Viewing page 8 of 8 pages. Previous 1 2 3 4 5 6 7 8
ADDENDUM 1 Aug 2012: The Real Final Word by Mike Chin
The comments in the forum discussion of this review provoked further communication between bequiet! reps and myself. The last email query was whether our voltage control of the bequiet! PWM fans might have affected the results. I pointed again to our carefully controlled and identical setups for testing and comparison, using either RPM or SPL as reference points. The comparison table in the review makes it very clear. I also reiterated my position that the non-flat, somewhat concave profile of the base was, in my opinion, the #1 weakness, which might be exacerbated by less than ideal pressure between the base and the CPU.
Here is the imprint of the TIM on the CPU and the base of the Dark Rock 2 from the last sample tried and tested. The striations of the TIM mean that the TIM was not sandwiched tightly between the two surfaces, and thus floated a bit to make this pattern. But because the TIM does not cover the entire surfaces of the base and the CPU, it's hard to see that along the top and bottom edges, the TIM has actually been pressed out, and there's much less of it remaining this is as it should be in the center.
I decided to test my hypothesis one last time. This time, rather than use the stock bequiet! mounting hardware, I borrowed the rigid, heavy-duty, machined brackets and bolts from a Prolimatech Megahalems cooler. This is probably the best mounting hardware I've come across during my >15 years of playing around with CPU heatsinks. Most of the Thermairight and Noctua CPU cooler mounting hardware is just about as good, but what's particularly useful about the Megahalems hardware is that it is easy to use with many other tower heatsinks.
The end result was virtually identical to previous results, with an improvement of perhaps 1°C over the best result we recorded for the three samples in previous tests. Time to examine the TIM imprint.
This time, you can clearly see that there is a line on the top and bottom edges of the heatsink, matched on the CPU, where the TIM has been pressed away, due to the high pressure of the Megahalems mounting hardware. Yet, with all that pressure, the TIM in the center still has those fractal-like striations. This is conclusive evidence that the base of the bequiet! Dark Pro 2 sample here is concave.
There's no question why all the added pressure of the Megahalems mounting did not make any improvement in cooling performance; the center of the base was still not tightly pressed against the CPU because the base of the cooler is concave. Only the edges made tight contact, but that's not where the heat of the CPU is concentrated, it is in the center. This is the primary reason that the three Dark Pro 2 samples all have such mediocre cooling performance.
One last point: An electronic, infrared, non-contact thermometer was used to check the temperature of various points on the heatsink during the above thermal test. With room ambient at 24°C, and the CPU core reporting 70°C, none of the fins read any higher than ~30°C, and the hottest portion of the heatpipes right at the base of the heatsink read just 37°C, My conclusion: Much of the heat from the CPU was simply not getting into the heatsink.
A photo of a TIM imprint on the base/CPU of a high performance cooler would really put the matter to rest, so here it is.
Desired TIM imprint.
The heatsink was turned 90 degrees for the above photo, and the TIM was smuged a bit during uninstallation, but you can clearly see the matching imprints of the TIM on the base and on the CPU. Both have a central area where there is very little TIM; it's been pushed from the center due to very tight contact. Only around the periphery are there any striations of the TIM. Incidentally, this is a new photo taken just for this addendum, and the heatsink is the original Megahalems sample we tested a couple years ago. Its performance here was 12°C better than the bequiet! Dark Pro 2, with the latter's stock fan at full speed.
* * *
this article in the SPCR forums.
|Help support this site, buy from one of our affiliate retailers!|