Fan constantly running at high speed

Hi all,
I've been reading on many threads (on this and ASUS support forums) about people encountering what it seems to be the same problem or variations thereof, but none of the suggested fixes worked in my case.

Simply put: the fan on my W7S notebook runs constantly on high speed emitting a loud noise and this has been happening for the last few months.

Situation:
I've checked the temperature on my parts and the temperature monitoring softwares I am using (SpeedFan and CoreTemp) say the hottest part is the GPU (Nvidia8400M) at approx 67-72 celsius. Whereas the other parts sit at about 55 (core) and 47 (HD). To me these temperatures are normal operating temperatures *as far as I know* but the fan, quite obviously, disagrees with me and it seems to be set to constant full speed. In fact, even after a cool night of rest, as soon as I boot the machine it starts running at high speed from the very first second. Even before the OS starts! This is something it never happened before say May '09 and it keeps happening even if I thouroughly cleaned and dusted the fan and the other visible parts with compressed air as elsewhere indicated.
I am afraid there might be something wrong with the GPU fan settings.
If that is the case, I am not sure that substituting the part would address the problem.

Questions:
1) What would you consider "normal" idling temperatures for GPU, core, HDU?
2) Do you have an explanation for what is happening?
3) What course of action would you recommend?

Thanks in advance and regards,
Luca

 

1. Those are average normals for any notebook(except the g series >_<)

2. Sounds like a faulty thermo sensor.

3. If you can't live with full fan speeds 100% of the time I would send it back for an RMA.

 

Tx mk1freak, I have read elsewhere ( here and here) that the latest drivers for my bios (tx to my automatic updates install feature) seems to/may include a default setting for the graphic card fan which is set 100% all the time to prevent the card from frying. This, of course, is not substantiated by any official communication and it'd seem as a quick fix to a graphic card heat issue that Nvidia reports more or less officially for my chip. Or at least, that is what a computer illiterate such as myself understood from reading up about this issue in forums and such. Have you ever heard of all this?
IF that is the case, would you think an RMA can fix this? I am afraid they would just swap the card with an identical model (of course) and potentially substitute the fan thermosensor with no apparent solution to the 100% driver setting. This, of course, if I got the whole thing right.

 

Those temp seems perfectly fine, but the important question is the 67-72 C is your temp under heavy load(3d gaming/benchmarking)or just idling temp? If thats your idling temp, you are definitely having a problem there.

Here is just something i thought that might happen to you that caused you the problem as i had similar experience before, your GPU probably doesnt downclock and constantly runs at full speed, most of the time its because of a bad driver, i suggest you check the operating clocks of your GPU while gaming/idling and see if it automatically downclocks itself, GPU-Z or Rivatuner is all u need to do this.

 

you can try speedfan to manually control the fan speeds link here
but even if ASUS replaced the mainboard/vidcard, then who's to say it won't solve the problem? better they(ASUS) do the work than you (if you can live without the lappy for abit) at least they'll keep at it untill its resolved (or replace with a new unit )

 

Tx guys,
@lidowxx: that is idling temps, I thought they were ok! Tx for the downclock suggestion: I tried GPUZ and this is what I see while idling:
GPU core clock: 400 MHz,
GPU Memory clock: 400 Mhz,
Default clock: 400 MHz,
GPU Temp: 70C.
Don't have any 3D app, any idea on how should I check for the downclocking issue (benchmarking sw perhaps)? Which values should change and how?

@mk1freak: somehow SpeedFan won't show the fans rpms nor let me manually control them (differently to what happens on other laptops on which I run SpeedFan). Some of the buttons/windows are in fact greyed/blanked out.

 

That is definitely too high for an idle, my 260m which produces a fair bit of heat only idles at 50-55C tops.

 

I just checked the default clocks of 8400M gs, its 400 for core and 600 for Memory. Sounds like your GPU indeed might have a problem downclocking itself if thats what u are seeing while idling. Download 3dmark06 from futuremark, run the test and check the GPUz after running and see if the clocks change after u finish the test.

Also did you clean the inside of your laptop recently? If u didnt, the collected dust could somewhat clog the air vents and cause the GPU to overheat.

 

I would definitely suggest this as well. You won't be able to remove the fan, so you'll need to go through the fan blades with Qtips, preferably dampened in sanitary alcohol (not water!).

Additionally, keep in mind that the W7S is a small laptop with a relatively powerful GPU, so it will get hotter than many other, bigger or less GPU-powerful laptop.

And also that the W7S GPU is affected by the nVidia "bad bumps" problem, which in later stages also causes the GPU to overheat.

If you think the fan behavior is caused by the BIOS, why not downgrade the BIOS to an earlier version and check? But you're running the risk of killing your GPU sooner due to that (see above)...

 

First of all, tx all of u guys for great support in troubleshooting this, really appreciate it
@lidowxx:
I did as you suggested and besides getting a, lousy, 1023 3dmark06 score nothing else changes before VS after running the test. Temperatures are always the same (stuck around 70C), clock same (400 and 400), the fan keeps spinning at max speed before, during, and after the test.
@E.B.E.:
Yep, yep: I did the cleaning procedure as first thing (before posting here!) as I routinely do on my other laptop (a "vintage" L5G ) and my desktop as well. Quite a pro at it actually. But thanks anyways for the lead. Also, I understand your point about the powerful GPU in a tiny laptop, it makes perfect sense to me: strange thing is it started having fan issues only recently with no apparent change in the usage pattern (room temp, graphics app, etc.). As you also suggested I should probably try to roll back some updates (BIOS, GPU driver, whatever) and see if the fan quits spinning and the temp remain the same. That should tell if it's a fan setting issue.
Thing is I am not comfortable at all in digging through those updates and messing with bios and drivers by myself. I sort of hoped I could diagnose this in some other way Since the machine is still warranty covered I am going to hand it in tmrw to the Asus center here in my city and see what they say. If it comes back with a brand new GPU and still fan spinning like crazy I think it's definitely a Nvidia heat issue there. What else can you do about it?
Any other input or comment is always welcome. I shall keep you posted on the next steps anyway.

 

Seems your GPU doesnt want to downclock itself automatically as i guessed, i had the same problem before usually caused by driver corruption/bad driver installation, which should be resolved by uninstalling GPU driver/clean installing new driver. The correct way to install new driver:
1. uninstall Nvidia driver from your control panel-add/remove programs
2. restart and boot into safe mod, use driver sweeper(download it if u dont have) to clean the old driver residues.
3. restart and then install the latest official driver from Nvidia,restart

You can give this method a go and see if it helps, if does not, then send it back to asus and let them deal with it.

 

Hi guys,
quick update:
took three full days to get an RMA (call, then ticket, then email, then RMA), then brought laptop in to the tech support center, got it back after five days with a new board: same malfunction. Somewhat worse, GPU now sits around 84 C.
Currently waiting for a new RMA.
I cannot say I am happy with their RMA-assistance procedure, left alone their quality in service. Come on, they didn't even care checking if the fan kept spinning or not, just merely swapped motherboards and shipped it back.
This might be useful info for someone pondering what is the level of assistance of Asus in this country (posting from Italy).

 

I assume 84 C u mentioned is your idle temp? If so. it's absolutely unacceptable
. Another story of poor customer service from ASUS

 

Hi all,
I'll try to bring a sense of closure to this thread. I eventually got back my machine after the second trip to the tech support center. Basically they changed the main board and the thermal module. Got a call from the tech who explained that the fan keeps spinning and the temp are to be considered "normal" for this model. He recognizes that the fan setting (100% speed always) are a fix for a potential GPU heat issue thus indirectly confirming the Nvidia policy of preventing faulty GPU issues by spinning the fans like crazy. All this was of course done under warranty, nevertheless the amount of time it took to deliver this diagnosis is plainly unacceptable. At the moment I am thus stuck with a W7S with brand new components and the same old problem, which apparently is to be considered "normal" from now on for this machine. I thank you all for the advice and beware of Asus customer service! Cheers!