[lm-sensors] Reported fan speed oscillates 730 RPM -> 337500 RPM/675000 RPM

[simon]

Hi Jean, 
Thank you.

Your analysis appears absolutely correct. The fan is speed controlled
and problem is pronounced when it is running slowly. Normal speed is
around 750-800rpm. The problem appears to diminish as the speed rises;
above ~1000 rpm, the readings appear stable.
 
I would assume that Gigabyte have used the automatic control facilities
of the it8718 chip. Where is this going to be set up - from the bios? Is
there anywhere in the linux system that this can be done? Or is it all
automatic.

At any rate the thing seems to work, so I don't really want to play with
it (let sleeping dogs lie etc). It's just the oscillating reading is a
bit annoying. Wouldn't there be anyway of filtering out such values in
lm-sensors? (Maybe a configuration setting)

Many thanks,

Simon

On Sat, 2008-05-31 at 23:18 +0200, Jean Delvare wrote:
> Hi Simon,
> 
> On Wed, 28 May 2008 09:39:45 +0200, simon wrote:
> > The following output shows periodic output from sensors for my CPU
> > cooler over a short period.
> > 
> > The value continually oscillates between a real value (around 720-750
> > RPM in idle) and out-of-range values 337500 RPM/675000 RPM. When the
> > machine is loaded, and the fan spins faster, the fan value stabilizes at
> > a normal value.
> > 
> > The fan 2 sensor has no problems.
> 
> What's the usual speed value of fan2? Does its speed also change
> according to the temperature? The problem typically only shows for fans
> those speed is controlled and only for low speeds.
> 
> > Anyone have any ideas what's happening?
> > 
> > Configuration: 
> > lm-sensors v3.01 
> > Motherboard Gigabyte GA-MA78GM-SH2 
> > CPU Cooler: Thermaltake TMGA3 (3 wire)
> > 
> > Output from sensors over ~2 mins:
> > it8718-isa-0228
> > 
> > fan1:          726 RPM  (min =    0 RPM)
> > fan1:       675000 RPM  (min =    0 RPM)
> > fan1:       337500 RPM  (min =    0 RPM)
> > fan1:          738 RPM  (min =    0 RPM)
> > fan1:       675000 RPM  (min =    0 RPM)
> > fan1:          726 RPM  (min =    0 RPM)
> > fan1:       675000 RPM  (min =    0 RPM)
> > fan1:       337500 RPM  (min =    0 RPM)
> > fan1:       675000 RPM  (min =    0 RPM)
> > fan1:          725 RPM  (min =    0 RPM)
> > fan1:          725 RPM  (min =    0 RPM)
> > 
> > Many thanks for ideas!
> 
> This happens frequently when you control the speed of a 3-wire fan and
> the speed drops below a given limit. The speed is measured using a
> signal that comes from the fan, and when you control the fan, that
> signal gets weaker (think of it as proportional to the speed of the
> fan).
> 
> So apparently your fan is exactly at the limit below which the signal
> is too weak for reliable speed monitoring, which results in these fancy
> readings (which really mean that the IT8718F chip fails to sense the
> speed signal). If you can change the settings of the fan speed control
> so that the idle speed is slightly higher, the problem should go away.
> If the fan self-regulates then there's unfortunately not much you can
> do.
> 
> As a summary, 3-wire fans cannot always be controlled and monitored at
> the same time reliably. That's why we start seeing 4-wire fans, which
> do not have this problem.
>