Re: [PATCH] acerhdf: Fix fan activation with new thermal governor

[Peter Feuerer <peter@piie.net>]

Hi,

sorry for not answering to earlier mails, but sitting in front of the 
computer and writing English mails is still very exhausting for me.


Borislav Petkov writes:

> On Mon, Feb 25, 2013 at 11:06:40AM +0800, Zhang Rui wrote:
>> when the temperature goes up to 50C, will the fan be turned on or not?
> 
> Right, let's first clarify this:
> 
> so we want the fan to go on when it reaches and/or exceeds the 'fanon'
> temperature and to go off after it goes under the 'fanoff' temperature.
> It is supposed to remain on in the [fanoff,fanon] interval.
> 
> Now, my observation from yesterday was that, if we don't do the
> flattening of the abovementioned interval in acerhdf_set_cur_state() but
> we simply turn on the fan if we receive state=1 and turn it off when
> state=0, the fan gets turned on only for short bursts when the governor
> decides it is time to throttle, i.e. the temperature in the zone is
> rising and turns it off when the temperature starts to fall again. And
> it would turn on the fan at lower temperatures than fanon.
> 
> But this is annoying. The main reason why this acerhdf thing was added
> in the first place was to do a better fan control on those machines
> because the fans are noisy. Basically, enable the fan only when really
> needed - and leave it off when not because it is really annoying.
> 
> When enabled, however, leave it on longer until a safe, 'fanoff'
> temperature is reached.
> 
> Am I making more sense now? Peter?

This sounds good, couldn't have stated it more clear. Just one thing to add 
here: This is standard idea of a "schmitt trigger", we implemented 
in acerhdf :)
http://en.wikipedia.org/wiki/Schmitt_trigger

-- 
kind regards,
--peter;