On Sun, Jan 26, 2014 at 04:58:06PM +0100, Mathias Gerber wrote: > Hi Guenter > > On 24.01.2014 15:40, Guenter Roeck wrote: > > One interesting experiment might be to check what happens if you > > set pwm2_auto_point1_pwm to a value between 1 and 4. If that > > results in the fan running at full speed, we have some indication > > that either the fan or the board hardware is causing this > > behavior. > Same result. If the calculated pwm value (pwm2) is smaller than 5, the > fan spins up to maximum rpm. > > Today I disassembled my other pc (on which I'm able to completely stop > the fan by reducing the pwm value) with the same model of CPU fan > (Arctic Freezer 13) and hooked it up to the ASUS H87-PRO. The fan > looks a little bit different than the other, but the values are almost > the same. > > Arctic Cooling Freezer 13 (Old Model) > ------------------------------------- > > pwm2 rpm > 1 2064 > 2 2064 > 3 2064 > 4 2064 > 5 393 > 6 272 > 7 398 > 8 187 > 9 196 > 10 423 > 11 493 > 20 486 > 30 630 > 40 730 > 50 912 > 60 1060 > 70 633 > 90 859 > 128 1251 > 255 2261 > > Almost the same result with the original fan > > Arctic Cooling Freezer 13 (New Model) > ------------------------------------- > > pwm2 rpm > 1 2257 > 2 2253 > 3 2261 > 4 2180 > 5 551 > 6 451 > 7 407 > 8 370 > 9 366 > 10 370 > 11 392 > 12 411 > 13 433 > 60 1203 > 70 1355 > 90 1580 > 128 1836 > 255 2260 > > Then I took the boxed Intel fan and hooked it to the ASUS Board. > > Intel Boxed Cooler > ------------------ > > pwm2 rpm > 1 976 > 2 974 > 3 972 > 4 976 > 5 978 > 6 972 > 7 975 > 8 973 > 9 973 > 58 976 > 59 1005 > 60 1018 > 61 1021 > 70 1072 > 90 1184 > 128 1393 > 255 1953 > > And now the interesting part ;-) Connected the original fan to > CHA_FAN2 (controlled by pwm3) > > Arctic Cooling Freezer 13 (New Model) > ------------------------------------- > > pwm3 rpm > 1 0 > 2 0 > 3 0 > 4 0 > 5 0 > 6 0 > 50 0 > 55 430 > 60 512 > 70 633 > 90 859 > 128 1251 > 255 2261 > > On this pwm3 output I'm able to let the fan spin down until it stops > (around pwm 50-55) > > So it seems that the CPU fan output is somehow special - but I guess > only ASUS knows how exactly and why. Odd results. Guess there is nothing we can do about it. Thanks a lot for testing! Guenter _______________________________________________ lm-sensors mailing list lm-sensors@xxxxxxxxxxxxxx http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
