Hello! Thanks for the reply.
Hi Simon,
Which motherboard is it? On some boards passing module parameter
On Tue, 16 Jul 2013 19:04:39 +0200, Simon Sandström wrote:
> I'm having trouble with the asus_atk0110 kernel module. The speed for fan3
> (CHASSI2 FAN) often just shows zero, but once in awhile it shows the actual
> speed of the fan. I've checked the fan, and it does spin at all time and it
> also shows the correct speed at all time in BIOS.
new_if=1 helps, you may want to give it a try.
Also, the asus_atk0110 driver heavily relies on the BIOS, so a BIOS
update may help.
Asus P5K Premium. I'm pretty sure it has the latest BIOS version (can't check now, see below). I tried using the option new_if=1 and it just screwed up all readings. For example it now shows 1.54V for all voltages, 1500~RPM for all fans and 150°C for both CPU and MB, which is way off.
Note that the fact that the speed in the BIOS is right doesn't
necessarily mean that the BIOS is right and the asus_atk0110 Linux
kernel driver is wrong. If the fan has any form of automatic speed
control applied, it might simply be that the BIOS keeps your machine
busy enough for the fan to stay in the measurable speed area. Did you
enable fan speed control in the BIOS? Does the problem go away if you
(temporarily) disable it?
For 3-pin fans, speed measurement is easy and reliable when the fan
runs at full speed. However as soon as control is applied, measurement
becomes harder. This is true when PWM control is used, but even more
when DC control is used: the speed signal weakens as the voltage is
lowered, and after some point it's difficult for the monitoring chip to
distinguish between the speed signal and the noise.
So maybe your fan is spinning at the border between readable and
unreadable.
Yes, I have fan control enabled both for CPU and chassi fans. I can not disable it at the moment as the server is running some services which needs to be up. CPU is set to automatic and the chassi fans is set to run at 70% duty. Both chassi fans are the same model, but only one of them is giving strange readings with lm-sensors. So I don't think there is a problem because of the speed control.
I will try to disable the fan control and also try to switch the connectionof the two chassi fans to see if I can come up with anything.
Not necessarily surprising. I don't know how exactly the Asus BIOS code
> There is also a slight delay when reading the input with "cat
> /sys/class/hwmon/hwmon0/fan3_input". When the speed is zero, it takes
> approx. 0.3s and when it shows the correct speed (700RPM) it takes 0.1s.
> All other fan inputs respond with no delay (0.001s).
works, but I can imagine that they try harder when the fan speed they
read at first looks suspicious. This would explain the extra delay.
> (...)
> Does anyone else have this problem, or any similar problem? I could justHow do you keep them at exactly 700 RPM?
> connect the fan to fan4 and ignore fan3, but for some reason the minimum
> speed for fan4 is 800RPM (the others are 600RPM), and I like my fans
> running at 700RPM..
See above.
--
Jean Delvare
Simon
_______________________________________________ lm-sensors mailing list lm-sensors@xxxxxxxxxxxxxx http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
