Linux Tempreratures

Hello, I apologize if this counts as double posting, but I am asking a question in a different light (at first I thought it was a driver issue until I went home and hooked it up to my cooler):

Before a few weeks ago my idle temperatures would not change much based on the presence of a notebook cooler, but recently without the notebook cooler, the idle temperatures are starting to hit load temperatures. Could this be because the notebook ceases to properly draw in air?

The following temps are after a while (~1 hr) and fluctuated very little
The GPU clocks are at minimum (169/100 MHz) so powermizer is working
The CPU clocks are also at a minimum (800 MHz)

Temps with Cooler:
CPU: 17 C
HDD: 34 C
GPU: 53 C

Temps without Cooler:
CPU: 35 C
HDD: 52 C
GPU: 70 C

The former is at home and the latter at lab, but the ambient temps are similar.

Thanks in advance for any replies!

 

A fan is likely broken.

 

Any suggestions on how to remedy this problem? I had this problem in windows too but really noticed it in Linux. I suppose RMA?

 

Sorry for the pedestrian questions... but are you on the latest BIOS? Also did this just start happening ... because maybe your vents and/or fan is clogged with dust and gotten to the point where it's a problem expelling heat.

 

I cleaned out my vents pretty well I think (I opened up the laptop as much as I could). I could try updating to the latest BIOS. I've had some trouble with that but I could try again (was using easyFlash, but for some reason it didn't work).

Also, I do think I can hear the fan going on... but maybe its not working correctly.

Thanks for the suggestion!

 

Sometimes the BIOS code gets updated with better thermal management stuff, so it's def worth a shot. If you already cleaned it out, then that's off the checklist. Since you're still in warranty, checking out the thermal pads is a *bad* aka warranty voiding idea, but that'd be my next suspicion next to a bad GPU.

 

Thanks! I'll give it a try ASAP (maybe not for a while since I have to do work in the lab) but until then I'll just have to bring my cooler since those temps are actually pretty good.

Would a bad GPU be affected by the presence of a cooler? I thought that the temps would go up irregardless if it was a faulty GPU.

Thanks again for your help!

 

Well... since you've ruled out the driver package, it's some sort of lower level problem. Whether that's firmware (BIOS), or physical, is hard to tell. The cooler is just going to mask the problem. It shouldn't get that hot by itself. I don't really expect the GPU is totally faulty, but it could just be a bad one making too much heat. But my main suspicion is that the thermal pad that's between the heatsink and the die is maybe bad, installed improperly from the factory, or degraded to the point where it doesn't move heat anymore effectively. Hard to tell without voiding the warranty, so it'd be best to rule the rest out before sending it back for warranty service.. good luck!

 

Well if it helps, I first noticed the problem when I came back from a ski trip so maybe that triggered something? I would think that the problem is likely with the fan since it happened suddenly (aka fine one day, problem the next). Also, the cooler could act like a substitute fan which would explain why it alleviated the problem.

Thanks again for the suggestions and I'll try the BIOS update ASAP.

And how should I go about submitting this problem to ASUS? I've never had to do anything like this before... plus its research crunch time so I can't afford to be without the computer for long. Maybe a local repair shop?

 

Yeah, maybe it got jostled a bit and messed something up?

The thing with the warranty is that you have to register your notebook:
http://member.asus.com/login.aspx

If you have it serviced at a regular shop they'll void the warranty too... so maybe just keep using the cooler for now. Your temps aren't ridiculously high anyway, actually not *that* bad, but if they're noticeably higher than before it's an issue. My GPU (9700M GT) gets to 100C easy under heavy heavy load.

 

Might be interesting to install speedfan or some other program that can monitor the fan speeds and see what it has to say.

Something i haven't seen anybody mention: Can you feel air coming out of the vents?


Something is definitely wrong there IMO, a notebook cooler shouldn't make that large of a difference.


On a related note my brothers laptop start giving him fan warnings a month or so ago (it's a gateway and it runs a fan test as part of POST, if the fan doesn't spin it refuses to boot and turns itself off), it turned out to be a small lint/dust ball tucked in the corner of the fan housing where it could just barely touch a fan blade if the fan stopped at the perfect point, at which point the motor didn't have enough torque to start the fan.
I'm not familiar with your laptop type specifically, but on some you can remove the fan without disturbing the heatsink(s) or tearing the entire laptop apart, if you can i'd try that.

 

Update: The problem was with the BIOS and was repaired with a BIOS update.

Something happened that caused the laptop to read the CPU temperatures about 30C lower than they actually were (I was getting 0C at one point) and so the laptop would never activate the fans, causing the other components to heat up.

I'm glad that I decided to try flashing the BIOS (thought it was a shot in the dark) before fiddling around with the hardware or taking it to a shop.

Thanks to everyone that responded (esp groceries with both the correct solution and constant help )!

 

Even with a cooler, 17C for the CPU sounds ridiculous.

 

It was because the BIOS was reading it incorrectly. I think that the sensor was measuring it 25C too cool. Right now my CPU is idling 30-40 (after a BIOS flash).

The problem with reading the CPU too cool is that the fan then never activates.

 

I've had computers boot up at 12 degrees. It was fairly cold in the room...