From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Ian Dobson" Date: Sun, 06 Feb 2011 10:00:03 +0000 Subject: Re: [lm-sensors] Standalone driver for W83677HG-I, NCT6775F, Message-Id: <373965DBFB2040D2A2318382AC5C8D46@portable2> List-Id: References: <20110205175852.GA26672@ericsson.com> In-Reply-To: <20110205175852.GA26672@ericsson.com> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8bit To: lm-sensors@vger.kernel.org -------------------------------------------------- From: "Guenter Roeck" Sent: Sunday, February 06, 2011 2:00 AM To: "Ian Dobson" Cc: Subject: Re: [lm-sensors] Standalone driver for W83677HG-I, NCT6775F, NCT6776F > Hi Ian, > On Sat, Feb 05, 2011 at 04:46:22PM -0500, Ian Dobson wrote: >> > [ ... ] >> >> Not bad, the results look almost the same as the standalone "hack" driver >> for the n6776 I'm working on, looks as if I can stop now. The only thing >> I'm >> seeing is that on fan headers where no fan is attached I'm seeing 164RPM. >> > Can you try again ? I updated the code to address the problem. > >> The NCT6776F seems to use a 13-bit counter for the fans. The RPM formula >> is >> RPM = 1.35e6 / count. The value 1.35e6 value might come from a 22.5 kHz >> counter. Now if the counter reaches the maximum value, which is 8191 for >> a >> 13-bit counter, we get exactly an RPM of 164.815041. >> >> nct6776-isa-0290 >> Adapter: ISA adapter >> in0: +1.18 V (min = +0.95 V, max = +1.26 V) >> in1: +1.02 V (min = +0.90 V, max = +1.20 V) >> in2: +3.36 V (min = +3.01 V, max = +3.50 V) >> in3: +3.34 V (min = +3.01 V, max = +3.50 V) >> in4: +1.03 V (min = +1.00 V, max = +1.10 V) >> in5: +2.04 V (min = +1.90 V, max = +2.00 V) ALARM >> in6: +0.80 V (min = +0.77 V, max = +0.90 V) >> in7: +3.42 V (min = +3.01 V, max = +3.50 V) >> in8: +3.30 V (min = +3.01 V, max = +3.50 V) >> fan1: 530 RPM (min = 0 RPM) ALARM >> fan2: 1030 RPM (min = 0 RPM) ALARM >> fan3: 164 RPM (min = 0 RPM) ALARM >> fan4: 784 RPM (min = 0 RPM) ALARM >> fan5: 164 RPM (min = 0 RPM) ALARM >> temp1: +23.0°C (high = +35.0°C, hyst = +37.0°C) sensor >> thermistor >> temp2: -60.0°C (high = +50.0°C, hyst = +55.0°C) sensor = diode > > This doesn't look good either. Can you report the raw values you get when > reading > registers 0x150/0x151, 0x75/0x76, 0x152, and 0x200 ? > > I have been wondering if I should retrieve the temperature values from > 0x73/0x75/0x77. > This is another set of temperature registers supported by both chips and > used as > fan monitoring source. 0x150 and 0x200 specify the temperature source for > 0x151 and > 0x75, respectively. > > Thanks, > Guenter There's also something wrong with the fan alarm limits (w83627ehf.c file from 05-Feb-2011 18:48):- SYSFANIN: 526 RPM (min = 28125 RPM) ALARM CPUFANIN: 994 RPM (min = 15340 RPM) ALARM AUXFANIN0: 0 RPM (min = 12980 RPM) ALARM AUXFANIN1: 769 RPM (min = 9642 RPM) AUXFANIN2: 0 RPM (min = 8035 RPM) ALARM The limits should be label fan1 "SYSFANIN" set fan1_min 400 label fan2 "CPUFANIN" set fan2_min 600 label fan3 "AUXFANIN0" set fan3_min 1000 label fan4 "AUXFANIN1" set fan4_min 600 label fan5 "AUXFANIN2" set fan5_min 1000 I'm also seeing this in the syslog [127250.935448] w83627ehf: Found NCT6776 chip at 0x290 [127433.043365] w83627ehf w83627ehf.656: fan4 clock divider changed from 1 to 16 [127433.043413] w83627ehf w83627ehf.656: fan5 clock divider changed from 1 to 8 [127829.468663] w83627ehf: Found NCT6776 chip at 0x290 [127837.241797] w83627ehf w83627ehf.656: fan4 clock divider changed from 1 to 16 [127837.241842] w83627ehf w83627ehf.656: fan5 clock divider changed from 1 to 8 Regards Ian Dobson _______________________________________________ lm-sensors mailing list lm-sensors@lm-sensors.org http://lists.lm-sensors.org/mailman/listinfo/lm-sensors