[ltp] Re: fan problem on R51

[Thomas Breitner]

Am Mittwoch, den 03.11.2004, 16:44 +0100 schrieb Moritz Karbach:
> Hi,
>=20
> > which version of XFree do you use?
> > I have debian sarge's 4.3.0.dfsg.1-8 running, with fglrx 3.14.1.
>=20
> It's the same here. But, as I wrote, I don't use the fglrx driver. First =
I=20
> did, but finally all these dependency problems pissed me off so I returne=
d to=20
> the radeon driver of the XFree project.

Here it's fairly easy to use ati's fglrx with debian: just follow the
instructions on
http://xoomer.virgilio.it/flavio.stanchina/debian/fglrx-installer.html

>=20
> > same here, cpufreqd manage the clock-rate of my cpu, and while on
> > battery it's "hard-wired" to 600 mhz.
>=20
> Do you think it makes any difference wether ac is connected or not? Eg th=
ings=20
> like "if I'm connected, I switch the fan on in any case"?

Don't know exactly what you mean, but in my case it makes no differenz
if I'm on ac or on battery: the fan *is controlled* by the temperature,
which is directly affected by the cpu-clock-rate. I even limit the
cpu-clock to about 1200 or 1400mhz on *normal* operation (on ac),
because of the fan. only when the systems requests more speed, it
switches to 1800mhz, and after the work is done it switches back.

>=20
> > same here, but also doesn't know what's meant by these different
> > C-states.
> > hm, here it's very human-readable explained:
> > http://acpi.sourceforge.net/documentation/processor.html
>=20
> Thanks for the interesting link. Now it seems, that these C-states should=
n't=20
> have much to do with the problem...
>=20
> > > - my CPU temperature is 32=C2=B0C
> >
> > mine ist - at the moment (2h online, constantly 10% cpu) - at 48=C2=B0C=
. but
> > it *seems* the fan is off.
>=20
> Well then, if you get close to the air slots, you should hear the fan if =
it's=20
> running :-)
>=20
> > perhaps a newer bios-version? ibm deliveres some kind of minimal
> > changelog with the bios-revisions (check online here:
> > http://www.tpdrivers.com)
>=20
> That's a thing I could try. I knew it was good to leave windows in a corn=
er of=20
> my system...
>=20
> Btw, how long is your uptime when running on battery? I reach about 2h30m=
in,=20
> roughly half an hour less then they promised... I obtained this value wit=
h=20
> tunable settings, of course [lowest display brightness, wlan driver unloa=
ded,=20
> cpu running on 1500Mhz, no throttling, no DynamicClocks], but do you thin=
k=20
> this gains an extra 30 mins?

last time on battery I *measured* the following:

uptime	equals	percentage of voll battery
__________________________________________
2:39 h 		50%
3:39 h 		33%
4:10 h 		25%

... then I plugged ac ...
btw: it's one of these *high capacity battery-packs* (9-cell or
something like that).

thomas

>=20
> Cheers,
>=20
> - Moritz
--=20
_____________________________________________
Thomas Breitner

mail@tombreit.de   |   http://www.tombreit.de