G51 Fan vBIOS Mod

My cpu idles around 44 degrees, but my gpu temps are terrible.
My GTX will hang around 55 degrees idle, and the fan will shutoff. Within two minutes the temperature will start climbing to 60 degrees, 65, 70, 75, 80, 82 degrees before the fan actually turns on the starts kicking it back down to 55.
At which point the whole thing starts again.
There is no way my idle temp should be yo-yo-ing between 55-82 degrees C.

Has anyone had some more success with getting the fan to turn on at lower temps?

EDIT: I havemy gtx running at 0.9v, which definately helps with my actual load temps. Could this be whats causing my idle temps to yo-yo?

 

thalanix made a post about this here...
http://forum.notebookreview.com/showthread.php?t=405231#3

Edit btw the linux commands to extract the dsdt are:

Code:

cat /proc/acpi/dsdt > dsdt
iasl -d dsdt
iasl -ta dsdt.dsl

Then that will generate dsdt.hex (which you can view in any text editor).

 

Those are ugly idle temps.

Your temperatures shouldn't breach 70 at idle and light loads. How are your load temps?

 

CPU load temps are 70, gpu load temp is 85.

 

Is that your "usual" temperature, or just the max?


Were you having problems with these fluctuations before the voltage mod?

 

Max temps are good.

CPU and GPU are cooled by the same fan (if I'm remembering correctly, I've been up for a while), so I doubt the fan just "shuts off."

Can you get me a graph of your GPU temps? Might just be spikes.

 

Max temperatures aren't the alarming issue. Like you said they're pretty damned good.

What's confusing is how his temperatures will go all funky during idle.

 

I had this same problem with an old dell d620 where the fan wouldn't kick in until it got too hot. it was due to the sensor not reading temps properly because it was not seated or connected properly, so the temps it was reading was way off, resulting in the lag time for the fan kick on/off. could be why yours id doing what it is.
here's a question though,
does undervolting the gpu also undervolt the temp sensors on the card? seeing as it is a dedicated.

 

Undervolt the temperatures sensors?

That's not how it works, ha ha. Your temperature readings are going to be fine.

 

I think he undervolted his brain.

 

lol let me clarify:
does undervolting the GPU only undervolt the GPU or does it do so for all the power going into the video card itself(which includes the onboard sensors i would think ), not just the gpu, or am i wrong and they are powered seperately?
i was under the impression undervolting does so for the whole card, or am i wrong?

and Soviet its late! LOL

 

Yes, you are indeed correct, mk1freak. It is late for you but great for me. I'm falling a bit behind on my smart*ss quota and you presented me with a well exploitable statement.

 

sorry OP dont mean to thread jack,

soviet, i cant wait till its late for you

 

I'm not worried. You're not the only one that stalks my posts to get revenge.

 

@Harkonnen

At idle, run GPU-Z, go to the second tab, and when the temps are rising to the 80s, tell us the shaders/memory frequency that GPU-Z is detecting. I'm suspecting that you're GPU is either stuck in Extra mode or switches to it when it doesn't need to