Articles: Cooling

Bookmark and Share


Table of Contents

Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 ]

Over a year has passed since we posted our last fan roundup. A lot of ne fan models have been released since then in both: 120 mm as well as 140 mm form-factor. We managed to collect an enormous number of new 120x120x25 mm fans and quite a few 140x140x25 mm ones. The latter products will open a series of fan roundups on our site today.

The fans of this size are currently pretty popular. It is not only because processor air-coolers more and more often use 140 mm fans these days, but also because they gradually replace the common 120 mm fans inside contemporary system cases, which is exactly what happened back in the days with the 92x92x25 mm fans. This is a pretty logical process, although it will be some time before 120 mm fans go away completely. Nevertheless, we believe that 140 mm fans will continue increasing their presence not only in the CPU cooling products, but also as system case components.

Well, today we are proud to introduce to you 13 fan models from nine cooling solutions manufacturers, such as: Deep Cool, Evercool, Nanoxia, Noctua, Noiseblocker, Revoltec, Scythe, Thermalright and Xigmatek.

We would like to start by telling you a bit more about our new fan testing equipment and methodology.

Fan Testing Methodology

We used the same measuring tools as last year: our special controller card, CENTER 321 noise meter and ProskitMT-4005 anemometer:



noise meter


Let me start by saying that we used a new controller revision that allows measuring the voltage with a finer increment of 0.1 V instead of the formerly used increment of 0.3 V. That is why this time we set the fan rotation speed using 0.5 V increments in the interval between 4 and 12 V. For the purposes of unification of our testing approach, we measured the rotation speed of PWM controlled fans using the same voltage setting. Moreover, now our controller can also display the average fan rotation speed after the stabilization period. Of course, we didn’t leave out the fan current and startup voltage. We took all our measurements after 5-7 minutes to ensure that the motor has warmed up enough.

The methods used for measuring fans acoustic characteristics have been seriously modified. Now we didn’t just use a polyurethane foam stand, but hanged them inside a special wooden frame:


This frame weighs 2.3 kg and has four silicone stands attached to the bottom of it:

Moreover, the fans didn’t touch the frame and were fastened using special soft silicone mounts:


In addition to that the noise meter is now installed on a tripod with rubberized legs, while in our previous session last year it was sitting sideways on top of two cardboard boxes:

The distance to the receiver of our noise metering device was 150 (±1) mm and the device itself was centered precisely along the axis going through the center of the fan rotor. All these modifications and improvements of our testing techniques should ensure more precise measurement results that would be less dependent on the external factors. Here I would also like to add that all acoustic tests were performed between 1:00 and 3:00 AM in a closed room about 20 m2 big. There were no computers or any other electrical appliances working in this room except for the controller. The subjectively comfortable acoustic maximum is at 36.0 dBA, acoustic quiet – at 33.0 dBA.

The airflow was measured in feet per minute, in order to calculate its volume in the most typical unit for cooling fans – CFM. This methodology has also been significantly improved. First of all, we used an air guider. In the beginning we decided to go with a Nexus BeamAir 120. However, since the results obtained with this air guider didn’t seem credible enough, we decided to make one ourselves. Here is what we came up with:


Our custom-made air guider was built with almost 700 cocktail straws, 8 mm in internal diameter and 210 mm long, with very thin and smooth walls. It allowed us to eliminate possible turbulence and additional spinning of the anemometer blades. To ensure lower airflow resistance the adjacent straws were shifted vertically by 4-5 mm.

Anemometer was inserted into a round of polyurethane foam:

Since the air guider was inserted in the center of the pipe, the distances between the guider and the fan and the guider and anemometer were the same and equaled 145 mm.

We made two more polyurethane foam rounds to install the fans: one for the ones with a square frame, and another – for the fans with a round one:

Both rounds were inserted into the ends of a 500 mm long plastic pipe segment with 200 mm internal diameter:


The foam rounds sat very tightly inside the pipe, so there was absolutely no airflow loss: all of it went through the air guider and then through the anemometer fan module. We tested each fan twice in its entire rotation speed range and took the average value from both test runs (unless there were some serious peaks or drops, in which case the test was performed one more time).


To calculate the volume of the airflow we multiplied the measures airflow speed by the size of the anemometer blades surface times three (so that it could match more or less closely to the airflow readings declared in the fan specs). We measured the airflow sped with the air guider as well as without it. The graphs and charts show the airflow readings with the air guider but we will also provide a link to a complete Excel-file with all the results and fan specs.

Besides the pipe with an air guider inside, we also measured the airflow when there were three different heatsinks installed right next to the fan: Scythe Ninja 3 with the fins set wide apart, copper Thermalright TRUE with very dense fin array, and a two-array Thermalright Silver Arrow with the highest heatsink fin density of all: 

Scythe Ninja 3

RUE Copper

Silver Arrow

However, the readings taken off both Thermalright heatsinks were the same to the nearest tenth, while with a Scythe Ninja 3 heatsink the airflow didn’t change at all. Therefore, we decided to provide only the results taken with a two-array Silver Arrow heatsink (besides the results with an air guider).

Since all fans worked under load inside the pipe, their rotation speed was lower than in an open testbed during our noise measurements. You can check out the speed difference in a table provided at the end of the article. However, in order to scale acoustic and airflow measurements identically, we increased the airflow readings by a coefficient illustrating the relation between the fan speed inside the pipe and on an open testbed.

Well, I assume this detailed description of our testing methodology should be enough to move on to the actual test of our 140 mm participants. With the exception of Noctua, we had at least two units of each 140 mm fan model. We tested acoustics of all available fans, while airflow tests were only performed for one unit of each model. All fans discussed in this roundup are sorted in alphabetical order.

Testing Participants Technical Specifications

We are going to mention individual specs of the tested fans later in the corresponding chapters. Here we would like to sum up their complete specs in a reference table for your convenience:

Click to enlarge

I would only like to add that all fans spin counterclockwise.

Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 ]


Comments currently: 16
Discussion started: 09/29/10 05:11:28 PM
Latest comment: 04/21/13 08:27:14 AM

View comments

Add your Comment