Only 2 Deg rise 2500+ under load?

Cooling Processors quietly

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
marc999
*Lifetime Patron*
Posts: 230
Joined: Sat Oct 26, 2002 4:05 pm
Location: Calgary, Alberta, Canada

Only 2 Deg rise 2500+ under load?

Post by marc999 » Mon Jan 12, 2004 12:38 pm

Broke down this weekend and bought an XP 2500+ cause the price was just too good. :D

With it I got the 7000A AlCu and ASUS A7N8X-X and when testing it shows only a 2 Degree difference between Idle and full load (tested for like 5 hours just to be sure). Sounds kinda weird to me. Is this normal or is something wrong with my temp reporting (using MBM5)?

P.S. - Shows 48 Degrees Celcius under full load with Zalman at L setting on Termaltake H/M/L switch. Don't know what L is, probably somewhere like 5V or so.

PretzelB
Posts: 513
Joined: Tue Feb 11, 2003 6:53 am
Location: Frisco, TX

Post by PretzelB » Mon Jan 12, 2004 12:54 pm

That does sound too good to be true. Are you sure you're getting full load? I like to use Folding@Home since it does a good job at putting the cpu at 100% all the time.

You might want to do a search at the amdmb.com forums and see if others have reported strange temps with the board.

marc999
*Lifetime Patron*
Posts: 230
Joined: Sat Oct 26, 2002 4:05 pm
Location: Calgary, Alberta, Canada

Post by marc999 » Mon Jan 12, 2004 2:06 pm

I tested with both CPUBurn and Prime95. Same temps.
I will probably try the ASUS temp reporting utility tonight and see if it reports the same temps as MBM5.

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Mon Jan 12, 2004 2:18 pm

I get similar results. My 2500 is overclocked to 2.2ghz (~amd rating 3050) and my idle temp is 46-47, load 51-52.

fyi, i have a 7000a-cu on an asus a7n8x-e deluxe. I think it may be a characteristic of the 7000a cooler? One possible explanation could be that the bottom part of the hsf, being more dense, holds the heat energy very well, and this heat energy only gets dissipated when it reaches the outer fins? *shrug*

marc999
*Lifetime Patron*
Posts: 230
Joined: Sat Oct 26, 2002 4:05 pm
Location: Calgary, Alberta, Canada

Post by marc999 » Mon Jan 12, 2004 4:05 pm

Thanks for the info chylld. It's helpful to know someone else is getting similar results. Just for the record, are you using MBM5, the included ASUS program, or something else to measure the CPU temp?

BTW, did you have to bump up the Vcore to OC?

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Mon Jan 12, 2004 4:43 pm

mbm5.

nup vcore is at stock 1.65v

Harry Azol
Posts: 228
Joined: Fri May 02, 2003 9:21 am

Post by Harry Azol » Mon Jan 12, 2004 4:57 pm

I have a similar setup with similar results.. :)

a7n8x-x with xp1700+ @ 2.2ghz (1.525 volt vs 1.5v default), and 7000A-ALCU.. the temperature does not fluctuate very much between load/idle

marc999
*Lifetime Patron*
Posts: 230
Joined: Sat Oct 26, 2002 4:05 pm
Location: Calgary, Alberta, Canada

Post by marc999 » Tue Jan 13, 2004 7:38 am

OK, well I tried the included ASUS program and got very different results (sort of). The temperatures I quoted were from MBM5, sensor W837785TS-S.

The ASUS program shows a lower temp. When overclocked to 2200Mhz, it started at about 27-28 and rose to 38 degrees. These temperatures coincide with the MBM5 sensor ASUS1. So I just assumed that W837785TS-S was the CPU because it was higher, but it must not be because the ASUS program can't be set up incorrectly (it knows what your MOBO is). So I guess the question is, is what the heck is W837785TS-S measuring on MBM5 if it's not the CPU? :shock:

Unfortunately, while testing last night, when OC'd to 2200 Mhz on stock voltage Prime95 gave me errors after 1 hour and 53 mintues. I think this may be a power supply issue though because I think I remember seeing 12.48V on the the 12V line, which is 4% off. Who knows, only further testing will tell.

Harry Azol
Posts: 228
Joined: Fri May 02, 2003 9:21 am

Post by Harry Azol » Tue Jan 13, 2004 8:56 am

AFAIK, the asus program uses the thermistor underneathe the socket for cpu temps..

in MBM5 you can select the diode on the cpu to get a more accurate temperature (will be higher)

marc999
*Lifetime Patron*
Posts: 230
Joined: Sat Oct 26, 2002 4:05 pm
Location: Calgary, Alberta, Canada

Post by marc999 » Tue Jan 13, 2004 9:20 am

Mmmmm, are you sure? I thought that only P4s had a temperature diode in the CPU (and thus all Athlon XP temperature measurements relied on a thermistor underneath the socket). :?:

randalee
Posts: 34
Joined: Fri Jul 11, 2003 9:57 am
Location: Salt Lake City, UT
Contact:

Post by randalee » Tue Jan 13, 2004 9:35 am

The XP CPUs definitely DO have a thermal diode built-in to the chip. I believe the older Thunderbird CPUs are the ones that didn't have a diode.

Randy

nbac
Patron of SPCR
Posts: 142
Joined: Sat Jul 05, 2003 11:27 am
Location: Sweden

Post by nbac » Tue Jan 13, 2004 11:25 am

ASUS are infamous for their inaccurate temperature readings.
If I want to lower my CPU temps 10'C, all I have to do is Flash
the BIOS... :roll:

tkortens
Friend of SPCR
Posts: 18
Joined: Thu Apr 17, 2003 6:26 pm

ASUS temps

Post by tkortens » Tue Jan 13, 2004 11:43 am

I am intrigued by the comments re ASUS CPU temp calibration errors.

I'm cooling a P2.4C with a Zalman 7000Cu and both MBM5 and Asus probe only show a 10C rise from normal use (29C) to full load with a CPU burn-in program at 100% CPU load (39C stable). Idle temps are case ambient - according to Asus probe and MBM5.

Regardless of the actual temps I assume that the rise in temp is relative, i.e. even if the initial temp was 39C instead of 29C I am still actually only getting a 10C rise.

Point is that I find the Z 7000 to be an awesome performer. (I admit I am using a good sized Global-Win server case - but with AcoustiPack).

marc999
*Lifetime Patron*
Posts: 230
Joined: Sat Oct 26, 2002 4:05 pm
Location: Calgary, Alberta, Canada

Post by marc999 » Tue Jan 13, 2004 2:47 pm

chylld,
What setting are you using in MBM5 for CPU temps?
W837785TS-S or ASUS1?

I guess it's possible that W837785TS-S is the CPU diode and ASUS1 a thermistor. But if that's the case, then that's pretty incompetent of ASUS to set up their temp monitoring software to measure the thermistor and not the diode. :!:

Post Reply