[ltp] Bad fan behaviour in X61s caused by GPU temperatur sensor
Fri, 04 Dec 2009 09:33:01 +0100
This is an OpenPGP/MIME signed message (RFC 2440 and 3156)
Content-Type: text/plain; charset=ISO-8859-1
> The bad thing is, that the temp4 sensor causes the fan to be nearly
> always running, unless I stop X. I already disabled compositing, and
> switched to 16 bit color depth, without success. The thing is, I just
> can't believe that mobile hardware has to be cooled by a fan at 40=B0C
> already. However, I also don't want to run any fan script control
> scripts, as I tend to trust the hardware vendor more than some script
> writers. :-)
I am also using a fan control script on the X61s. My conclusion when I
freshly got this laptop was that it is a design error. In my case I came
to the conclusion that the fan is triggered by the wireless temperature
while it is just to far away (physically) from the wireless card to cool
it. So it would stay on all the time and the wireless would still stay
at the trigger temperature.
That made me conclude I could simply reduce that trigger temperature at
the wireless. (Now if the BIOS was open source ...:) )
Then there is this issue that it seems to be much better when running
windows (fan actually turns off again!). But well this is not an option
in any case. I think there could be better powermanagement for gpu and
wireless because lenovo made those drivers for windows. At least that
was my conclusion 2 years ago.
Its all proprietary, so maybe lenovo has bugs in the BIOS part
controlling the fan, which they work around inside the windows driver in
the way I sketched in the first paragraph? So, when you trust your
hardware vendor more than some script writers keep in mind that the
hardware vendor maybe solved hardware problems in proprietary software
(bios, windows drivers). I use the same 'tpfancontrol' script I
downloaded and read 2 years ago. And the kernel watchdog assures that I
have a cool thinkpad, should it ever fail!
> So, is this normal behaviour? Any hints how to avoid the fan to be
> running that often?
The fundamental theorem of algebra is open source. Like any other
mathematical theorem it can be applied free of charge and everybody
has access to its proof and can convince himself how it works. Why
should software be any different?
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename="signature.asc"
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
-----END PGP SIGNATURE-----