Weird temp issue, please advise.

Samury

Limp Gawd
Joined
Feb 7, 2006
Messages
199
I don't know if this is the right place to post this issue, but it's about cooling in some way, so here goes:

I woke up today to find my computer, which had been DLing something all night and doing nothing else in a room in the mid 60s F, but Hardware Monitor and ASUS Thermal Radar both showed the CPU and System temp as 123 C(!). I quickly felt the air coming out of the top two fans and it felt fine. I powered down anyway and removed the door to feel inside. Nothing. It felt fine. I touched just about everything I thought I should and nothing was even warm: the memory, the various heatsinks around the CPU, the H80 block. All cool to the touch.

Any ideas? Right now HWmonitor shows a SYS temp of 29, CPU 60 (the cores are in the 20s); Thermal Radar reads 24 CPU, 29 MB. Is this a faulty sensor? Should I update my bios if possible? Wait for an update to HWmonitor? I'm obviously not going to OC anything until I get this figured out.

So here's another mystery to solve.
 
Well now it's just gotten weird. When I've used the comp, HWmonitor has lowered the CPUTIN temp to 19C (the cores are still in the 20s). Meanwhile Thermal Radar is at 25 for the CPU.

I should note that I just played a game, then reviewed the temps. So when I actually use the machine HWmonitor drops the temps, but when I don't, they skyrocket. Strange.
 
It happened again, while DLing my Steam holiday purchases. I wasn't doing anything else but reading Sports Illustrated when my temps jumped to 123C for the SYSTIN and CPUTIN in HWmonitor and the ASUS monitors. But after a shutdown and restart everything was fine, in a minute or two. I guess it's time to talk to our trusty ASUS rep for some advice. The cores never broke 40C, or even came close. All other temps and voltages were standard at the time. Just weird.
 
Try not running more than one monitoring tool at a time. If you run multiple tools that poll the onboard Super IO chip, it will generate these types of errors.

-Raja
 
Those types of monitoring programs generally do not read the right temperature sensors. For checking your CPU temperatures, you should use something like Real Temp or Core Temp. You will likely find that your CPU is in reality far below 100C even at full load.
 
123C and -123C are the default numbers used for a dead sensor, non existant sensor, or sensor read error. but it more then likely sounds like a sensor reading error because you have to many programs trying to read it.
 
123C and -123C are the default numbers used for a dead sensor, non existant sensor, or sensor read error. but it more then likely sounds like a sensor reading error because you have to many programs trying to read it.

I'm trying to find a way around this. At the first time I just had HWmonitor and ThemalRadar (in the background, and didn't realize it). Now when I just have CoreTemp running I never have an issue. It's only when I also have HWmonitor also running to show the other values that temps start to rise. Maybe I just need to wait for a new HWmonitor update....
 
try using speedfan instead of HWmonitor see if that works. otherwise just ignore the temp readings.
 
Back
Top