> On Wednesday, February 19, 2014 9:10 PM, Jean Delvare <jdelvare@xxxxxxx> wrote: > > It turns out that a very frequent pair of scaling resistors is 10 Ohm > and 56 Ohm, which leads to: > > compute in6 @*(1+56/10), @/(1+56/10) > > This maps your raw reading (1.896) to exactly 12.513 which is one of > the values you saw in the BIOS, and also maps integer register values > to 12.091 and 12.460. > Ah, I see. Thank you for that input. Now it maps exactly with the BIOS values that I have observed. > You must have changed the wrong file, or the wrong section in the right > file. It should work. > Yes, it works now. I'm almost ready with my configuration. Now here's my +5V scaling computation: +5V BIOS samples: 5.040, 5.064, 5.136 5.064 - 5.040 = 0.024 5.136 - 5.064 = 0.072 5.136 / (24/8) = 1.712 1.712 is close enough to in5 (1.688). So my compute statement is: compute in5 @*3, @/3 Apparently, the scaling factor I use above maps exactly with the BIOS values I have seen. So if everything checks out and you concur with my computation, then I can send in my configuration. _______________________________________________ lm-sensors mailing list lm-sensors@xxxxxxxxxxxxxx http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
