It
would of been impossible to test all these heatsinks on a CPU.
So I decided to use a technique, which I first saw at 2cooltek.
Instead of a CPU, a peltier unit is used for the heat source.
Slap the heatsink on the hot side of the peltier and a thermometer
on the cold side. The difference between ambient temperature
and the cold side of the peltier represents the heatsinks performance.
Here is a picture if my test board with the Golden Orb running:
On
the left you can't see the 250w power supply, which powers the
peltier, and the fan(s). On the board is the Golden Orb, which
sits on top of the 68w Hitech peltier unit, which can be found
at 2cooltek. Then there are the 2
CompuNurse thermal probes, which I got from the CoolerGuys.
One of the probes is under the peltier, measuring the cold side
temperature. The other one measure the ambient temperature all
the thermal pads where removed from the heatsinks and replaced
by Arctic Silver thermal goo provided by Nerdsbyte.
This
is probably the most precise test you can do at home. However,
I don't believe it is 100% perfect. First of all, the CompuNurse
has an accuracy of 1ºC. Then there is the amount of thermal
paste used and the pressure between the peltier and heatsink.
There are also ambient factors like air drafts and ambient temperature.
Now that this is said, here are the results I got using the
heatsinks and their default fan(s):
Model
|
Fan(s)
|
Ambient |
Cold
Side |
Performance |
Globalwin
VOS-32 |
2x
60mm YStech |
18.7 |
-13 |
31.7 |
Alpha
P3125 |
2x
60mm Sanyo Denki |
19.7 |
-11.1 |
30.8 |
Globalwin
FLK-38 |
1x
60mm Delta |
18.1
|
-10.4 |
28.5 |
RDJD
K702 |
2x
50x15mm |
21.1 |
-7.4 |
28.5 |
Vantec
K7D-5032 |
2x
50x20mm Top Motor |
20.9 |
-6.9 |
27.8 |
Globalwin
FLK-32 |
1x
60mm YStech |
18.7 |
-2.8
|
26.9 |
Vantec
P3D-5030 |
2x
50mm YStech |
19.9 |
-3.7 |
23.6 |
Globalwin
VGS-08 |
2x
50mm |
18.7 |
-3.9 |
22.6 |
Thermaltake
Golden Orb |
1x
50x20mm |
23.3 |
1.6 |
21.7 |
Next
>>
<<
Previous
|