• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

E2180 reporting wacked temperatures.

A

AmongTheChosenX

Supreme [H]ardness
2FA
Joined
Sep 24, 2007
Messages
7,299
Alright, so now my PC is running again. (specs in sig)

but it seems that this new board Asus sent me is doing wierd things about temperature.

The BIOS reports that at stock speed its idling at 30 degrees, like 10C higher than it used to say.

I tried Asus AI Suite, and that reports that its idling at about 30 as well (dips to 26, and rises to 34 at times).

under load, this is where i'm concerned. with no voltage increases, and OC'd to 2.8 ghz (lower than i had it before), coretemp reports 68C ON BOTH CORES. Speedfan reports the same, with a CPU temperature reading of 58C! AI suite is also reading 58C!
however its passing prime95 totally stable with 100 percent load on both cores...

my old board used to tell me 44C under load, OC'd to 3 ghz. it took 1.525 volts to get this thing to even close 60C. I'm running 1.3 and its telling me 58C. somethings not right.

but is it safe to assume the software/BIOS are wrong? maybe they effed up the sensors on this RMA'd board?

does anyone ahve an idea? I tried BIOS 1004 (final), and i''m currently on 0602. I have Tuniq TX-2 paste (same as before), and i originally had Diamond thermal paste on it that i got from microcenter, which actually brought these numbers i gave down by like 3-5 degrees at idle.
 
Have you tried speedfan or coretemp? I've seen some software report temps 15 degrees high with E2180s. IIRC Coretemp reports 15 degrees high, but speedfan is on point.
 
yes, as i mentioned in the first post, i tred speedfan, coretemp, and AI suite.


Coretemp was reporting 85C, so yes that is 15C too high, but then again 60C is not what i should be at.

and another wierd thing, my E2180 suddenly requires an extra .03 volts over what i used to give it to hit 3ghz... idk if that means anything.
 
i think it's either the board or this processor combined with this board, because my old board used to do this thing where if I had the processor OC'd to say 333 x 9 for 3 ghz, any program would report that it was 333 x 10, and thus running at 3333 mhz, which it was not. when i had it at 395 x 8, it would tell me the processor was running at like 4 ghz, which it definitly was not.

so maybe the processors bugged.

EDIT: flasshed to 1013 BIOS, OC's are normal now, but not it says my CPU when clocked at 333 x 9 says its at 85C, which is definitly wrong. i looked at some notes i took down when I OC'd with the previous RMA'd board, and i wrote that it ran at 43C, so i'm assuming the thermometers of the board are 40C off, give or take a degree...
 
I think you mean 70c (85-15=70) ;). And the more voltage thing is understandable. No two board will be exactly the same in how they react to a given CPU/other hardware. I had to bump my CPU volts up another notch too after frying IP35 Pro #1, and shelling out another $150 for a replacement.

I would say the newer (?) BIOS might be the culprit, but that wouldn't explain CoreTemp's readings. Have you tried RealTemp (similar to CT)? I doubt it will be any different, but it's worth a try. And it's SpeedFan that's off by 15c (too low), CT/RT are the ones you want to pay attention to. Some say the delta is a bit different, ie: 10c or so, but SF (unless you specified an offset in the advanced options) is off. You may be right and the sensors (or calibrations) could plain be off on the board - if you've tried multiple times remounting, multiple reapps of paste, and multiple programs, the culprits are few. Do you have a spare HSF around to make *sure* it's not the Gemini II that's the problem here?

Oh, and from what I understand of the situation, NO program will truly read the temps right. Everybody is taking a stab in the dark, whether we're talking CT/RT/SF/else.. Intel, from what I read, doesn't publish all the data needed for (proper) temp monitoring for desktop CPUs (only notebook?), and as such people have to make an educated guess (at the actual TjMax rating as I understand is the main issue). Ultimately, unless you are experiencing instability or notice the system throttling due to an over temp situation, I wouldn't worry *that* much.

[edit] And how is your case cooling? The Gemini II is one big cooler and while not the best (certainly for <$10 AR it's a great BFTB), it should be ok.. But there's only so much it can do if you have very little airflow through the case.
 
Touch the heatsink while its loaded up. If it feels cold, then you dont have proper contact between heatsink and chip.
 
no i touched it, its warm to the touch, but not hot. i remember how hot it got when i made the thing run 60C with 1.5 Vcore, and everything around the core was scorching. its not that hot, so its definitly not at 80C.
 
I have the same processor and heatsink as you, though mine is still at stock. With a 19C ambient and a single 400RPM fan on the GeminII, it idles at about 23C and loads in the lower 30s. I don't have a solution for you, other than that you should try RealTemp.
 
yeah with my old board before I RMA'd thats how i idled...

i think the sensor is bad. I jsut ran the "test sensor" thing in realtemp, and it said i probably have a bad sensor as theres too much "movement".

so i'm gunna go with taht assumption and use my OC notes that i have for temperature references... since its all the same hardware anyway.

Raxxath, what are your core0 and core1 temps?
 
At the moment, both are 20C, but my side panel is off.

In Real Temp, the TjMax is listed as 85C. My sensor movement is 5-7C.
 
It turns out the TjMax for the e2180 should be 100C, so add 15C to my temperatures.
 
unclewebb said:
Where is that info from?

I tested an E2160 with an IR thermometer and TjMax is definitely only 85C. Does the 100C number come from someone that has done some actual testing or are they just guessing?

http://www.xtremesystems.org/forums/showpost.php?p=2844886&postcount=423

Here's the latest version of RealTemp.
http://www.techpowerup.com/realtemp/
Click to expand...

Agreed, I followed that thread and was debating that same thing on the coretemp forum a couple of months ago. 85c tjmax makes a lot more sense than 100c.
 
unclewebb said:
Where is that info from?

I tested an E2160 with an IR thermometer and TjMax is definitely only 85C. Does the 100C number come from someone that has done some actual testing or are they just guessing?

http://www.xtremesystems.org/forums/showpost.php?p=2844886&postcount=423

Here's the latest version of RealTemp.
http://www.techpowerup.com/realtemp/
Click to expand...

Its from here. From Gigabyte's post:

All M0 steppings which the E2180 is an M0 are 100C Tj max. Use CoreTemp 0.96 if you can find it, if you can't find it I can upload to FileFront later if you want. I had months of disscussions with the CoreTemp devloper and managed to get him to change M0 Tj max to 100C, since then he changed it back for an unknown reason and does not seem to believe any of the proof I show to him.

I know this is like XP vs Vista.. this battle for 85 or 100C Tj max for M0's has been going for a while now, but Tj max for M0 steppings is indeed 100C! So add 15C to those core temps.
Click to expand...

I'm inclined to believe him, simply because Real Temp was reporting that my E2180 was running at slightly sub-ambient temperatures, which we all know is impossible with a simple heatsink. I still love your program, though. :D
 
Raxxath said:
Its from here. From Gigabyte's post:



I'm inclined to believe him, simply because Real Temp was reporting that my E2180 was running at slightly sub-ambient temperatures, which we all know is impossible with a simple heatsink. I still love your program, though. :D
Click to expand...

Look at post number 73
http://www.overclock.net/intel-cpus/305028-tsk-s-crazy-theory-tjmax-dts-8.html

It shows that some cpu need idle calibration, I think unclewebb will be better suited to answer your question.

I personally have the older realtemp where there is a button for + / ++ and I use the + to adjust for my idle temps.
 
Doodies said:
Look at post number 73
http://www.overclock.net/intel-cpus/305028-tsk-s-crazy-theory-tjmax-dts-8.html

It shows that some cpu need idle calibration, I think unclewebb will be better suited to answer your question.

I personally have the older realtemp where there is a button for + / ++ and I use the + to adjust for my idle temps.
Click to expand...

You know, I read all of that a while ago and forgot about it, apparently. Maybe that's why my E6700's idle temps are higher than they should be?
 
You must log in or register to reply here.
Back
Top