MCH Temperature Registering 83C

[Lufkin]

I have a an Intel® Desktop Board DQ965CO (it's a MicroBTX board) BIOS version 5150 with an E6400 Core 2 Duo processor and mushkin eXtreme Performance 1GB 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Desktop Memory. I am not overclocking (Intel BIOS is pretty barebones). Here are the pertinent cooling specs.

COOLER MASTER Centurion B541 RC-B541-SKN1
Two 92mm exhaust fans (running around 600 RPM)
OCZ ModStream OCZ52012U 520 Watt power supply (I list it because it contains a 120mm fan)
ThermalTake Silent BTX Cooler DHCLP0191 (runs at around 350RPM when processor is idle)
Arctic Silver 3

I am concerned that the 83C degrees registered by the MCH on the Intel Desktop Utilities ver. 3.0 is excessive. Which of these is the better first option to reduce the MCH temp. Should I replace the BTX cooler fan so it runs at a higher RPM (I will miss the silence)? Or should I remove the heatsink and try replacing whatever suboptimal thermal conductor exists on the MCH with AS3 and then reapply the stock heatsink? This is only possible if Intel didn't glue the heatsink on (it has a clip, but it might have been glued as well).

I know I am a bit behind the times as Arctic Silver 5 is out now, but I still have half a tube of AS3...

Hardware Monitor Screenshot

 

[Lufkin]

<bump> Anyone?

 

[imported_lavalamp]

I'm quite concerned that your CPU fan only runs around 350rpm, is the fan speed temperature controlled by the motherboard? It may be spinning fast enough to cool your CPU, but any surrounding componants that rely on the CPU fans blow-off to be cooled might not be getting enough air moving over them. Try running your CPU at full speed to see if it makes any difference to the MFT temp.

If it doesn't make a difference then you'll have to try something else, but if it does then you'll have to find a compromise between the temperature and the sound level you're willing to endure.

 

[pkme2]

I would add 120mm Yate Loons. Front intake 120mm YL. Top exhaust 120mm, Side cooling would be greatly added with two side 120mm intake fans. Cooling would improve at least 7-10C I took my original temps and compared them with my final, and really noticed great improvement.
Intels can utilize them in certain case designations to improve cooling.

 

[pkrush]

I'd put better thermal grease on the stock chipset cooler first, and if that doesn't help, I'd mount a more powerful front fan in the case so the chipset gets more airflow. By the way, why did you buy a BTX board and case anyway?

 

[Lufkin]

Thanks for the replies. The CPU fan is connected to the motherboard. It does spin higher (up to 670 - 750 RPM) when I run the freeware stress test byPassMark that came with the motherboard.

Pkrush, I was really interested in building a smaller, quieter box than my last two system builds. I am also not a hardcore gamer so I was not that concerned about warm exhaust from the Type I BTX cooler pouring over a molten SLI setup. (I am currently using the onboard Intel video). I read all of the material online I could find on BTX, and I understood that is not that popular a platform with homebuilders. But in the end I wanted to try something new so I chose a micro-btx motherboard. Not the best system for overclockers (since it is an Intel board the BIOS is far from flexible). It does have a PCI-Express 16x slot if I do decide to make Civ IV a priority, but until then I hope it will be a quiet, steady performer with a smaller footprint than my old system.

The high temperatures are bumming me out though. I had read several less than stellar reviews on the BTX cooler I chose, but opted to give it a try because it was so quiet...The ICH (63C) and MCH (84C) temps are really unacceptable. I will try replacing the CPU fan with a faster RPM version...so much for quiet.

 

[Lufkin]

This is the first weekend I really had a chance to sit down and troubleshoot my thermal issues. There are only three fan connectors on the motherboard, 1 3-pin connection for outlet fan, 1 3-pin connection for inlet fan, and 1 4-pin connection for the active thermal management fan.

In my original setup, I had two exhaust fans, one hooked up to the 3-pin outlet header and one hooked up to the 3-pin intake fan header. My BTX Type I cooler was hooked up to the 4-pin header. In this setup, my processor fan would rarely spin faster than 350 RPM which meant the MCH heatsink at the back end of the BTX thermal module was getting very little, if any airflow.

Being a noob and thus not knowing how that active thermal management crap worked on the 4-pin fan header, I thought that maybe the active managment was trying to balance the intake/exhaust air flow. Even though I had two exhaust fans, one was connected to the intake so I purchased a splitter and hooked up both exhaust fans to the 3-pin outlet fan header. I rebooted and there was no change. The processor fan was still moving at 330RPM and the MCH was still reading 83 degrees C.

For my next try at resolving my heat issues I hooked up the 4-pin connector from my thermal module to the 3-pin intake fan header. I rebooted and checked out the fan readings and now the processor fan was spinning at around 1250RPM. The MCH temp had dropped to 50C.

I then read on Intel website that the acitve thermal management doesn't spin the processor fans at full speed until 72C is reached.

The Thermaltake fan is supposed to spin at a maximum of 2500RPM, but I'm guessing it never will. At this point I am happy enough with the cooling/silent performance of the module to avoid the hassle of an RMA. While the 4-pin fan header might be the bees knees for a noisy prescott fan in an ATX system, it seems that the 4-pin fan header should not be used in a BTX environment where other components depend on airflow from the processor fan.

...Now if I could just bring down the ICH temp which is still a hot 63C
Temp readings after troubleshooting