Hi Juergen, > > > -----Urspr?ngliche Nachricht----- > > Von: "Juerg Haefliger" <juergh at gmail.com> > > Gesendet: 08.09.07 15:41:35 > > An: "Juergen Bausa" <Juergen.Bausa at web.de> > > CC: lm-sensors at lm-sensors.org > > Betreff: Re: dme1737 module > > > > > The contents of the temperature registers match the output of sensors, > > so it's not a conversion/display issue. The dme1737 does advertise the > > measured temp as 16C. The only thing that's left is that the CPU diode > > is not a 3904-type diode and the dme1737 gets the calculation wrong. > > It wants to see a 3904 diode attached to its remote diode pins. If you > > want closure, you need to check the docs for your CPU. > > > > So the actual sensor is on the cpu and the processing is done on the board? Yes. > Is processing done by the bios or by dme1737? In the dme1737. > The type of my cpu is very new. At boot up, the bios displays "unknown > cpu type". So, this might be the reason why it does not know how to handle > the sensor output. Can you send the output of 'cat /proc/cpuinfo'? > If this is true, next bios revision might solve the problem. A BIOS upgrade might fix the display of the temp numbers on the BIOS config screen it will not affect the output of sensors. The Linux driver reads the temp directly from the dme1737, the BIOS is not involved at all. If it turns out that the temp diode in your CPU is not a 3904 type, the only 2 options are: 1) Tweak the dme1737 to recognize the diode type (not sure if that's possible, I have to look at the datasheet). 2) Come up with a conversion algorithm and put it into sensors.conf. ...juerg > Is this correct? > > Juergen > ______________________________________________________________________ > XXL-Speicher, PC-Virenschutz, Spartarife & mehr: Nur im WEB.DE Club! > Jetzt testen! http://produkte.web.de/club/?mc=021130 > >
