On Mon, May 31, 2010 at 22:44, Camaleón
<noelamac@gmail.com> wrote:
On Mon, 31 May 2010 02:24:57 -0700, freeman wrote:
> Today my CPU seemingly jumped to 85' C and remained there without one
> change during three 15 min. sessions.
Did you check that values from BIOS or other sources?
(...)
> I guess I'll never know. How could a dust buildup cause a sudden change
> in the course of one session? I didn't see anything that seemed to have
> been sucked in all at once.
Bad heatsink or old fan?
> The Debian part is, could sensor reporting by ACPI, I8K (for Dell) and
> libsensors be dead-ended at 85' C?
"lmsensors" reads the values provided by the BIOS but can they be wrong
unless you load the right modules.
> I am thinking that physically removing and replacing the cooling unit
> maybe got me a lucky realignment of sensors or something. 85' C would
> just have been a default on failing.
As you changed "nothing" is quite strange, but I would just replace the
whole heatsink with a newer one, put a new layer of thermal paste and
check for any BIOS update.
> But the fan had to be in on the bad information too. Does it's
> information come from the kernel.
It comes from BIOS. You better check the BIOS values to reassure.
> BTW, this is a Pentium M, 1.6 GHz., which is suppose to handle heat
> well.
IIRC, Pentium M saga was not very "wattage hungry", I mean, it had a very
low TDP (<30W) :-?
Archive: http://lists.debian.org/pan.2010.05.31.17.14.27@gmail.com
I found a problem like that once with athlon x2, it used to go to more than 100C during some oridinary work. I opened the case, used a vacuum cleaner to clean everything thoroghly, and after that the cpu remained at 45C only.