Re: [ACPI-sppt] fan speed control techniques (again)

On Wed, Jul 09, 2003 at 08:53:33AM -0300, Derek Broughton wrote:
> From: "Emma Jane Hogbin" <[email protected]>
> > http://www.analog.com/library/analogDialogue/archives/34-04/fan/
> 
> I haven't read it thoroughly yet, but I don't believe it said that.  It's
> showing a specific implementation which turns on the fan at 60.  It also shows
> one that turns on the fan at 50.  When the fan should turn on is determined by
> each manufacturer based on numerous conditions.  The specific capabilities of
> the CPU probably being the most important.

*nod* It's possible I misunderstood it the first time--what I did pick
up is that there are magic numbers at which point the fan should be
on/off. The magic numbers depend on what the manufacturer has set. 
There are two (sets of) numbers which are used in either active or
passive cooling modes. Active cooling modes use the fan; passive cooling
modes use CPU throttling--which means the CPU slows down the speed at
which it works: the slower the machine, the cooler it will run. (I think
I've got all of that right.)

> Not imo.  I wasn't sure what Pavel meant by "Fix thermal management", either,
> but the R/W was specifically to allow someone to override the trip points.  I'm
> not sure why that _should_ be included in Linux ACPI implementation, since it's
> an open invitation to destroy your machine, but then we Linux geeks are supposed
> to be smart enough to be trusted with such awesome power :-)

Heh. I'm not sure I want to change my trip points. As far as I can see
they are VERY high (80C and 92C or thereabouts). There must be another
number that is much lower (perhaps 45C) where the fan decides it needs to
be on. I can't find a number that says anything close to that though--so
I can't figure out why the fan goes on.

> I don't think there's anything wrong with the fan itself.  Does it really come
> on when the thermal trip points say it should?  Then the acpi implementation
> works :-)  Whether the laptop maker has done a good job of defining the trip
> points is another matter, but I'd rather trust them than risk my machine.

Agreed there's nothing wrong with the fan--it's a symptom of a problem,
but I don't believe it is the problem. 

If the thermal trip point is:
debian:/proc/acpi/thermal_zone/THRC# cat trip_points 
critical (S5):           100 C
passive:                 86 C: tc1=2 tc2=5 tsp=300 devices=0xc12c5a40 

Why is my fan on when it's only:
debian:/proc/acpi/thermal_zone/THRC# cat temperature 
temperature:             42 C

thanks,
emma

-- 
Emma Jane Hogbin
[[ 416 417 2868 ][ www.xtrinsic.com ]]


-------------------------------------------------------
This SF.Net email sponsored by: Parasoft
Error proof Web apps, automate testing & more.
Download & eval WebKing and get a free book.
www.parasoft.com/bulletproofapps