[ltp] Fan-questions again...

[Benjamin Buch]

Am Mittwoch, den 10.01.2007, 17:45 +0100 schrieb Gilson Laurent: 
> Moien
> 
> > What I did to lower the noise:
> > 
> > -undervoltet Pentium M
> > -underclocked ATI x300 with fglrx
> 
> You forgot the WLAN-card. "iwconfig" will show you if the power-managment is set or not.

Thank you, I was not aware of that.
Is there a conf-file where you can set up the changes or do you have to
do it always manually with iwconfig?

> > -with the underclocking, the CPU-temperature usually doesn't rise above
> > 56 - 57° when not under load (with the fan disabled).
> 
> underclocked and undervolted (PentiumM 725, 600 MHz @ 0.7V) i get below 40°C in idle desktop and below 45°C for office/browsing. Are you sure you already undervolted your CPU ? Do you have a CeleronM or PentiumM ?

I tested again, and you are right.

Pentium M 750, 800 MHz @ 0,72V: 

idle - fan out (/proc/acpi/ibm/thermal)
42 38 30 43 28 -128 22 -128

office with fan on
42 39 29 43 28 -128 22 -128

office with fan off (after 1/2 45 minutes)
49 45 36 49 32 -128 25 -128

I think those temps are quite okay, but I'm no expert. Are there any
tables about wich part is allowed to get how hot?

I turned the fan on again after the 45 min off, it runs now for a good
ten mins, and the temps are 44 41 32 44 30 -128 26 -128.
It seems that the fan is not very effective, or is this slow tempchange
with fan on normal? (The blown out air is rather cold...)

What is the need to turn it on when not under load? 
There is only 7° difference in the temps with fan on, so why not turn it
off?
In "normal" mode (no underclocking, undevolting etc) the difference is
even smaller: the temps are 46 42 31 49 31 -128 26 -128 with
office/browsing and fan rotating with noisy 3257 RPM. 

> > -why are there different fan speeds within the same speed level?
> 
> Uhm... explain a bit more. Setting the fan to level X (via /proc/ibm/acpi/fan in experimental=1 mode) will always produce XYZ rpm.

If I change the speed level manually (ecdump or with tp_fancontrol), the
same level does not result in the exact same RPM. The range within level
1 is from ~3110 to ~3250 (du to my very rare observations).
I had the impression that the higher the speed was before the
level-change the lower the RPM were afterwards.

Thanks,

regards

Benjamin