBo is really from the HM series or from the EM series? Is there a way to identify it? And, if it is from the EM series, what are the consequences of the missing the GPU detect resistor?

Again, thank you very much.

 

Did it ever use optimus?

 

Meaker@Sager, please forgive my ignorance, what do you mean by "optimus"?

 

My EM series is modified to be more like an HM so I am trying to guess at what a true HM model does with no GPU installed.

An unmodified EM system will boot and display an image without a GPU installed.

An HM system will definitely not display an image. It may not even boot. It may just turn on the power LEDs and backlight and do nothing else. I do not know what you mean by "nothing happens" so I am trying to be descriptive on what is expected. There is a huge difference between absolutely nothing, like no power LEDs, backlight, versus the motherboard powering everything up but freezing on startup, which some people may describe as "nothing happens".

 

Hi Khenglish,

Well, in my case the "nothing happens" means exactly absolutely nothing, like no power LEDs, no backlight, no fans, no bips.
So, based on what you described, the MoBo really seems dead

There is just something that intrigues me a bit, the fact that even with doing nothing, the MoBo still lets the power reach from the PSU to the battery and charge it. The charging LEDs also indicates when the battery is full. Of course, all this with the notebook turned off. This makes me wonder if the charging circuit is independent of the remaining MoBo circuitry.
One small question, in your first reply you said you verified the schematics of the P150HM MoBo, could you please tell me if it is possible to find it available on internet?

Thank you very much for taking the time and checking all this details.

 

The thing is the charging circuit is not independent. If the battery can be charged, then the motherboard's main power rail is getting powered up.

Motherboard schematics are in the P150HM service manual.

 

It is more or less isolated, there is the input protection circuitry, and then the battery charger IC and change-over to/from battery and AC power are all at the early stages, and the battery IC can work standalone, if the led powers up, the EC seems to be functional as well.

By main power-rail, do you mean, the 19v rail?

Or in Clevo's is stepped down first?

Most of my experience is with Asus mobos, and the battery charger IC is powered right after the reverse polarity protection MOSFETs, the 3.3v aux rail needs to be present, but that one is a bit hard to kill usually.

 

Clevos do not have a step down on the system voltage. The battery is charged by the 19V system power rail with a buck converter to step the voltage down for the battery. If the battery can charge, then the system must have a functional main power rail.

 

Same as the Asus, HP's, Acer, and maybe all others.

The battery charger IC depends mostly on the EC working so it can talk with the battery pack via PMBus, the chipset, or the CPU VRM's can be dead and the charger section can still work.

Can you try to remove your BIOS battery, battery, and AC, leave it for a couple minutes, plug the AC back in and see if there is any activity of the status leds/fans?

 

Does it fire up if it has no CPU? If it does the CPU VRM is dead.

 

Hi Meaker@Sager,

No, it does not power up when I remove the processor. Actually, my ultimate test was to remove everything (GPU, RAM, processor, discs) until only the motherboard remained, and even in that case no power up, either connect to the PSU or to the battery only.

 

Do you have a multimeter?

 

Hi Senso,

So, I did as you said and remove every power source (BIOS battery, battery, and AC) and left it like that a couple of minutes. Then I plugged the AC back but sadly nothing changed. Still no apparent response from fans or LEDs.

Yes, I have a multimeter. Any advice?

 

Its not unusual for fermi cards to short. I've had 3 GTX 570Ms all shorted.

It's likely that the board has some damaged componment that prevents its startup. Since this is very likely caused by the short of your GPU, it's very likely that you will be able to locate the region of the board with the naked eye where it got damaged since a short usually ends up melting a chip or other components.

 

This is strange because the HM series does not have a GPU power switch. The GPU dying can only kill the 3 FET switch between the wall power and the main power rail.

I'm thinking the battery saying its charging is a false reading. I can make a picture of where to probe to check if the main rail really is powered or not.

 

Hi Danishblunt,

Yes, after reading many post here in the forum I discovered that my 560M kind of had a long life (~ 7 years).

I did verified the MoBo in the area around the graphics card trying to find any sign of defective component but I did not had much luck, everything seems normal.
I think I will try to get an amplifying glass and give it another try. Maybe I let something to pass.

Thanks.

 

Hi Khenglish,

It really seems strange but the battery is really being charged and also discharged over time when I let it plugged to the notebook, as I could observe.
About the picture, sure it would be of great help. However, I hope this is not a nuisance to you.

Thank you.

 

Not every failure is visible, you would need to do circuit analysis to dive further (the start would be measuring power rails)

 

Hi Meaker,

Yes, I know. The problem is I never did it, at least not in a notebook MoBo. I will try to learn a bit more the days to come.

Thanks anyway.

 

Sorry for the delay in uploading this. Power should be present at the pads above even with the system off. There are many pads to check if the main power rail works, but this spot should be easiest to probe.

Make sure you don't accidentally short out these pads with the multimeter. If you do you should just make the PSU emergency shutdown with an electric flash, but you might get something worse. To reduce short changes you can use one pad from the pair that I boxed, and another from any of the 5 identical pad pairs close by to the right.

If power is not present, put your finger on the the square chip in the bottom right with "ite" on it. This is your embeddeed controller. See if it feels hot. It should be room temperature.

Let me know if there is no power. Then we will test EC control lines and the power switch FETs.

 

The mxm power rail should be easy to measure too.

 

HM systems do not have a separate MXM power rail. The board has FET pads for it but it is bypassed with jumpers.

 

Hi Khenglish,

Don't worry about the delay, I imagine you are busy.
Thank you very much for the schematics. I will do the measurements and then report back here.
But first I have to disassemble everything and get the MoBo off the chassis. I will do it this weekend (kind of busy too).

I get back as soon as I get some results.
Thanks again.

 

Hi Khenglish,

I finished doing all the measurements you pointed out.
There is 19V in the pads you indicated. In fact, there is also 19V in all the other four pads to the right of the one you marked.
Besides, the embedded controller is not getting hot. I also did not found any other component getting warm.

So, since this indicates that there is power on the main rail, this is bad or good?

If it would not be a burden to you, do you have more advises?

Thank you.

 

There are multiple phases that deal with the main input which is why you got the same voltage at those other pads. This is good, it means the main power is reaching the board, that's the first step.

 

This is mixed. It means your main power rail works. If it was dead though, replacing the FETs on it is a fairly easy fix.

When you press the power button do the wall power and battery LEDs turn off (this means there is a short), or is there absolutely zero reaction?

You're sure you didn't break your power button?

 

Hi Khenglish,

No, there is no reaction when I press the power button. Once I plug the battery and the PSU to the MoBo the two related indicating LEDs become yellow (after some time charging the one indicating full battery becomes green). When I press the power button nothing happens, no sounds, no LEDs blinking, no component heating.

Well, I am pretty sure the power button was working before the GPU died, and I did not put any strong pressure over it while doing the tests. Nevertheless, to be sure I will test it with the multimeter.
I will reply back when I am done.

Thanks.

 

Hi again Khenglish,

I tested the power button and it is working well.
However, I think I found something. While checking the components close to the battery connector I found a small IC with a burned spot on its surface. I scratch a bit with the nail and a carbonized powder was released.
The problem is that most of the markings disappeared and I am not sure I identified it well on the schematics. It seems is it the component named "PQ61", but I am not sure. I marked it on the image I am sending attached.

What do you think? does it make sense to you?

 

Actually, paying more attention to the letters and numbers still visible and checking on the schematics, I think the IC is the P2808A1.

 

I can't find a PQ61 in the schematics. Are you sure it's not maybe PQ60 or PQ64? These are power switches on the 3.3V rail, which the Embedded Controller needs to run. If they are malfunctioning they may be disrupting the 3.3V rail that the EC uses.

Find inductor PL10 on the motherboard. Check either side vs ground (you can use any metal post or negative end of a capacitor for ground) and see if it is correctly generating 3.3V. It will do this even when the system is off.

 

Hi Khenglish,

I have check PL10 and it is correctly generating 3.3V. I suppose this means that PQ60 and PQ64 are working correctly, right?
I verified the labels on the MoBo again and the schematics and that IC that I marked is not the PQ61, as you said.
I am still thinking that it is the P2808A1 depicted in the top of page 39-B of the service manual.
The problem is that I don't know how to test if it really failed.
I will try to find how to verify this IC.

Thank you for the advises.

 

Have a look at the data sheet, is it behaving normally? Are any of the pins shorted?

 

Hmm. So you have main power and 3.3V power. It sounds like you should have a laptop that turns on.

I will look into that component within the next day.

 

lol that switch detects if the power button is pressed. So now we know why the system does not turn on.

Bridge pins 5 and 6 on that chip together with a needle or something. This will bypass the chip and we can see if the laptop boots. Note that if the chip isn't dead, this bridging may kill the chip. It really does sound like it is dead though.

Using a heat gun would be ideal to replace the chip. Using an iron to replace chips like this is risky to the motherboard, but it is doable.

 

There are a lot of passives around it, that would be some nice soldering iron work to pull that off.

 

Hi Khenglish,

Sorry for the late reply, and thank you for the instructions.
I did the experiment you suggested, but sadly I did not get any response from the motherboard. It did not turned on.

At this point, provided my lack of electronics knowledge, I'm without new ideas.
Maybe I need to start to consider findind a MoBo to buy if I want to revive this computer.