Testing The Coolers!

Now that we had gone through all the heatsinks here, let's see just how well they perform. In the coming pages, we will put these coolers to the test and compare their performance.

First, let's take a look at our testbed and testing methodology.

Testbed

Testing Methodology

Idle temperatures were taken when the system was left idle, without any application running. For load temperatures, CPU Burn (with error checking disabled) was used to keep the CPU loaded.

The ambient temperature was taken by placing the probe near the fan intake, while the CPU temperature was obtained by placing another probe on one corner of the processor IHS. The temperature difference or delta was then obtain by deducting the ambient temperature from the CPU temperature.

Before each measurement was taken, we allowed the system to stabilize at that setting for at least 20 minutes. The ambient temperature and CPU temperature were recorded at 1 second intervals for 250 seconds. Finally, the temperature differences or deltas were calculated based on the average of the 250 seconds of readings.

Each heatsink or cooler were tested at least twice to look for inaccurate or abnormal results. Since it is near impossible to obtain the same or very close results, the results were considered acceptable if the variance did not exceed 1°C.

Accepted results will be averaged again, to obtain the final result. If the results do not come close, they will be discarded and the heatsink/cooler retested.

All the tests were performed outside the case. Incidentally, we only tested the CNPS9500 LED version because both non-LED and LED versions are the same. The main difference is the use of an LED fan, instead of a non-LED fan.

Air Flow & Noise

Before going to the test results, let's have a comparison of the air flow of each fan. We don't have the exact air flow rating for each fan, but these ratings are quite meaningless since every manufacturer has their own method to derive such values.

However, it is possible to 'feel' which fan has more airflow and which has less. When at full speed, it goes like this (from the most airflow to the least airflow) :

CNPS9700 (12V) > SUNON KD1212PTB3-6A > ICute generic > MiniTyphoon (12V) > CNPS9500 (12V)

Let's take a look at a comparison of their noise levels here.

SUNON KD1212PTB3-6A > CNPS9700 (12V) > MiniTyphoon (12V) >CNPS9500 (12V) > ICute generic

The ICute generic fan wins hands down. We can hardly hear any sound from the motor, only a slight noise from the air flowing through the fan. In case you are wondering... Yes, the MiniTyphoon is really noisy when it is running at full speed, even though it is a 92mm fan.

The SUNON fan was the noisiest, although it isn't the one that with the greatest air flow. All the fans are tested at 5V, except for the SUNON and ICute fans. This is because the SUNON fan barely moves at at 5V!

The ICute fan is already a low-speed fan and extremely quiet. In fact, its noise level actually beats other fans (even running at 5V!), so there isn't much of a reason to run it at 5V.