ThermalTake TR2-M4

Author
Nivram
Posted
October 6, 2003
Views
27422
Looking for a good HSF for that OEM CPU purchase or just a better than stock HSF for peace of mind. Read about one here.
Tags Cooling

Page 1: Background to Conclusion

<B>A Little Background</B>:
I have been working with electronics for over 30 years. Started by learning to operate and maintain high frequency radios in the military service. Because the components were a lot less integrated (yes, no vlsi) and 10kw power output requirement, these radios produced a lot of heat and therefore required extensive cooling. All types of cooling techniques were used: axial, ducted axial, and cage fans located within the equipment; externally mounted air conditioning and large ducted sump cooling systems.

Although not the size of those behemoths (think of restaurant size refrigerators), today's CPUs also produce a lot of heat for their size. What I find intriguing is the use of ducted and cage fans techniques for cooling from my introduction to electronics that are coming back into use today. What is vastly different of course is the small size when I compare them to my memory of those times.

Although unrelated to this review, I also remember these radios had windows (unetched of course). Not to look at the nice show from the neon and laser lites, but to notice dangerous component arcing or just exploding (luckily thick, reinforced glass).

<B>Introduction</B>:
Aron recently sent me a <a href="http://www.xoxide.com/thtramdp4co.html">ThermalTake TR2-M4 HSF </a> to test that he had received from <a href="http://www.Xoxide.com/">Xoxide</a>. It is, of course, based on cage fan technology mentioned above. I had only seen photos of this and other similiar HSFs, so I was pleasantly surprised by the overall, small size of the TR2-M4. Also, included in the package was a nice half page installation guide, it is true pictures are worth a thousand words.

<center>tr2m4</center>

<B>Specificatons</B>:
Below are screen clips from ThermalTake's website of the specifications and features for the TR2-M4. Of course, I like the use of a three prong heatsink mounting mechanism. This unit is set up to use a four prong power (almost 1/2 amp max draw) connector with a separate motherboard fan speed monitor. The fan speed control reostat is mounted on a card slot bracket for mounting in an unoccupied PCI slot. It would be easy to remove the reostat from the bracket and mount in on a front panel for easier fan speed control asi it has very long leads.

<center>hsfspecs features</center>

<B>Testing</B>:
The testing system is (no overclocking was attempted during this setup):

Shuttle Motherboard, Model# AK35S
AMD ATHLON XP 2200+ CPU
Kingston DDR RAM 256MB PC2700

<center>testsetup</center>

The testing scenerio was 1/2 hour to obtain idle temperture, then 1/2 hour of folding@home to stress the CPU to obtain the load temperature. Temperature and fan rpm readings were done using MBM5 taking high and low readings for temperatures and average readings for fan rpms.

I setup the tests using the stock AMD HSF with ceramique against the TR2 M4 using its stock thermal compound and ceramique. I then placed the results into a chart trying to make it easy to read and understand. The readings were corrected for ambient temperature variations, but were rerun if ambient changed more than one degree during testing periods.

<center>testtemps</center>

It was a pleasant surprise to see how well the ThermalTake thermal compound performed in competition with the ceramique. Also, I was wondering how the mechanical mounting of the copper plate to the the aluminum heat sink would work. Obviously ThermalTake did a good job in mating these two surfaces. I should note here that the thermal compound on the bottom of the TR2-M4 is not the standard thermal tape. I touched the edge of the compound (outside the contact area for the AMD CPU, so no foul here) with the tip of my finger and made a smudge, which I could never do with standard thermal tape.

<center>copperplate</center>

I did notice that the specifications called for a 2100-4500 rpm fan speed, but I registered readings of 2450 to 4750. You could attribute these to the motherboard misreading the fan speeds. Correcting the reading for the 10% factor noted with the speed specification brings the fan readings almost to the stated specification.

<B>Conclusion</B>:
It was a pleasant surprise to see how well the ThermalTake thermal compound performed in competition with the ceramique. I was wondering how the mechanical mounting of the copper plate to the the aluminum heat sink would work, but obviously (as noted above) ThermalTake did a good job in mating these two surfaces. With its relatively small size and weighing in at only 397g, mounting this unit in any position and on most motherboards should not prove to be a problem. I want to thank <a href="http://www.Xoxide.com/">Xoxide</a> for providing this opportunity for a wonderful week of testing.


<center><img src="http://www.aselabs.com/images/awards/2.gif"></center>

Title

Medium Image View Large