Abit AN-M2 w/Brisbane temperature sensor inaccurate?

All about them.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
displacedtexan
Posts: 3
Joined: Tue May 13, 2008 9:25 am
Location: California

Abit AN-M2 w/Brisbane temperature sensor inaccurate?

Post by displacedtexan » Sun May 25, 2008 7:39 am

After a bit of googling, it turns out that Abit AM2 motherboards don't properly measure CPU temps when used with Brisbane processors.

I am running an AN-M2 with a 4450e 45 watt processor and picoPSU with a Dell brick. This computer is running linux 24/7 as it is my MythTV frontend and backend.

Unfortunately, the temp sensors don't seem to work, and is apparently a known problem. I get the following output:

Core0 Temp:
-1A°C
Core0 Temp:
-5A°C
Core1 Temp:
+6A°C
Core1 Temp:
+3A°C

How are people measuring CPU temps with Abit motherboards and Brisbane processors?

Thanks!!

thejamppa
Posts: 3142
Joined: Mon Feb 26, 2007 9:20 am
Location: Missing in Finnish wilderness, howling to moon with wolf brethren and walking with brother bears
Contact:

Post by thejamppa » Sun May 25, 2008 9:01 am

Hi,

Yes, Brisbane core's temperature sensor is off by far and has caused problems for the peoples to get accurate measurement. There really is no accurate way but get temperature "gun" to measure temperature accurately. You could try guess how much off the temperature is and use that to calibrate SpeedFan or other program you use to measure temperatures to get it more right. I one point added +25C on current temp to get more normal readings.

Of course accuracy is the same as 16th century musket... Which is: Around there, something like there...

porkchop
Posts: 496
Joined: Thu May 25, 2006 1:19 am
Location: Australia

Post by porkchop » Sun May 25, 2008 9:43 pm

i find that the cpu reading in speedfan is pretty accurate.

currently my mobo is reading 17C core and 32C cpu(40-60% load [email protected]), touching the cpu heatsink- it feels barely warm which i think is about right.

the only way to really know is to measure the temp manually, and then do as thejamppa suggested to get the right number.

ryboto
Friend of SPCR
Posts: 1439
Joined: Tue Dec 14, 2004 4:06 pm
Location: New Hampshire, US
Contact:

Post by ryboto » Mon May 26, 2008 8:55 am

I just use the Abit software for monitoring a brisbane on an AN-M2. I just assumed it was correct. I run the CPU passive at 2.5ghz with 1.15v, and the Abit monitoring software sees it as hitting as high as 63C(1 fan in the case). Which seems reasonable to me.

porkchop
Posts: 496
Joined: Thu May 25, 2006 1:19 am
Location: Australia

Post by porkchop » Mon May 26, 2008 10:40 am

did a quick comparison between speedfan and abiteq and they display the same numbers for cpu temp- they increase and decrease at the same rate too.

my abiteq had a 'car' skin- pretty ugly looking...

ryboto
Friend of SPCR
Posts: 1439
Joined: Tue Dec 14, 2004 4:06 pm
Location: New Hampshire, US
Contact:

Post by ryboto » Mon May 26, 2008 11:51 am

porkchop wrote:
my abiteq had a 'car' skin- pretty ugly looking...
yea, it's definitely not minimalistic.

rouflaquettes
Posts: 1
Joined: Sat Jun 07, 2008 4:13 am
Location: France

Post by rouflaquettes » Sat Jun 07, 2008 4:29 am

Hi,
I have exactly the same problem with this mobo, which is annoying as my be-2350 is at 50°c, full charge (@3ghz): i'd be secured if i could get the right core temps. But i couldn't manage to. :(
Anyway, I would have another question for you: which temp would you take as the chipset temp: "AUX" or "System"?
Idle, mine are(°c):
CPU/32
AUX/46
System/36
So my guess would be AUX->chipset, System->motherboard ambiant temp.
Your opinion? What would be a correct chipset temp?
THX

Post Reply