Okay, so yesterday I installed my PC into a new case (Jeantech "Trinity") with an Cooler Master CoolDrive 4 (Aerogate II fan controller + hard drive bay), and I think the CPU's internal temperature sensor is acting weird.
The CoolDrive is set to knock all the fans up to maximum speed if the CPU or VGA sensors hits 65C, and if the case/HD sensors hit 55C. So, I thought I'd test it (Note to self: stop stress-testing everything).
Crank all the fans down to minimum, Prime95 the beast. 63C shows on the display, and the PC resets. Odd, I though, and tried again. 62C this time. Entering the BIOS (which, by the way, is a Leadtek nForce-2 board with the crappiest Health Monitoring features ever, and no voltage monitoring, argh), shows 68C for the CPU, which has been 46C up until now.
Thinking that's just weird, I crank all the fans up, and the temperature won't budge. It won't go up, either, it's just stuck permanently on 68C (even on a cold start). MBM5 (which doesn't like my board, really), is showing 80C. However, the CoolDrive CPU sensor, which is mounted next to the core, between the heatsink & susbstrate of the CPU, reads what very much appears to be the correct temperature.
Is my thermal diode broken? Did that sudden reset indicate that the board has had enough and reset to prevent damage to the CPU? I have remounted and reapplied decent thermal goop, which hasn't made any difference. There's no "Bad Temperature! Bad! Shutting down now!" feature in the BIOS, nothing. Here's a link to the board, "CPU Overheat Protection System" displayed, nothing to control it in the BIOS or manual.
BTW, it's a Barton 3000+, running at stock. With the fans cranked slightly, and Prime95 running now since half an hour ago, the temperature shown is 53.5C on the CoolDrive, but MBM is still reporting 80C temperatures. The heatsink is mounted correctly (believe me, I checked everything!), and feels fairly warm to the touch- not scorchingly hot, but hot enough to convince me that it's making proper contact.
Thoughts please?
Cheers!
The CoolDrive is set to knock all the fans up to maximum speed if the CPU or VGA sensors hits 65C, and if the case/HD sensors hit 55C. So, I thought I'd test it (Note to self: stop stress-testing everything).
Crank all the fans down to minimum, Prime95 the beast. 63C shows on the display, and the PC resets. Odd, I though, and tried again. 62C this time. Entering the BIOS (which, by the way, is a Leadtek nForce-2 board with the crappiest Health Monitoring features ever, and no voltage monitoring, argh), shows 68C for the CPU, which has been 46C up until now.
Thinking that's just weird, I crank all the fans up, and the temperature won't budge. It won't go up, either, it's just stuck permanently on 68C (even on a cold start). MBM5 (which doesn't like my board, really), is showing 80C. However, the CoolDrive CPU sensor, which is mounted next to the core, between the heatsink & susbstrate of the CPU, reads what very much appears to be the correct temperature.
Is my thermal diode broken? Did that sudden reset indicate that the board has had enough and reset to prevent damage to the CPU? I have remounted and reapplied decent thermal goop, which hasn't made any difference. There's no "Bad Temperature! Bad! Shutting down now!" feature in the BIOS, nothing. Here's a link to the board, "CPU Overheat Protection System" displayed, nothing to control it in the BIOS or manual.
BTW, it's a Barton 3000+, running at stock. With the fans cranked slightly, and Prime95 running now since half an hour ago, the temperature shown is 53.5C on the CoolDrive, but MBM is still reporting 80C temperatures. The heatsink is mounted correctly (believe me, I checked everything!), and feels fairly warm to the touch- not scorchingly hot, but hot enough to convince me that it's making proper contact.
Thoughts please?
Cheers!
Facebook
Twitter