Page 1 of 1
Incorrect Temps E6300 Rev. L2
Posted: Mon Mar 19, 2007 5:01 am
by armourer
Hey guys,
Both intel TAT and Coretemp are reporting VERY high idle and load temps compared to speedfan and everest.
I'm running at vCore 1.325 and am using a lapped and AS5 mounted Tuniq Tower 120, which is why I'm concerned about the difference in temps.
http://img524.imageshack.us/img524/4991/untitledgg2.jpg
http://img254.imageshack.us/img254/8440 ... ed2vx7.jpg
I think this is due to Coretemp being fooled into setting Tjunction to 100, rather than 85, which I believe is correct for the L2 stepping of the E6300
However, because it is also an Allendale, like the E4300,
the latest .95 update to coretemp may have caused the problem.
Please advise,
Evan
Posted: Mon Mar 19, 2007 10:02 am
by The Coolest
There are some problems detecting the Tjunction properly on these L2 CPUs. Most of them most likely have Tjunction of 100C while some may still have 85C, or you may have a very bad mount. I don't know.
I'm investigating this issue and will update when I have more information.
Thanks for the head's up
Posted: Mon Mar 19, 2007 11:35 pm
by armourer
I really appreciate your help,
you guys do great code.
Thanks again,
Evan
Posted: Tue Mar 20, 2007 3:03 am
by eric256
I get the same difference between 0.95 and 0.94
Mine is a 6300 L2 SL9TA
Tjunction is 100c in 0.95
Tjunction is 85c in 0.94
But 0.95 gets me closer to TAT.
My temps at load are:
TAT === 65/70
Speedfan 4.32 === 52/57
RMClock 2.2 === 39/45
Everest 3.80.948 === 52/57
Core Temp 0.94 === 52/57
Core Temp 0.95 === 65/70
Maybe all the differences are because of the Tjunction calculation ? RMClock is way off, but what is right ??
Posted: Tue Mar 20, 2007 5:51 am
by armourer
Hey Eric,
I think both TAT and Coretemp are incorrect.
I believe they both incorrectly place the tjunction at 100, therefore offsetting the real temps by 15 degrees.
I have a lapped CPU running with AS5 and a tuniq tower.
Idles that high are fairly unlikely.
~E
Posted: Tue Mar 20, 2007 10:50 pm
by eric256
That is what I tend to believe at this point. Strange that RMClock was way low too.
I have a Big Typhoon - and lapped both cooler and CPU. I really doubt that my TAT temps are accurate. Course TAT was developed for laptops anyway.
Now I just need to get back to CoreTemp 0.94 - but the download link is gone.
BTW - my temps are all load - not idle/load. I just listed load Core0/Core1.
...
Posted: Wed Mar 21, 2007 1:14 am
by BlkR
ey guys... ive just swich back my cpu from e4300 to e6300 and core temp on both versions are reading tjunction as 85C.... core temp 0.95, 0.94, speedfan and TAT are all reading exactly the same temp so its fine for me...
Posted: Wed Mar 21, 2007 2:26 am
by armourer
you have a B2 stepping I presume?
..
Posted: Wed Mar 21, 2007 8:56 am
by BlkR
.oh sorry yes i do
Re: Incorrect Temps E6300 Rev. L2
Posted: Thu Apr 05, 2007 1:04 am
by SLi_dog
armourer wrote:I think this is due to Coretemp being fooled into setting Tjunction to 100, rather than 85, which I believe is correct for the L2 stepping of the E6300
However, because it is also an Allendale, like the E4300,
the latest .95 update to coretemp may have caused the problem.
I suppose you have to ask yourself what's more believable?
An idle temp of 40C or an idle temp at or below ambient case temperature?
I know the Tuniq is good, but it's not
that good.

Posted: Mon Apr 23, 2007 9:27 am
by drkuli
IT IS that good, Scythe Infinity/Ninja/Mine and some more coolers also. And I didn't observe temps BELOW ambient on any of E4300/E6300. Instead I can observe ambient with very precise lab digital thermometer and it's ALWAYS lower than core 0/1, CPU/SYS (whatever it means) or HDDs through SMART.
Ambients HIGHER than any of these temps are erroneous because of the measuring method of ambients, not the rest of temps. It is that simple.
BTW, one of my HDDs IS LOWER than ambient and it's perfectly understandable. This HDD is in closed case, in direct air stream of 120mm intake fan, distant no less than one inch from it. Thermometer probe is in free field, about 4 inches from that fan.
Posted: Fri May 04, 2007 6:42 am
by SLi_dog
drkuli wrote:IT IS that good, Scythe Infinity/Ninja/Mine and some more coolers also. And I didn't observe temps BELOW ambient on any of E4300/E6300. Instead I can observe ambient with very precise lab digital thermometer and it's ALWAYS lower than core 0/1, CPU/SYS (whatever it means) or HDDs through SMART.
Regardless of how good the Tuniq is there is no way that an air cooler can keep a CPU at or below ambient temperature while it's running.
People where I am are seeing idle temps below ambient temperatures with E4300 CPUs which is just not possible. So far below that the error margin of a digital thermometer can't explain it.
Since the E4300, E4400, E6300 L2 and E6400 L2 use a different core to the rest of the Core2Duo line up, why is it so hard to believe that they may run hotter?
Also, the Scythe Infinity is also a very good cooler but the Ninja isn't in the same league.
Just configure CoreTemp v0.95 to read the pure DTS readout (Delta to Tjunction) and that should tell you all you need to know regardless of the degreesC value.
Posted: Thu May 17, 2007 12:21 am
by obvious
The Coolest wrote:There are some problems detecting the Tjunction properly on these L2 CPUs. Most of them most likely have Tjunction of 100C while some may still have 85C, or you may have a very bad mount. I don't know.
I'm investigating this issue and will update when I have more information.
Thanks for the head's up
I'm running an L2 E4400. How's your research coming along?
Thanks.