Question Clevo p950hr shutting down after disconnecting faulty fan

[B.D.united]

Hi!

One of the gpu fans in my laptop started making chirping/friction sounds. Unfortunately I wont be able to get the laptop back to the reseller for repair for a while so my temporary fix was to disconnect the faulty fan. Since disconnecting it I have experienced random shutdowns. I believed it to be overheating issues but thermals don't show anything out of the ordinary (gpu temp has been as low as around 50 on shutdown). A few times the mb speaker started beeping before shutdown, but ussualy it was just abrupt without warning or error. The windows event viewer shows error 41 (kernel.power) which is not very informative. The shutdowns have happened both under load but also right after startup.

Am I missing something here? Can the laptop be overheating in areas that dont show up in hwinfo? Can disconnecting the fan cause issues other than overheating? Any help would be greatly appreciated!

Picture of the gpu cooling unit:

https://imgur.com/a/amGLZDS

View: https://imgur.com/a/amGLZDS

 

[Ralston18]

Could well be overheating.

But I see something else.

The image and its expansion appear to show the GPU fans full of/caked with dust.

So airflow is lost (cooling) and the fan(s) will be off balance (chirping and friction).

During a routine check I discovered that a GPU (GeForce GT 640) fan had become quite packed with dust. Had a cover that needed to be snapped open and swung aside to blow out the debris.

That was in desktop pc.

For your laptop, check the laptop's User Guide/Manual for cleaning instructions.

A couple of canned air bursts may solve the problem. Again read all instructions first.

 

[B.D.united]

Could well be overheating.

But I see something else.

The image and its expansion appear to show the GPU fans full of/caked with dust.

So airflow is lost (cooling) and the fan(s) will be off balance (chirping and friction).

During a routine check I discovered that a GPU (GeForce GT 640) fan had become quite packed with dust. Had a cover that needed to be snapped open and swung aside to blow out the debris.

That was in desktop pc.

For your laptop, check the laptop's User Guide/Manual for cleaning instructions.

A couple of canned air bursts may solve the problem. Again read all instructions first.

Thanks for your answer. I will definitely try cleaning out the fans, but I am starting to suspect overheating is not the problem. On my last test the laptop shutdown with a gpu temp of 46 (and the temp was dropping), so it seems unlikely temperature is the issue.

Could it perhaps be the motherboard has some kind of failsafe that turns off the computer when no fan is connected? And if so, why is the timing of shutdown so arbitrary? Sometimes it runs for hours on end and other times it wont stay on long enough to open up hwinfo.

 

[Ralston18]

One thing at a time.....

Clean the fans. If the problem continues then "Step 2".

Step 2, in my mind is to take another look at Event Viewer and Reliability History/Manager.

Especially for something that happens before the shutdowns. What Event Viewer often captures is the restart - not what caused the original shutdown.


How old is the Clevo/Sager laptop? Do you have the model hardware specs?

 

[B.D.united]

One thing at a time.....

Clean the fans. If the problem continues then "Step 2".

Step 2, in my mind is to take another look at Event Viewer and Reliability History/Manager.

Especially for something that happens before the shutdowns. What Event Viewer often captures is the restart - not what caused the original shutdown.


How old is the Clevo/Sager laptop? Do you have the model hardware specs?

Cleaning the fans did not fix the noise unfortunately. The laptop is 1 year old.

The specs are:
i7-7700HQ (2.80GHz) 6MB Smart Cache, 14nm, DDR4-2400MHz, TDP 45W
Intel® HM175 Express Chipset
NVIDIA® GeForce GTX 1070 Max-Q 8GB GDDR5 Video RAM
1x 16gb ddr4 corsair vengeance memory
Samsung evo 960 m.2 ssd (1tb)


I have done some more diagnosis and am confident this is not a heat issue. I tested the system in linux (with the fan disconnected) to see if the behaviour was tied to windows and to have access to more detailed logs. The system crashed in similar fashion indicating this is caused at the hardware level. Like in windows the linux logs simply indicate unexpected shutdown, no overheating. Monitoring the temps in linux gave similar results as in windows. The hottest part was the pch (at ~55) which is not even actively cooled so wouldn't be affected by the fan anyway. CPU temps were all moving comfortably between 40 and 60 (TJmax on this system is 100) and highest monitored gpu temp was 45 (not sure what max temp is set for the gpu, but it handles atleast up to 80). In the end troubleshooting leaves me left with more questions than answers.

My best guess now is that this is a failsafe within the EC. Perhaps some combination of temperature and low/no rpm readings from the fan causes the EC to shutdown the system. The gpu temp at shutdown has however not been consistent (sometimes it was ~40, other times it was sustained around 60 for half an hour before shutdown).

Clevo offers quite a comprehensive manual including very detailed electrical schematics, but I do not have the specialized knowledge to understand those:
http://s472165864.onlinehome.fr/anyware/manuels/P95XHR.pdf

 

[Ralston18]

Without the fan (your temporary fix) it is my surmise that the CPU gets too hot too quickly and the safeguards do indeed shut the laptop down.

Before the temperatures even register perhaps.

Not sure what else to suggest and will hope for other input or ideas.

 

[Rob_in_Hood]

Exactly what I have here. I am not sure how to fix it, I have change the thermal paste, clean up the fan, change the AC adapter, change heat pipe, even use lower end i5 cpu as well, but none of them works for me. Did you fixed it?