From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Soeren D. Schulze" Date: Mon, 31 Oct 2011 07:46:43 +0000 Subject: Re: [lm-sensors] Something wrong with support for NCT6776F in Message-Id: <4EAE5263.3040309@gmx.de> List-Id: References: <4EA9CECA.4050007@gmx.de> In-Reply-To: <4EA9CECA.4050007@gmx.de> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable To: lm-sensors@vger.kernel.org On 31.10.2011 05:57, Guenter Roeck wrote: >> I have an ASRock H67M-ITX motherboard. >> >> The w83626ehf driver detects: >> >> [ 3904.608693] w83627ehf: Found NCT6776F chip at 0x290 >> >> "sensors" shows me: >> >> coretemp-isa-0000 >> Adapter: ISA adapter >> Physical id 0: +46.0=B0C (high =3D +80.0=B0C, crit =3D +98.0=B0C) >> Core 0: +46.0=B0C (high =3D +80.0=B0C, crit =3D +98.0=B0C) >> Core 1: +45.0=B0C (high =3D +80.0=B0C, crit =3D +98.0=B0C) >> Core 2: +46.0=B0C (high =3D +80.0=B0C, crit =3D +98.0=B0C) >> Core 3: +46.0=B0C (high =3D +80.0=B0C, crit =3D +98.0=B0C) >> >> nct6776-isa-0290 >> Adapter: ISA adapter >> Vcore: +0.96 V (min =3D +0.00 V, max =3D +1.74 V) >> in1: +0.18 V (min =3D +0.00 V, max =3D +0.00 V) ALARM >> AVCC: +3.39 V (min =3D +0.00 V, max =3D +0.00 V) ALARM >> +3.3V: +3.39 V (min =3D +0.00 V, max =3D +0.00 V) ALARM >> in4: +0.54 V (min =3D +0.00 V, max =3D +0.00 V) ALARM >> in5: +1.69 V (min =3D +0.00 V, max =3D +0.00 V) ALARM >> 3VSB: +3.41 V (min =3D +0.00 V, max =3D +0.00 V) ALARM >> Vbat: +3.31 V (min =3D +0.00 V, max =3D +0.00 V) ALARM >> fan1: 0 RPM (min =3D 0 RPM) ALARM >> fan2: 2180 RPM (min =3D 0 RPM) ALARM >> SYSTIN: +42.0=B0C (high =3D +0.0=B0C, hyst =3D +0.0=B0C) ALAR= M sensor >> =3D thermistor >> CPUTIN: +127.5=B0C (high =3D +80.0=B0C, hyst =3D +75.0=B0C) ALAR= M sensor >> =3D thermistor >> AUXTIN: -24.0=B0C (high =3D +80.0=B0C, hyst =3D +75.0=B0C) sens= or =3D >> thermistor >> PECI Agent 0: +46.5=B0C >> cpu0_vid: +2.050 V >> intrusion0: ALARM >> intrusion1: ALARM >> >> >> Here, CPUTIN and AUXTIN (whatever it is) are obviously wrong. CPUTIN >> *decreases* towards 17=B0C when the CPU gets hot (non-linearly, though). >> > Odd. No idea what is going on, though. The registers are quite straightfo= rward, > and the same for NCT6775F and NCT6776F. And it works on my system with NC= T6775F. Is there the option that auto-detection was wrong and I have a=20 completely different chip? >> This was the first problem. The second problem occurs when I try to >> control fan2 manually. The "pwm2" file in sysfs is 255 for anything >> above 2000 RPM, but as you can see above, the fan can do more (I've been >> at 2600 RPM -- maybe it can do even more), but pwm2 stays at 255. >> > Looking through the datasheet, I did not find anything obvious. It might > be possible that the PWM output frequency (register 2 bank 0) somehow aff= ects > the fan speed as well. What causes the fan speed to go higher in your sys= tem ? High temperature in Temperature Cruise mode. My firmware setup interface actually provides me with both the options=20 "Target fan speed" and "Target temperature", but I have the impression=20 that while the driver reports that I'm in Temperature Cruise mode, the=20 fan speed does not vary that much, but it rather tries to keep the fan=20 speed constant unless temperature is completely out of range. The fan speeds up only when I'm above 60=B0C, and even that with a certain = delay, so under full load, I get above 70=B0C when my fans are=20 firmware-controlled... That's basically the reason why I want software=20 control. _______________________________________________ lm-sensors mailing list lm-sensors@lm-sensors.org http://lists.lm-sensors.org/mailman/listinfo/lm-sensors