Tj max for E8200?

[dadi_oh]

So what is the correct tJ max for the E8200? It seems that the temp measuring SW is using the Tj max to calculate the core temperatures. See the attached screenshot.

Both HWMonitor and Coretemp think that Tj max is 105C. Realtemp thinks it is 95C. And the program readings of the cores are precisely 10C apart. Given that Realtemp thinks my idle temp is 12C while the other two think it is 22C. I would be much more inclined to believe 22C since it is certainly warmer than 12C in the room right now :)

[spearman914]

RealTemp is the "real" one. I think someone here on TPU did an experiment of what sensor monitoring program to believe.

[Duxx]

http://www.techpowerup.com/realtemp/

Check it out. Good read.

[dadi_oh]

RealTemp is the "real" one. I think someone here on TPU did an experiment of what sensor monitoring program to believe.

But that would fail to explain why Realtemp thinks my coretemps are 12C at idle. Without TEC coolers it is physically impossible for the cores to be cooler than room temperature. Maybe some calibration is required. I think that they had a process documented. Will need to look that up.

[dadi_oh]

But that would fail to explain why Realtemp thinks my coretemps are 12C at idle. Without TEC coolers it is physically impossible for the cores to be cooler than room temperature. Maybe some calibration is required. I think that they had a process documented. Will need to look that up.

Never mind. It seems that Intel does not publish Tj max which leaves us guessing. I found a good summary here (http://www.xtremesystems.org/forums/showpost.php?p=2995571&postcount=1124)

[ZenEffect]

But that would fail to explain why Realtemp thinks my coretemps are 12C at idle. Without TEC coolers it is physically impossible for the cores to be cooler than room temperature. Maybe some calibration is required. I think that they had a process documented. Will need to look that up.

sensors most likely are not calibrated properly.
realtemp derived its tjmax by using a ir thermometer on a cpu until throttling occurs.

[Dr. Spankenstein]

This (http://www.techpowerup.com/realtemp/docs.php) page can help you calibrate Real Temp. It really is a great temp program and only getting better.

[allen337]

just went there with wife last weekend ~~ http://www.tjmaxx.com/index.asp .

[dadi_oh]

This (http://www.techpowerup.com/realtemp/docs.php) page can help you calibrate Real Temp. It really is a great temp program and only getting better.

Thanks. I did the calibration procedure, setting my E8200 to 6X266 and voltage as low as my BIOS allowed 1.1V. In a room temp of 22C I was reading 12C on both cores. So assuming that my Watercooling setup is able to get Tj down close to ambient at such low clock and voltage settings I applied a correction factor of +2 to both cores which brings them to 22C (if I remember correctly... I think it is just correction factor X 5C). This was the opposite of what the Realtemp author had seen with his E8400 where he was seeing higher temps at idle than they should have been.

Under 100% load (Prime95 small FFT's) I am reading something like 38C on one core and about 44C on the other core. I took a screenshot but I don't have it with me. I still have a 6C to 8C difference between my cores which I am assuming is a sign of a poor IHS attachment. Also, my OC seems limited to about 3.7GHz and it always seems to be the hotter core that fails first in Prime95 above 3.7GHz. With a better IHS attach I might reach 4GHz but that is just speculation.

Still, I only paid $125 for my E8200 so getting it to 3.7GHz should make me happy. I'm just very demanding :)