CPU internal diode calibration

I have been having a problem with temp readings on my st6-celyt mobo combo that have left me pretty confused. It seems that for some reason I get consitantly higher temp readouts when compared to my other cely- mother boards. I have one abit vh6t, one Msi 6368 and one abit st6s. The first two are Via chip sets and have theristors located inside the cpu flip retainer near the center of the chip. The St6 is different though. either it does not have a theristor , it is surface mounted or there is none at all and the st6 reads the cpu internal diode temps. If this is the case then it explains everything because internal diod temps are always higher than any surface mounted theristor can read.

I did some research and found an article
that details how to calibrate your diode temps and I did it and it seemed to answer my problem according to the figures I got my cpu readings were about +7c. The question is does anyone know if the st6 mobo actualy reads temps from the internal diode and if not how come when I do the tests it seems to come almost exactly where I believe my temps are actualy at. My original load temps before calibration would consistanly read between 47c-51c but if I use the calibrated figures it would mean my actual cpu temps are more like 40c-44c full load (1.1a @ 1500 136fsb) 1.65vcore. I have a Vantec copper 6035/delta38 Hsf case open and system intake temps were 27c . I hope this is the case because it make everything I have been doing finally make more sense.

Davefred, I have read several of your posts with ol' man regarding the ST6 and CPU temps. I don't have an ST6, I have the VH6T and 1.1A Tually. But I was curious about what Tually temps should be, so I have followed the threads.

Any way, I found the following website which seems to say that the ST6 uses the cpu internal diode.

Quote - The board boasts three fan headers; more than sufficient for the comparatively cool-running Coppermine and Tualatin cores. Hardware monitoring is provided courtesy of a Winbond controller. The 83627HF provides monitoring of all fan speeds, all voltages, and makes use of Intel’s accurate on-die temperature diode.

Here is the link - www.hardwareanalysis.com/action/printarticle/1390/

Edward2 said:

Davefred, I have read several of your posts with ol' man regarding the ST6 and CPU temps. I don't have an ST6, I have the VH6T and 1.1A Tually. But I was curious about what Tually temps should be, so I have followed the threads.

Any way, I found the following website which seems to say that the ST6 uses the cpu internal diode.

Quote - The board boasts three fan headers; more than sufficient for the comparatively cool-running Coppermine and Tualatin cores. Hardware monitoring is provided courtesy of a Winbond controller. The 83627HF provides monitoring of all fan speeds, all voltages, and makes use of Intel’s accurate on-die temperature diode.

Here is the link - www.hardwareanalysis.com/action/printarticle/1390/

Click to expand...

Thank you very much that just about confirms my previous conclusions. This would explain everything. I can't wait for ol'man to see this. I wonder if he has ever calibrated his readings using the method described in my previos post. It only makes sense that I was getting good readings. They just can not be compared to on board theristor readings. My vh6t gives me much lower readouts but it is all relative. It also means that you must accuratly calibrate your diode temp readings or they will be misleading. Its all makeng sense now.