Be Cool: Ultimate 120/140 mm Fans Roundup

After a month of hard work and numerous experiments on a new unique testbed we are proud to offer you a roundup including 57 models of fans from 25 different product series collected from all over the world.

by Sergey Lepilov
06/22/2009 | 02:53 AM

Most contemporary cooling systems for CPUs that are worth considering as possible replacements for default (read “boxed”) coolers come equipped with 120x120x25 mm fans. It is becoming more common lately to equip them with even larger fans measuring 140x140x25 mm. Moreover, some cooling solutions makers do not even bother to bundle their heatsinks with any fans at all making this important choice users’ responsibility. It is even more acute to find powerful and quiet fans for the system cases, because very often case makers (even eminent ones) rig them up with whatever they can find. That is why we feel time has come for a detailed study of available 120/140 mm fans.

 

For that purpose we collected 57 fans of 25 different series from 15 manufacturers. Despite a large number of fans, this roundup doesn’t cover the entire existing variety available in the today’s market and we hope that from now on we will be able to offer you new fan roundups on a more regular basis. Since our first roundup called 11 Fans for Two Super-Coolers and One System Case we have developed a completely new testing methodology. We do not claim that it is the only correct approach to fans testing, but we are going to improve it in our upcoming reviews and hopefully with the help of your feedback as well. So, let’s start with the description of our testing methodology and equipment.

Fans Testing Methodology and Equipment

The primary fan testing was performed outside the system case without any cooling heatsinks involved. We decided to do it this way for two reasons. First, we had to exclude the effect of the noise generated by the system case on the results of our acoustic measurements. No matter how quite your system case and its components are, you can’t underestimate the effect they have on the ambient acoustics. Second, testing fans on top of heatsinks does have certain practical value and may be an important addition to the main test session, but you won’t be able to get exact results in this case. Especially, since some results obtained during our test session turned out pretty close to one another and there is no way to catch this difference using the temperature of the central processor hard drive or liquid circulating inside the cooling system contour.

During our test session we used a controller of our own proprietary design. This controller works independently and doesn’t need to be connected to a PC. It doesn’t generate any noise, and most importantly, allows controlling fans of all types – with three-pin as well as four-pin connectors.

 

Three-pin fans are regulated through voltage that can vary from 0 to 12 V with 0.3 V increment obtained with a liner stabilizer. Four-pin fans receive nominal voltage of 12 V and their rotation speed is regulated with a separate PWM signal with standard frequency of 25 kHz. In the latter case fan rotation speed is changed from 0 to 100% with 2.5% increment.

The fans are connected via low-resistance shunt, which allows us to measure the current. The maximum fan current that our device can handle is 1 A with 1 mA increment. The device was calibrated before the tests using reference load, so the measuring error for the voltage and current doesn’t exceed 1%. The unit is based on Atmel ATmega168 microcontroller. For current measurements it uses its own ADC together with AD8605ARTZ operational amplifier that amplifies the shunt signal. For fan rotation speed control via voltage it uses Analog Devices AD5245BRJ10 digital potentiometer and voltage regulator built with National Semiconductor LM7301IM5 operational amplifier and IRF IRL3502 field-effect transistor. In case of four-pin fans the PWM-signal is generated using built-in timer of the microcontroller.

Moreover, this device also measures fans rotation speed using the signal from its own diode that generates two impulses for each fan rotation. The operational mode (voltage or PWM control), voltage setting or relative PWM duration, as well as the measured current and rotation speed are displayed on a small LCD screen in real time.

During the tests the voltage of each fan changed from 3 V to 12 V with 0.9-1.2 V increment. In our opinion, it didn’t make sense to lower the voltage below 3 V, because not all the fans could work fine even at 5 V voltage setting. We recorded maximum current for each voltage setting and fan rotation speed. Moreover, we also determined the startup voltage for each tested fan (several start/stop cycles). Don’t be surprised if the operational voltage is lower than startup voltage: it means that once the fan blades reach stable rotation speed the voltage may be lowered and the fan will just keep on working fine.

Besides, for PWM-controlled fans we also determined the dependence of the rotation speed on the voltage for two reasons: to maintain unified testing methodology and because the tests performed at low voltage are considered heavier than tests of PWM-control.

Our second testing device was a thermal anemometer – a device used for measuring airflow speed and temperature. We used Pro’s Kit MT-4005 thermal anemometer with 0.01 m/s or 1 ft/min precision:

Although our anemometer only registers the speed of the airflow passing through its fan (which is very light and creates almost no resistance), we can use the size of this fan to calculate  the airflow in more common units used by most fan makers - cubic feet per minute (CFM).

To ensure higher measurement precision, namely, to make the entire airflow generated by the fan go through the anemometer blades we had to be creative. We took a 5-liter plastic bottle and removed its bottom and neck. We cut off the bottle neck so that it matched the anemometer frame exactly, then attached the anemometer to it and sealed the contact area with insulating tape. The bottom of the bottle was removed to accommodate the tested fans. For that purpose we made a special polyurethane foam ring with a square cutout measuring 115x155 mm. The diameter of the ring was 3 mm bigger than that of the bottle. During the tests we inserted a fan into the square opening in the polyurethane foam and then inserted the ring with the fan into the bottom of the bottle:

 

As a result, this device was very airtight from the sides and had an admission opening and a discharge outlet. For example, if we inserted the fan into the bottle without the polyurethane foam ring, the airflow speed was only half as high. The distance between the anemometer fan blades and the tested fan was 190 mm. Things were a little more complicated with 140 mm fans because they were too big and didn’t fit into the bottom of the bottle. Therefore, we simply pressed them against the bottom part of the bottle and sealed everything with sticky tape. The only exception was Scythe Kaze Mary fan that fit into the bottle together with the foam ring, just like all 120 mm fans.

The whole thing was placed at the edge of the computer desk to ensure free airflow to the fan almost over its entire diameter:

 

Before we took any of the readings down each fan worked for about 5 minutes to warm up the bearings and make sure the characteristics have stabilized. We took the maximum airflow readings for each fan voltage for the performance diagrams.

The noise level of each fan was measured between 1:00AM and 3:00AM in a closed room about 20 m2 big using CENTER-321 electronic noise meter. The noise meter was always placed along the same axis as the fan rotor at an exact same spot. To ensure precision we set special marks for the noise meter as well as fan placement. The distance between the noise meter receiver and the fan equaled 350 mm:

As you can see, we set the fan on special stand made of polyurethane foam (with placement markings as well). We had to give up shock absorbing retention and reduce the distance between the fan and the noise meter, because CENTER-321 registers 30 dBA as the minimum reading. And since some fan models generate less noise at minimum and even medium rotation speeds, we had to resort to this specific measuring technique. We will provide not only the actual measurement results, but also our subjective opinion on the noise level of each particular fan model.

We are going to provide an individual graph with test results for each fan. The vertical axis on the left stands for the fan airflow and noise level, the one on the right shows fan power consumption in watts. The horizontal axis indicates fan rotation speed. All results are summed up in a table below the graph.

Technical Specifications of All Testing Participants

We are going to mention some of the technical specs during the review of each particular fan model. Here we would like to sum up all the specs in a comparative table for your reference:


Click to enlarge

Akasa Silent Color 140 mm

Akasa is one of the largest fan makers out there; they currently offer over 60 models of fans of different size and application. We managed to get three fan models for our today’s roundup. We would like to start with Akasa Silent Color fan of 139x139x25 mm size. This fan comes in a light-blue cardboard box with a large cut-out window in the front:

The back of the box shows the exact fan dimensions and distance between the retention holes, and also lists all the fan technical specs. Akasa Silent Color comes with four long shock-absorbing spindles and a cable with a PATA power connector:

This fan has 7 blades and a rotor with 45 mm diameter. Its distinguishing feature is acid-blue color of the fan blades and a holographic spindle sticker:

The blades are of pretty common shape, they expand from the beginning towards the tip. They are larger than the blades of a 120 mm fan, of course. Four supporting rods holding the fan cannot boast small size or aerodynamically effective shape:

 

The fan works at 1000 RPM. The airflow at this rotation speed is claimed to be 47.1 CFM – higher than by 120 mm Akasa fans working at higher rotation speeds. However, the static pressure is way lower and is declared to be only 0.700 mmH2O. The noise is expected to be quite moderate: around 19.5 dBA.

The peculiarity of this particular fan model is its ability to fit into the slots designed for 120 mm fans. That is why the frame has additional retention holes in it (for 115, 120 and 125 mm slots):

According to the marking, the fan was made by Young Lin Tech Co. LTD:

The sleeve bearing inside should last at least 20,000 hours without failure. This fan should consume no more than 1.4 W of power. The latter spec can be easily confirmed on our testbed:

The manufacturer reported the power consumption absolutely correctly and the fan turned out really economical, but the airflow is a little too weak even for a 120 mm Akasa fan, not to mention the best fans tested today. The fan doesn’t produce too much noise, it remains acoustically comfortable at 630 RPM. We didn’t detect any parasitic noises from the fan motor. Its startup voltage is pretty high - 10.5 V.

Akasa UltraQuiet Amber (AK-183-L2B)

The next model is Akasa UltraQuiet Amber (AK-183-L2B) that ships in clear plastic blister package with a cardboard insert inside:

The back of the package enlists brief fan specs in five languages. The bundled accessories include a PATA power connector and four self-tapping screws:

The fan is made in China and consists of a semitransparent plastic frame and amber-red fan sitting on four supporting rods and rotating counterclockwise:

The fan measures 120x120x25 mm. It has 7 blades and a rotor with 41 mm diameter. It has a pretty short power cable of only 300 mm:

 

Among the distinguishing features of Akasa UltraQuiet Amber fan I should mention blades with large sweep area and more aggressive angle and a small lug right at the rotor:

The fan rotates with constant speed of 1400 RPM. At this speed the fan should create 44.8 CFM airflow, which is a very modest number for this speed. The static pressure is claimed to be at 1.100 mmH2O, while the level of noise should be no higher than 18 dBA.

The fan uses a double ball bearing with 80,000 hours MTBF. The “KM121225LB” marking on the rotor sticker indicates that it was originally made by Yen Sun Technology Corporation (YS Tech):

The same sticker lists the electrical specs of the fan. It is priced at $14.

Let’s see what Akasa UltraQuiet Amber fan is capable of:

The fan generates a decent airflow in its entire supported range and consumes half of the power claimed by the specs. It starts at 3.6 V and can keep rotating up until the voltage drops down to 1.7 V, which indicates high quality of the employed bearing. However, all this good is totally ruined by the crackling sound produced by the bearing, which is so strong and annoying that I nicknamed this fan “Chinese economy scooter”. This fan’s acoustics is the worst of all tested today. There is no subjective acoustic comfort around this fan, but our noise meter showed that the noise dropped to 33 dBA around 660 RPM.

Akasa White LED (AK-FN055)

Another 120 mm Akasa fan model participating in our today’s tests comes in a cardboard box with a fan picture on the front side and its detailed technical specifications on the other:

The bundled accessories are exactly the same as those of the above described UltraQuiet Amber fan. However, the new fan looks a little different: it has a clear frame and glossy white fan blades:

The fan is 120x120x25 mm big. We don’t know how much it weighs, as there is no mention of it in the specs. Each fan has 7 blades and a 43 mm rotor diameter. The fans come with a 300 mm power cable:

 

The shape of the Akasa White LED blades is very similar to the shape of Akasa UltraQuiet Amber blades; however, there is no sharp lug at the beginning of the blades:

The fan rotates at 1200 RPM and creates 38.4 CFM airflow. Its static pressure is even higher than that of the amber Akasa and equals 1.170 mmH2O. The noise level is declared not to exceed 17.5 dBA due to a sleeve bearing with guaranteed MTBF of only 20,000 hours.

The marking on the fan rotor reads “DFS122512L” and indicates that this fan belongs to Young Lin Tech Co. LTD:

The declared power consumption levels are much more modest than those of Akasa UltraQuiet Amber and shouldn’t exceed 1.8 W.

The fan is equipped with four white LEDs that look very attractive in the dark:

 

Akasa White LED is made in China and is recommended to retail at $13.

The obtained test results show that the white Akasa fans make much more pleasant noise than the amber ones:

The fan retains acoustic comfort up until 900-930 RPM, the bearing makes no crackling of any kind at all. The airflow is pretty average for our today’s testing participants, which means that this fan doesn’t work any wonders in this respect. However, the claimed power consumption didn’t get confirmed, because the fan consumed about 2.24 W at maximum speed instead of the declared 1.8 W. the startup voltage is one of the lowest in our today’s tests – only 2.7 V.

Enermax Apollish

We are going to start our discussion of Enermax fans with a brand new model called Apollish. Clear plastic blister with an open front reveals almost the entire fan to the user’s eyes, so you won’t even need to open it to check the fan out:

The fan comes bundled with six (in case you lose two of them?) silicone mounts and a PATA power connector. There is a cable with the on/off button for the built-in lighting and a cable with a thermal diode attached to the fan:

 

The cardboard insert inside the package lists the key features of the new fan model and offers detailed instructions for fan installation and removal for cleaning purposes:

The fans are made in China that can at this point be called a “fan country” :)

Enermax Apollish is available in four different colors: green, red, silver and blue. We managed to get two latter models for our today’s roundup:

 

These 120x120x25 mm fans are very beautiful. We don’t know their weight, but each of them has 9 blades and a rotor with 40 mm diameter. Frankly speaking, at first glance these fans look more like shiny toys that serious cooling devices:

 

Unlike other fan models, the blades of these fans are not curved too aggressively, we can even say that they are almost flat. The gap between the fan blade tip and the inside of the fan frame is 3 mm. the supporting rods are rectangular, 3 mm wide (except for the one with the cable inside).

Enermax Apollish fans rotation speed varies automatically according to the thermal diode readings. At 25 °C it will be 700 RPM, and at 55 °C will increase up to 1700 RPM.

The created airflow should be between 28 and 72 CFM while the noise level should vary between 10 and 19 dBA. I have to say that this is very optimistic acoustic data…

One of the peculiarities of the new Enermax solutions is the removable fan blades unit. All you need to do is apply even pressure to the back of the fan, so that it could be removed for cleaning purposes or in order to check out the motor design:

 

By the way, Enermax Apollish uses a special bearing called Twister that, according to the manufacturer, combines the long lifespan of two ball bearings with low noise level of the sleeve bearings:

For example, at 1000 RPM this bearing creates no more than 11 dBA of noise and its declared MTBF is 100,000 hours! Very impressive specifications, I have to say.

As for the electrical circuitry, things are not so rosy here: the fan requires 0.35 A current and 4.2 W of power:

However, some of this power is used for extremely beautiful lighting created with LEDs integrated into the inside of the fan frame that starts working at approximately 7.8 V:

 

You can disable the lighting, but who would want to give up this beauty? Besides, why would anyone want to buy Enermax Apollish if, among other things, not for the unique lighting? Let’s try finding out these other reasons that would determine your buying choice in Apollish’s favor. Here are the results of our practical tests:

Well, looks like there are no other advantages Enermax Apollish could offer compared to other solutions tested today, because the fan demonstrated pretty modest airflow and worked quite strangely during rotation speed adjustment. Namely, the speed is increasing smoothly up to 630 RPM and then jumps up to 1020 RPM and after that – to 1650 RPM accompanied with the corresponding noise increase. It must be the thermal diode’s fault. I think it could be better if they made it possible to turn off the diode, rather than lighting. Enermax Apollish can be considered acoustically comfortable at 870 RPM, the bearing doesn’t produce any annoying noises over the entire speed range. The fan consumes quite a lot of power and starts at 10.2 V. Enermax Apollish is priced for retail at $20.

Enermax Everest, Cluster, Magma

Now let’s talk about a more technologically serious family of Enermax fans. Here we have three models with the blades of the same shape, but each with its own distinguishing features. These fans are called Everest, Cluster and Magma and are shipped in clear plastic boxes with an information insert inside:

 

The fans come together with a PATA power connector, four silicone mounting spindles, four self-tapping screws and a manual with installation and cleaning tips. Everest and Cluster models feature LED lighting, but only Everest is bundled with three stickers with yellow, red and blue stripes:

 

Everest and Cluster fans feature a miniature On/Off lighting switch and Everest also has a thermal diode:

 

Besides different technical specifications, the fans have different frames and blades of different color. Everest and Cluster modding fans are of clear and white color, while the most powerful Magma model comes with a black frame and bright-red fan blades:

  

The fans measure 120x120x25 mm, their weight is not specified. Each fan has 9 blades and a rotor with 40 mm diameter. The fans’ power cables are 500 mm long:

The key peculiarity of the fans from this Enermax series are the so-called BATWING-blades. Each BATWING-blade consists of two distinct parts, each looking like an individual small blade with a sail-like curve:

Besides, these two parts of the blade are turned relative to one another. According to the manufacturer, these two-sectional blades generate 30% higher airflow at the same speed as common single-section ones. It is a very impressive promise and we really hope it will prove true.

The fans from this series have different technical specifications. The Everest model works at variable speed changing automatically depending on the temperature coming from the thermal diode from 500 to 1000 RPM. Cluster fan equipped with a PWM controller works at 500-1200 RPM, and the most powerful Magma fan works at 1500 RPM. The airflow and static pressure values are provided in the spec table. The maximum noise from these fans is promised not to exceed 18 dBA.

Enermax uses the same Twister bearing for these fans as in Apollo. The fan blade unit can be removed easily:

 

From the power standpoint, these Enermax fans are far from being called economical: they require a current between 0.25 and 0.40 A and consume between 3 and 4.8 W of power.

  

LED lighting kicks in at 3.3 V on Everest fans and at 3.0 V on Cluster fans.

Enermax Everest, Cluster and Magma fans are priced at $19 each.

Now let’s check out the results:

As you see, Everest equipped with a thermal diode acts strangely, just like Apollo, but it works quieter. The two other models performed better, though they are still far from the leaders in airflow as well as acoustics. Cluster and Magma fans reach their acoustic comfort zone at 960 and 1020 RPM respectively, but the noise generated by the former is a little higher. It must result from the fact that these fans use different frames: the Cluster frame is perforated that creates additional noise. The actual power consumption of these fans turned out lower than the declared one. The startup voltages were 5.7 V for Everest, 3.0 for Cluster and 4.8 for Magma.

GELID Solutions Wing 12 (FN-FW12-15)

Our regular readers should already be familiar with GELID Solutions Company (we have already reviewed GELID Solutions cooler not so long ago). Now let us introduce to you GELID Solutions fans.

GELID Solutions Wing 12 fan ships in a cardboard box with a cut out window on the front side that reveals the rotor sticker, part of the fan blades, four silicone shock-absorbing spindles of fluorescent-green color and fan rotation speed controller:

 

The packaging contains detailed info on the fan technical specifications and peculiarities. GELID Solutions Wing 12 is made in China. I would like to add here that this model series also includes 80 and 92 mm fans.

Inside the cardboard box there is a plastic organizer with separate sections for the fan and accessories:

We have to give our respects to the manufacturer for such thorough approach to packaging of a simple fan. As for the bundled accessories, their variety is also pretty impressive. GELID Solutions Wing 12 comes with water- and dust-resistant regulator with a piece of double-sided sticky tape, a 500 mm cable, four silicone shock-absorbing spindles and a company logo sticker:

The fan boasts very unusual and attractive looks. Glossy black frame goes very well with nine acid-green semitransparent blades of unique shape:

The main focus on the manufacturer’s web-site as well as fan package is made on the fact that GELID Solutions Wing 12 is best optimized for efficient work at low rotation speeds. However, at the same time this fan is positioned as a solution “for gamers and enthusiasts”, who do not always care about the level of noise. Besides, the manufacturer points out that they use high-quality environment-friendly materials to make their fans. But, this is all beautiful words; let’s get down to practice.

The fan measures 120x120x25 mm and weighs 119 g. It has 9 blades and its rotor is 40 mm in diameter. It has a braided 500 mm long cable attached to it:

 

When you look at the fan you notice one thing - vertical blade tip extensions called winglets that get taller towards the forward edge of the blade. The interesting thing is that we have already seen exactly the same winglet on the fans of the new Enermax power supplies – MODU82+ and PRO82+. We didn’t manage to find out who of these two companies is actually making the fans for another one, or if they both get them from a third-party.

According to the manufacturer, the main and only purpose of this winglet is to lower the level of generated noise. It must be achieved due to lower turbulence between the blade and the fan frame, or due to increased fan inertia moment and smoother stroke. Anyway, 25 dBA of noise seems really low for a fan with adjustable rotation speed that can be changed from 600 to 1500 RPM (from 5 V to 12 V) using a compact regulator.  The claimed maximum airflow is 64.3 CFM and the declared static pressure – 1.925 mmH2O, which is not bad compared with the other manufacturers’ solutions working at the same speeds.

GELID Solutions Wing fans are built with Nanoflux Bearings (NFB). They are self lubricating ball bearings with magnetic flux:

According to the manufacturer, NFB should combine all advantages of a sleeve bearing and a ball bearing. The former is known to be very quiet, however, it doesn’t last too long. The latter, as a rule, works several times longer, but is way louder. Thanks to non-abrasive nano-materials GELID Solutions engineers managed to create a universal bearing with guaranteed MTBF of 100,000 hours. Although it is no record, long lifetime like that is definitely quite significant by all means.

The specs claim that the fan can start at 5 V already. Our tests showed that GELID Solutions Wing 12 can start at 5.1 V, which is very close to the claimed voltage setting. At 12 V the current is 0.22 A, which means that peak power consumption of this fan model may reach 2.64 W. Our results turned out considerably more modest.

In addition, fan blades as well as silicone mounts glow in UV light. The recommended price for GELID Solutions Wing 12 fans is pretty high and is set at $19. The fans come with 5 year warranty.

Let’s move on to the obtained test results:

During the tests of GELID Solutions Wing 12 fan we see liner increase of its airflow depending on the rotation speed. The CFM readings for this fan are above average (we are going to offer all comparative summary charts in a separate section later in this roundup). However, we can’t say the same about its acoustic characteristics, because the level of generated noise increases sharply between 800 and 900 RPM, which our noise meter registered immediately. Besides, we start to hear the bearing rumble, though not too distinctly. Moreover, noise level barely increases between 930 and 1050 RPM (it doesn’t seem to get any different from my subjective standpoint), but as the rotation speed keeps growing, the noise level starts to increase linearly. GELID Solutions Wing 12 operates within acoustically comfortable range at 900 RPM. As for its power consumption, this model is quite average and doesn’t stand out in any way, although the obtained numbers are way lower than the claimed specs. I would like to remind you that this fan starts at 5.1 V.

GELID Solutions Silent 12 PWM, TC (FN-P(T)X12-15)

Two next fan models also come from GELID Solutions and are also made in China. However, they belong to a lower price range. The Silent fan series ships in small not very brightly colored boxes with small triangle cut out windows on the front:

The accessories bundle of these fans is also much smaller and includes only four short silicone mounts and a sticker with the company logo:

GELID Solutions Silent fans can’t boast unique or remarkable looks: matt black frame and seven light blades on four plastic rods:

However, despite the simplicity of these fans, we can’t complain about their quality as well as about the manufacturer’s attention to little things (such as long 500 mm braided cable or round rods, for example). The fans measure 120x120x25 mm and weigh 130 g. Each of them has seven blades and a 43 mm rotor (in diameter):

 

The blades of these fans don’t have anything innovative about them and look pretty common.

The fans rotation speed can be adjusted from 750 to 1500 RPM, but in different ways. As you may have already guessed, the fan with “PWM” in the model name uses pulse-width modulation method, while the one with “TC” index is controlled by a thermal diode. The maximum airflow for the fans of this series is claimed to be 58 CFM and the noise level varies between 12 and 25.5 dBA. In other words, Silent fans shouldn’t be any louder than Wing 12 fans at the same rotation speed. However, they should have lower static pressure – it is claimed at 1.6 mmH2O.

The fans from Silent series are built using fluid dynamic bearings with 50,000 hours MTBF.

Silent 12 fans have slightly lower current and power consumption than Wing 12 fans: 0.18 A and 2.16 W respectively. Their recommended retain price is pretty low: $12 for the PWM model and $13 for TC one. They come with the same 5-year warranty.

Let’s take a look at the test results obtained for Silent 12 fans:

Although GELID Solutions Silent fans are cheaper than Wing, their airflow is almost the same and is also linearly dependent on the rotation speed. However, these models run noisier. These fans perform in acoustically comfortable range only at 810 RPM and 6.6 V voltage, but the good thing is that we noticed no rumbling of any kind at any speed. Of course, since they used a different bearing type (fluid dynamic bearings usually work very quietly). GELID Solutions Silent refused to work at 3 V. The startup voltage setting was registered at 4.2 V.

GlacialTech GT12025 SDLA1, EDLA1 and BDLA1

Three budget 120 mm fans from GlacialTech Company are among the most popular and widespread retail solutions due to the good reputation of the manufacturer as well as pretty democratic prices. We are very curious to see how well they will be able to compete against more expensive and distinguished solutions. Maybe we should spend that much money on fans after all?

The fans manufactured in China come in simple OEM packaging without any accessories. They look quite simple: black frame and black blades sitting on four supporting rods:

These fans are 120x120x25 mm big and their weight doesn’t exceed 147 g. each fan has 7 blades and the rotor is 40.5 mm in diameter. The fans work with constant speed of 950 RPM (±150 RPM). At this speed they should create 0.75 mmH2O static pressure, 36 CFM airflow and no more than 18 dBA of noise.

 

The fans with pretty unpretentious exterior have aggressively curved blades with rather large sweep area that expand from the rotor towards the frame:

The difference between the fan models within this series is only in the bearing type. There are four different bearings used for these fans: SLDA1 - sleeve bearing, BDLA1 - double ball bearing, EDLA1 – enhanced sleeve bearing, and HDLA1 – hydrodynamic bearing. Unfortunately, we didn’t manage to get a fan with the latter bearing type:

  

The sleeve bearing MTBF is 26,000 hours; the double ball bearing MTBF is 50,000 hours, the enhanced sleeve bearing MTBF is declared at 35,000 hours and the fan with a hydrodynamic bearing should last at least 80,000 hours. The interesting thing is that GlacialTech doesn’t differentiate between the fans according to their noise levels depending on the bearing type used, because the declared noise for all fan models is the same - 18 dBA. The fans also do not differ in their official power specs: all of them can work at voltages between 7.5 and 13.8 V, support 0.12 A current and consume no more than 1.44 W of power. At the same time, our measurements showed that the fans differ indeed significantly in power consumption and acoustic readings, which depend on the type of the bearing employed.

Take a look yourselves:

GlacialTech fans featuring almost the same airflow characteristics turned out dramatically different in noise levels. The solution with a sleeve bearing didn’t appear the quietest – it reaches the acoustic comfort zone only at 730 RPM. We can hear the bearing rustling over the entire speed range. The ball bearing model made even louder noises, but it was still ok to tolerate at 810 RPM. The quietest of all was the EDLA1 model that was free from any bearing defects. Its speed in acoustically comfortable mode was 930 RPM. The airflow readings of all fans were a little above the average of the today’s testing participants. Their power consumption is very low, which means that they must have played it safe in the specifications. The startup voltages are quite high for all fan models: SLDA1 – 10.2 V; BLDA1 – 9.0 V; EDLA1 – 9.3 V.

Nanoxia FX12, FX+12

The next four fans come from the German Nanoxia Company that is a sub-division of the large Xen Micro GMBH. All fans ship in identically designed plastic boxes. You can see the fan inside through transparent front of the box:

 

There is a lot of information on the package starting with the list of key fan features and finishing with their detailed specifications. FX+12 models differ from FX12 by PWM support and corresponding rotation speed control function. Therefore, only the rotation speed regulator is different:

 


Nanoxia FX12


Nanoxia FX+12

Besides the regulator, these fans are also bundled with four silicone mounts. They are of similar acid-green color like the ones from GELID. These fans are made in China.

At first glance, it seems that Nanoxia fans are very similar to GELID Wing fans: fan blades, frame and even 500 mm braided cable are of the same color and look alike. However, if you take a closer look at them, you will notice that FX(+)12 has seven blades instead of nice and they don’t have the winglets on their tips:

In fact, Nanoxia FX(+)12 fans do not have any distinguishing features (except for the unique color), so let’s check out what the manufacturer has to say about them. It turns out that the fan blades are made from high-tech Makrolon polycarbonate – very durable, water-resistant and capable of working in wide temperature range for a long period of time. That’s pretty much all they have to say about the peculiarities of the fans design :)

The fans measure 120x120x25 mm. Their weight is not specified. Each fan has 7 blades and a 40 mm rotor:

 

Speaking of the fans specifications we have to point out that there are two models within each fan type: one with 380-1250 RPM rotation speed, and another with 600-2000 RPM rotation speed. Maximum airflow for the first model is claimed to be 47.39 CFM and for the second – 79.14 CFM, their noise levels – 12-17 dBA and 17-33 dBA, and their static pressure – 1.11 and 1.57 mmH2O respectively.

As I have already said, it is hard to find anything unique about the fan blades:

Nanoxia FX(+)12 fans use ball bearings made using nanotechnology and minimum of abrasive materials. Due to special bearings design, the fans are water- and dust-resistant and should last up to 150,000 hours (at 25 °C). Very impressive number, I should say. And there is no reason for us not to trust the manufacturer on this one, as they provide their solutions with a 10-year warranty! The manufacturer claims that all Nanoxia fans are balanced manually, however, we didn’t see any traces of this on the fan blades.

 

As for power consumption, Nanoxia also did a great job here. Besides low startup voltage of 4 V, the fans consume comparatively little power. For solutions with 1250 RPM maximum rotation speed the current is at 0.1 A and maximum power consumption shouldn’t exceed 0.6W. As for the solutions with 2000 RPM maximum rotation speed, these parameters equal to 0.28 A and 1.92 W respectively. They owe this low power consumption to the ECO-motor that was developed together with X-Spice Technology Co. Ltd. Of course, all these technologies and high-end materials affected the price of the fans. The 1250 RPM model sells for $16, and the 2000 RPM model – for $19.

Let’s take a look at the obtained test results for the Nanoxia FX(+)12 fans:

The junior Nanoxia fan lost to the previous two GELID fan models in airflow readings, while the higher-end fan turned out significantly ahead of them. Nevertheless, the results of Nanoxia fans airflow measurements fall into the middle category. However, when it comes to noise levels, both Nanoxia fan models look way better than GELID solutions (certainly, the top one only up until ~1500 RPM). Nanoxia FX12-2000 fan is one of the airflow leaders of our today’s test session. We didn’t notice any parasitic sounds during work: nano-bearings and fan motor deserve a good B+ here. We reach acoustic comfort zone at 900-930 RPM. As for the power consumption, Nanoxia doesn’t have anything to boast here, because the results we obtained in our lab are almost twice as high as what the manufacturer claimed (the results demonstrated by PWM models are identical to those by regular FX12 fans. The startup voltages are 3.6 V for the FX12-1250 fan and 2.1 V for FX12-2000 fan.

Nexus Basic (D12SL-12)

The next fan definitely won’t leave you unfazed. So, if you have been looking briefly through fans descriptions and only checking the test results up until this moment, I strongly urge you to take your time and read about the Basic fan from Nexus, because it is absolutely worth every minute of your time.

So, the fan comes in pretty common package – a small clear plastic box. There is a woman’s head on both sides of it that either suggests that we should all keep quite or symbolizes an extremely quiet device inside:

Among the really useful data available on the box there are fan rotation speed, airflow and noise. There is nothing else there. The fan is bundled only with a three-pin → PATA adapter. The latest fan revisions also include silicone mounting spindles. All fans are made in China. Nexus Basic comes with a 3-year warranty.

At first glance there seem to be no obvious distinguishing features, unique traits or innovative technologies. Nexus Basic is available in two color schemes: solid orange color or black frame with a white fan. We got the second modification:

Nevertheless, the manufacturer claims that Nexus Basic is the “world’s quietest fan”. Don’t think that this fan works at some 200 RPM speed. It is quite normal for a 120 mm fan and equals 1000 RPM. So, what makes the manufacturer make such ambitious claims? Let’s try figuring this out.

The official company web-site offers all the details about their solutions including acoustic tests results and photographs. Moreover, Nexus stresses for those who check out their web-site that many fan makers out there do not bother to perform adequate acoustic tests in specially equipped anechoic test chambers, but simply extrapolate the old test results onto their new solutions with similar specifications. Self-promotion and competition aside, I have to admit that this does in fact make sense. You may also have been surprised before when a 2000 RPM solution was claimed to have a noise level of only 16 dBA. Of course, these numbers are unrealistic and shouldn’t be taken seriously.

But let’s get back to our reviewed fan. It measures 120x120x25 mm and weighs 123 g. Each fan has 7 blades and a 40 mm rotor. Note that Nexus Basic has one of the smallest rotors of all 120 mm fans. Only Enermax fans have a rotor of the same diameter of 40 mm, and the smallest rotor belongs to Scythe Slip Stream fans – 36 mm, and SilenX iXtrema Pro IXP-74-09/11 – 33 mm. moreover, the rotor sits on four supporting rods, three of which are round and only 3.5 mm in diameter:

 

So, we see the Nexus engineers intended to maximize the effective sweep area of the fan blades and minimize the airflow resistance. The fan blades are quite classical in design, aggressively angled, with a 2.5 mm gap between the blade tip and the inside of the fan frame:

They are very similar to the blades of GELID Silent fans, but are more curved. There are no notches, winglets or dimples of any kind. At 1000 RPM the fan should create 36.87 CFM airflow – it is the closest declared parameter to what we obtained during our test session. The noise in this case shouldn’t exceed 18 dBA.

It is interesting that there is barely any info on the official site about the bearing employed in this fan. The only thing we managed to find out is the type of the bearing used – common sleeve bearing. We don’t even know its MTBF. Looks like Nexus didn’t think this information was important and deserved special attention, although they should, because case fans are replaced several times less frequently than CPUs or graphics cards.

The startup fan voltage according to the specifications is 7 V, while the actual one is 6.5–13.8 V.

It is really interesting that the UL number on the packaging belongs to Yate Loon Company that does in fact have a D12SL-12 model in their product range. Yate Loon fans are not very well-known among users, they are most often seen in mainstream power supply units, but have pretty good reputation. Besides, Yate Loon products are also quite inexpensive: 120 mm fan from this maker usually costs about $3-$4, although they usually retail faster and noisier D12SH-12 models).

Maximum fan power consumption shouldn’t exceed 3.6 W at 0.3 A of current. I would also like to add that Nexus Basic MSRP is set at $15.

Finally, here are our test results:

If anyone asked me to name the best fan from the prospective of subjective acoustic comfort (without the noise meter readings taken into account), I would undoubtedly choose Nexus Basic. It works unbelievably quietly (I am not using the term “Noiselessly” because I will be inevitably bashed by some maniacs of absolute quiet :) ). Look at the straight line of the noise level: you won’t hear this fan at all up until 840 RPM. After that you can hear the airflow created by the fan, but not the fan itself. By the way, the airflow is pretty significant, even though it is a little lower than by the leading solutions of our today’s test session. Good job, Nexus! Very modest power consumption, but the startup voltage of 6.9 V could have been lower.

Nexus PWM (D12SL-12 PWM)

According to the information on the official company web-site, Nexus, inspired by definite success of their Basic series, released another 120 mm fan model with a simple and logical name – Nexus PWM. The box of this new fan is very similar to that of the Basic series, just the colors and the information on the package are different:

The fans are bundled with four silicone mounts of acid-purple color, but no PATA power connector:

The fans are made in China and come with 3-year warranty, just like the Basic fans.

Nexus PWM looks more interesting than the previous fan model, because the fan itself and the frame are made of clear plastic:

Nevertheless, there is no LED lighting, although it could make this fan even more beautiful and competitive. But, it is no big deal, serious products very rarely come with modding-friendly features anyway. The fan measures 120x120x25 mm and each has 7 blades. The rotor is 45 mm in diameter:

 

The supporting rods are now a little bigger. Nexus PWM fan blades are dramatically different from those of the Basic series: now they are expanding away from the rotor and end with a distinct “tip”:

Nexus doesn’t explain why they have changed the blades shape so dramatically, but focuses mostly on the automatic adjustment of the rotation speed between 500 and 2000 RPM using pulse-width modulation method (PWM). The maximum airflow from this fan makes 76 CFM, which is a pretty typical number for this rotation speed. The noise level has, definitely increased and now varies from 15.5 to 26.9 dBA. Nevertheless, it didn’t prevent Nexus from calling their fan “the quietest in the world” again. A quick comparison of the other participating 2000 RPM fans’ acoustics did, however, show that the claimed noise spec is in fact the lowest of all.

Nexus PWM’s startup voltage is claimed to be at 6.5 V, which is 0.5 V lower than that of the Basic fans. The operational voltage interval remained the same, but power consumption decreased and now makes 2.04 W at 0.17 A current, although the sticker on the fan rotor claims the same 0.30 A:

In conclusion I would like to add that Nexus PWM is priced at $18. Now it is time to check out the test results and see how much better or worse the PWM model is compared to Basic:

Well, looks like Nexus PWM has a completely different character. Yes, this fan generates high airflow; yes, it consumes not so much power even at maximum rotation speed; yes, it features PWM controller and looks attractive. However, the acoustics is really poor. The fan doesn’t crackle too loudly, but this crackling is stable and exists over the entire operational range. As a result, it can be considered more or less acoustically comfortable at 900-930 RPM, which is 100 RPM lower than by Nexus Basic. Therefore, it loses to the Basic series in airflow at identical noise level. Low startup voltage of only 3.6 V will hardly be a consolation for the Nexus PWM owners.

Noctua NF-S12B ULN, FLX

The next two new fan models that we will review and test come from the Austrian Noctua Company known for their scientific approach to any cooling system component, such as a heatsink plate or a fan blade. Frankly speaking, we can’t really call Noctua NF-S12B ULN, FLX fan new, because they are in fact enhanced and improved modifications of the NF-S12 fan that has already been discontinued.

The boxes are designed in Noctua’s traditional manner: highest quality and maximum functionality. It has everything: a cut-out window in the cardboard panel to attract the potential customer, a plastic casing inside that should protect the fan against transportation and other damages, an in-depth description of the fan features, including its detailed specifications:

Moreover, the back of the box can be unfolded to reveal even more details on the fan features and employed technologies:

Noctua NF-S12B FLX fan comes bundled with four silicone mounting spindles, four self-tapping screws, a PATA power connector and two adapters with built-in resistors lowering the voltage to 7 V and 5 V respectively:

The ULN accessories bundle is exactly the same, but without the 5 V adapter, not required for this particular fan model.

The fans can’t boast remarkable exterior, although they do look unique against the background of black and clear solutions. The fan frame is of light-beige color, while the fan itself is dark-brown:

The fan is 120x120x25 mm big, its weight is not specified. Each fan has 7 blades and a rotor with 41 mm diameter. The braided cable that comes with it is 400 mm long:

 

The key feature distinguishing the new fans from the old NF-S12B is the so-called “Bevelled Blade-Tips”. The front side of the fan blades is longer, while the tips have three specific notches on them:

According to Noctua engineers, this blade design increased the airflow from the fan as well as the pressure by 10%. Noctua engineers claim that the new NF-S12B FLX and ULN work quieter and create higher airflow than NF-P12 fan model, although I have to say this is hard to believe. However, FLX and ULN should actually lose to NF-P12 solution in static pressure, which is quite logical keeping in mind the shape of the blades.

ULN fans can work at 1200, 900 or 600 RPM depending on the resistor. The FLX version supports 700 and 500 RPM modes. As usual, Noctua provides the airflow values in cubic meters per hour. In cubic feet per minute these values look as follows: 59 CFM for ULN and 33.4 CFM for FLX at maximum rotation speeds. The static pressure is promised at 1.31 mmH2O for the top model and 0.44 mmH2O for the junior one. The latter should generate maximum 6.8 dBA of noise, while the ULN modification of the NF-S12B fan can produce as much as 18.1 dBA.

The new fans use the same bearing called Noctua SSO bearing, which is a fluid dynamic bearing with a magnet. A bearing of this kind equipped with a magnetic stabilizer of the rotor axis increases the fans MTBF up to 150,000 hours. This way Noctua can provide 6-year warranty for their fan products.

 

As for the electrical specs of the NF-S12B FLX and ULN fans, we could point out their low current and power consumption. The ULN model requires only 0.05 A and 0.6 W, while the faster FLX model needs precisely twice as much. The new fans are pretty pricy and go for $25 each.

Since Noctua themselves compare NF-S12B FLX and ULN against NF-P12 fan, we decided to add the results for the latter into our charts for a more illustrative comparison:

The new fans turned out better than NF-P12 in airflow. As you can see from the charts, despite the developers’ claims, the nine-blade model with larger sweep area pumps more air through than a fan with seven flat blades, even though they feature an enhanced tip. However, in terms of acoustics, NF-S12B FLX and ULN outperform their fellow fan. And while we have no complaints about the acoustic comfort provided by the ULN modification, the FLX one generates low-frequency rumble that is most likely generated in the gap between the blades and the fan frame. The bearing is extremely quiet. The subjectively comfortable acoustic range for the NF-S12B FLX model is around 900 RPM, while ULN model works quietly within the entire supported operational range. The power consumption of both models is really low and is very close to the declared values. Fans startup at 3.9 and 4.5 V, while NF-P12 required at least 6.6 V.

Noiseblocker NB-BlackSilent, XL1, XL2, XLP3 Rev.3.00

Another group representing German fan makers consists of three models in NB-BlackSilent series from the Noiseblocker Company, which name speaks for itself. The fans are packed in clear plastic casing with a cardboard insert at the bottom:

The detailed technical specifications of the fans are provided on the back of the packaging:

  

The fans are shipped with four long silicone mounting spindles that should absorb vibrations during work and slightly reduce the level of generated noise:

All three models are manufactured in Germany, belong to ultra-quiet series and third revision (2008 models). XL1, XL2 and XLP fans differ by their supported rotation speeds, and the latter model is also equipped with PWM rotation speed regulator.

NB-BlackSilent series fans look very attractive due to crescent-shaped semitransparent blue blades:

The fans measure 120x120x25 mm, weigh 120 g and 125 g for the PWM model. Each fan has seven blades and a 45 mm rotor diameter:

 

The rods holding the fan blades are very thin and their top part has semicircular cross-section. Both these things help, even though just slightly, to reduce the airflow resistance and to lower the noise.

Unlike all 120 mm fan models we have already discussed above, the inside of the NB-BlackSilent fan frame has no sharp edges, which should also have a positive effect on the acoustics:

XL1 fans rotate at 1000 RPM (±10%), XL2 fans – at 1500 RPM, and XLP fans equipped with a PWM rotation speed controller can work in the interval between 1000 and 2000 RPM. As a result, these fans generate 41, 58 and 75 CFM airflow and produce 13, 21 and 14-21 dBA of noise respectively, which is unusually low for rotation speeds like that. At this point we can only add that Maximum static pressure is claimed at 0.659 mmH2O for XL1, 1.044 mmH2O for XL2 and 2.653 mmH2O for XLP.

NB-BlackSilent fans use sleeve bearings with guaranteed MTBF of 80,000 hours, which is more like the lifespan of solutions using ball bearings than a common sleeve. In terms of power consumption, these fans can’t boast anything special and don’t support low startup voltage (only 6V).

  

XL1 and XL2 fans should consume no more than 3.24 W at 0.27 A current, while the power consumption of the XLP model varies depending on its rotation speed between 0.6 and 4.8 W with 0.05-0.4 A current. The recommended retail price for these fans is set at $13 for XL1 and XL2 models and $17 for XLP. They come with 3 years of warranty.

In terms of practical performance, NB-BlackSilent fans did very well and were among the leaders of our today’s test session:

I was especially impressed with the level of noise generated by XL1 fan model, namely, with the absence of any noise in the entire rotation speed range! It is a perfectly balanced fan, with an absolutely noiseless bearing and extremely successful combination of the fan blades shape and fan frame with smooth internal surface. The latter two features are also true for the other two models. They do run noisier, of course, but only because of the extremely high airflow they create, which is, actually, one of the highest in the entire test session. NB-BlackSilent fans reach acoustically comfortable performance level at 1020 RPM. They do not consume much power, even less than the specs claim. Their startup voltages are 3.6 V for XL1, 2.1 V for XL2 and 4.2 V for XLP.

Noiseblocker NB-Multiframe S-Series MF12-S1, -S2, -S3HS, -P

Another four Noiseblocker fans we are going to talk about today belong to the NB-Multiframe S-Series MF12 and are shipped in small cardboard packages with partially open front panel:

 

 

The package states that these fans run 30% quieter than standard 120 mm fans (it doesn’t specify which ones). NB-Multiframe S-Series were designed by German engineers but are manufactured in China. They come with twice the warranty of 6 years.

These fans measure 120x120x25 mm and weigh 150 g and 155 g. Each fan has 7 blades and a rotor with 41 mm diameter:

The blades of NB-Multiframe S-Series MF12 fans are shaped almost the same as those of the NB-BlackSilent fans, but the actual frame is different and we are going to get to its description shortly. The supporting rods are very thin, except the cable one, and their side facing the airflow is of semicircular shape. The cable is 440 mm long. The fan blades are of semitransparent gray color, which makes the fans look more serious, I would even say, stylish:

 

The models differ from one another by their rotation speeds and all other characteristics connected with them. The junior model with S1 index works at a constant speed of 750 RPM , S2 – at 1250 RPM, S3HS – at 1800 RPM, and P – feature a PWM rotation speed regulator and its speed varies between 1000 and 2000 RPM.

The airflow created by these fans is 34, 51, 73 and 42.8-78 CMF for all models listed above respectively. Their static pressure equals 0.45, 1.11, 2.302 and 2.838 mmH2O. The most interesting parameter of the NB-Multiframe fan series is the level of generated noise, to be more exact, very low level of generated noise. For example, they claim that the S1 model produces only 8 dBA of noise, which means that you won’t be able to hear it from a 1m distance even in absolute silence. And that is at 750 RPM speed! Some fans working at exact same speed rattle like an old tractor, and this one should be completely quiet. I wonder if this is really the case. For the S2 model they claim 19 dBA of noise, for S3HS – 27 dBA and for P in the interval from 12 to 29 dBA.

The distinguishing feature of NB-Multiframe fan series is advanced anti-vibration system. It uses not the traditional silicone spindles, but four soft rubber triangles that should be inserted into the corners of the fan frame:

I have to say that it is a pretty unique and actually more convenient solution than the traditional mounting spindles, but only if you are going to use it as a case fan. You may have some difficulties attaching these fans to CPU heatsinks (take for instance the retention clips from Thermalright or Scythe).

However, I think that the key peculiarity of these fans is not the soft shock-absorbing triangles or extremely smooth blades, but the bearing employed in the rotor. NB-Multiframe fans use ultra-quiet NB-NanoSLI bearings designed by the German Company themselves. Of course, no details are revealed. And they have every reason to keep the technical details secret, because these bearings together with only one more fan model proved to be the quietest of all (I would even say completely noiseless), which is totally amazing.

 

 

The bearing employed in the S1 fan model is guaranteed to last 160,000 hours. The one in S2 model will last a little less – only 140,000 hours. And the bearings in both higher-speed models have 130,000 hours bearing MTBF. S2, S3HS and P models can start at only 4.5 V, while the slowest S1 model will need a little more than that – 6V. However, according to the official specifications, NB-Multiframe fans are not that energy-efficient at all, because even the slowest model requires 2.74 W of power. The actual power consumption rates again turned out lower than the declared ones. Noiseblocker fans from NB-Multiframe series are priced between $20 and $26 depending on the model. It is very expensive for fans, but they are worth every penny.

Now let’s discuss the practical measurements results. Check out the diagrams below:

Our verdict about the previous Noiseblocker fan series was highly positive already, but the NB-Multiframe series is divinely good. They run even quieter than the NB-BlackSilent fans, but their airflow is at least as good. Amazing fans, truly ones of the best out there! They reach acoustical comfort zone at 1080 RPM. In terms of power consumption of their NB-Multiframe series, Noiseblocker deceived us, as the slowest model doesn’t any anything even close to 2.74 W and can work just fine with only 0.5 W of power. The declared specs for the higher-end models are closer to reality, as you can see. The startup voltage for the “S” models is 4.2 V, while the PWM modification of the same fan requires 4.5 V (exactly like the specs claim). NB-Multiframe fans refused to work at anything lower than that.

Scythe S-FLEX (SFF21E)

One of the leaders in the air-cooling segment, the Japanese Scythe Company, is represented by five different fan models in our today’s roundup. Three of them are brand new solutions, while the other two should already be familiar to you from our previous articles, so we simply added them to the test session. The first model we will talk about belongs to the S-FLEX series. We got only one fan like that without the appropriate package and accessories. The official company web-site has some pictures of the S-FLEX fans packaging. They come in clear plastic blisters with cardboard inserts:

These fans are made in China and are priced relatively high: at $19.99.

Scythe S-FLEX fans looks quite common. I would even say that they look very unpretentious and even too simple. Matt black frame and black fan blades, that’s all it is:

The fan measures 120x120x25 mm, the weight is not specified. Each fan has 7 blades and a rotor 48 mm in diameter. Four oval-shaped rods hold the motor and fan blades. The smallest diameter of all rods is 4 mm except for one that has three-pin cable hidden inside:

 

S-FLEX fan lineup includes four fan models with rotation speeds ranging from 800 to 1900 RPM. We managed to get our hands on SFF21E model working at 1200 RPM that creates 49 CFM airflow. As usual, Scythe doesn’t mention the fans static pressure anywhere in the specs, considering this parameter to be secondary. The level of noise generated by this particular fan model shouldn’t exceed 20.1 dBA.

The blades curve slightly and have rounded edges. The gap between the tip of the blade and the inside surface of the fan frame doesn’t exceed 2 mm. I noticed that the fan rotor is of pretty big size. It seems that S-FLEX won’t be able to create a significant airflow.

However, these fans do have one peculiarity about them. It hides in the center of the fan. I am talking about the S-FDB bearing that stands for SONY Fluid Dynamic Bearing. As you may have already guessed, this bearing is developed by a well-known Japanese corporation, so only complete skeptics or those who have been plain unlucky with SONY products could doubt the high quality and extensive lifespan of this bearing. This is what its internal structure looks like:

It is a sleeve bearing with the bush sitting inside the metal ring that prevents it from drying out. We you understand, this bush will last way longer being permanently oil-lubricated. Moreover, there is an additional rotor suction magnet built-in to ensure better rotor centering. The motor is regulated by brand name Linear Drive controller. All this allowed the engineers to significantly lower the level of generated noise and reach the claimed 150,000 hours of fault-free fan operation (I wonder if anyone has ever checked that out?).

This fan should be pretty energy-efficient, because its claimed peak power consumption shouldn’t exceed 1.8 W at 0.15 A current. This Scythe S-FLEX fan model is declared to startup at 4.5 V max.

Since we do not expect Scythe S-FLEX to impress us with its airflow readings, the most interesting part of our test session will be the noise level. How did the use of S-FDB bearings pay back? Let’s find out:

I am telling you, S-FDB proved totally worth it! Scythe S-FLEX fan turned out one of the quietest fans of our today’s test session. Up to 1000 RPM this fan works in acoustically comfortable range for a mainstream user. You can’t hear its bearing in the entire operational range. However, despite the fact that there are only seven blades on a rotor of bigger diameter, this fan creates good airflow that yields just slightly to the leaders of our today’s test session. What an excellent fan! Extremely low power consumption and only 4.8 V startup voltage. Please, send me the entire S-FLEX series now! :)

Scythe Gentle Typhoon

The next family of Scythe fans that started selling in the end of last year is called Gentle Typhoon. And while the first word in the model name suits this solution pretty well, the second one doesn’t. But let’s start from the very beginning.

120 mm fans from Scythe Gentle Typhoon family are available in five different modifications and all of them will participate in our today’s test session. They ship in similarly designed boxes with different information printed on them. The plastic casing on the front of the box reveals the entire fan. You can also see the fan rotor through a cut-out round window in the back (below are the photos of the senior and the junior models in the lineup):

 

The package contains a lot of useful info about the fans inside, including not only their technical specifications, but also detailed description of the fan bearing with schematics. Gentle Typhoon fans are made in Indonesia and are priced at $16.96. Each fan is bundled with a Molex power adapter and four self-tapping screws.

The fans were originally designed by the Japanese Nidec Servo Company that have been working closely with Scythe for a long time now. Gentle Typhoon fans look very unusual and stand out among other fans due to black fan frame and nine dramatically curved light-gray blades attached to a large rotor:

They measure 119x119x25 mm and weigh 200 g. As I have already said, each fan has 9 blades and a rotor with 52 mm diameter. Just like all other developers, Nidec Servo Corporation tried to accomplish two goals when they designed Gentle Typhoon fans: to lower the noise and to increase the airflow intensity. According to the manufacturer, these fans are as efficient as 32 mm fans due to more blades with large sweep area:

 

Besides, unlike other Scythe fans, the inside of this fan frame is absolutely smooth, without any edges, and the side of the frame facing the incoming airflow has been rounded even more to lower the level of generated noise. Unfortunately, they didn’t manage to make the supporting rods thin and rounded because of pretty serious cable that goes into one of them, which slowed down the airflow a little.

Another peculiarity of Scythe Gentle Typhoon fans are the notches on the inside of the fan blades right next to the rotor:

The manufacturer didn’t explain their purpose, but I can only guess that these small notches around the fan rotor should probably create turbulent airflow. But for what purpose and why in this particular area? These notches are way too small to really affect fan operation in any way. They could probably lower the motor workload…

The rotation speeds of the fans in this lineup vary from 500 to 1850 RPM depending on the model (with 800, 1150 and 1450 RPM models in-between). The declared airflow is measured in cubic meters per hour and ranges from 31 to 99, which translates into CFM as 17.7-56.5 CFM. As you can see, there is nothing supernatural about the airflow by Scythe Gentle Typhoon fans. There is no mention of the fans’ static pressure, but they do reveal the noise readings: 5 dBA for the junior model and 28 dBA by the senior one. Taking into account that the top model works at 1850 RPM rotation speed, the claimed 28 dBA of noise is very little.

Scythe Gentle Typhoon fans have the largest rotor diameter of all fans tested today – 52 mm. It is more likely the result of a double ball bearing employed in them, each balanced manually. Together with the frame and fan blades design, the bearing was built to produce noise in a specific acoustic range that is not perceived by the human ear.

Besides, ideally balanced fans shouldn’t transfer vibrations onto the case parts. The manufacturers claim that Gentle Typhoon fans won’t stand out against the noise of the working system power supply (though they don’t state which one). The bearings have 100,000 hours of MTBF at 30 °C or 60,000 hours at 60 °C, which is more than enough for these fans to outlive several system cases.

They haven’t forgotten to mention the fan electronics either. The manufacturer claims that Gentle Typhoon consumes 30% less power than other fans due to high efficiency of the motor. According to the technical specifications, the developers did succeed here, because even the top model in the family should consume no more that 1 W at 0.083 A current. And that is at 1850 RPM rotation speed! It is very impressive, although not so utterly important, considering how tiny this number is against the background of the rest of the system case. The fans startup voltage is the following: 500 RPM – 6.5 V, 800 RPM – 5.5 V, 1150 RPM – 4.5 V, 1450 RPM and 1850 RPM – 3.5 V.

There is a sticker on the rotor of each fan that has the fan model number and some other details printed on it:

  

 

In conclusion to our review of Scythe Gentle Typhoon fans I would like to add that the fan frames have short stands in the corners between the retention holes. These stands may become a problem during fan installation onto cooler heatsinks that require attaching the fans by the inside of the fan frame opening. Besides, the 300 mm cable could have been a little longer. The fans are priced at about $17.

Let’s take a look at their performance:

Well, these are 5 more great fans from Scythe, no doubt about that. Gentle Typhoon demonstrated excellent acoustics, so they can work at higher rotation speed generating about the same noise as their competitors. For example, they work within acoustically comfortable range at about 1440 RPM – an unattainable speed for other solutions. Had their rotor been not so humongous, they could have become indisputable leaders not only in acoustics but also in airflow tests. And at this point, unfortunately, 52 mm rotor eats away the effective sweep area from the fan blades that is why Scythe Gentle Typhoon becomes only one of the best, which is also quite rewarding. We didn’t register any parasitic noises coming from the motor or bearing during the entire test session. Only when we changed the fan rotation speed, we could hear some short-term clanging sound that would disappear only when the rotation speed became constant. Scythe Gentle Typhoon fans boast undoubtedly the best power consumption: even the most powerful model consumes less than 1 W of power. The interesting thing is that the fan blades keep spinning for about 30-35 seconds once the power has been shut off – it is a great illustration of how well-balanced the bearing is. The startup voltages for all five fans are: 6.9, 5.4, 3.0, 3.9 and 3.6 V.

Scythe Kaze Maru

One more series of fans from a well-known Japanese Scythe Company is called Kaze Maru. These fans first came out in mid 2008; however, they haven’t yet become very popular among computer users. Although I think everything is still ahead of them. They are shipped in Scythe’s typical packaging: clear plastic box with a cardboard sheet inside:

 

Besides the fan, there is also a three-to-four pin PATA power connector and four self-tapping screws:

These fans are made in China.

Scythe Kaze Maru fan family includes three 140x140x25 mm models. These fans look very powerful and serious:

Eleven sharp crescent-shaped blades come out of the rotor with 49 mm diameter that sits on four curved supporting rods with a trapezoid cross-section. The fan frame is round with flat inside surface. It has four retention loops attached to it on the outside at 120 mm distance from one another. In other words, 140 mm Kaze Maru fans can be installed into 120 mm fan spots and that is the key peculiarity of these fans. However, in this case the frame should also fit into the 120 mm slot. This is certainly no issue for CPU coolers, but in system cases Kaze Maru may not always fit into the available fan slots because of the frame size.

Kaze Maru fans are represented by three models with 500, 1200 and 1900 RPM rotation speed. It is pretty strange that there is no 800 RPM fan in this family, because I think that this could be the most popular model offering moderate noise levels and impressive airflow for the fans of this speed. Kaze Maru fans weigh differently: 131, 137 and 175 g for the junior, medium and senior models, respectively.

 

The airflow from a fan working at 500 RPM should make 29.5 CFM, at 1200 RPM – 61.3 CFM and at 1900 RPM – 104.5 CFM! As for the noise, only the junior model can be considered quiet enough, as its declared noise level is 10.8 dBA. Faster fans will generate 26.6 and 41.1 dBA, respectively.

Unfortunately, there is no mention of the fans static pressure in the official specs. I think a fan of this size and with blades shaped like that should create pretty high static pressure:

As for the bearing and electronic components of the Scythe Kaze Maru fans, there is nothing special here. This model uses common sleeve bearing with 30,000 hours MTBF. The fan model number and supported current are printed on the rotor sticker:

 

The maximum power consumption of these fans is as follows: 500 RPM – 0.48 W, 1200 RPM – 2.28W, 1900 RPM - 2.88 W. The startup voltage for the 500 RPM model is declared at 8 V, and for 1200 and 1900 RPM models – at 3.5 V. The power cable is 300 mm long, but as you know, there is an adapter included with the fan that will add about 400 mm to that. Scythe Kaze Maru is priced at only $8.

Well, 140 mm fans from Thermaltake called TurboFan didn’t do too well in our tests. Let’s see if Scythe Kaze Maru fans will confirm the benefits of having a larger fan in your system. Here are the results:

Kaze Maru fans made a mixed impression. Firstly, they turned out different in the level of noise they generate and that comes not only from the difference in rotation speed. The junior and medium models works absolutely fine without any parasitic noises, which was not the case for the top model: we could clearly hear rumbling of its motor/bearing. Secondly, Kaze Maru failed to impress us with their airflow readings. They even lost to their 120 mm brothers and to 140 mm Thermaltake fan (working at the same speed, of course). Only the top Kaze Maru model performed well enough, though at a higher noise level. These fans were within the acoustically comfortable range at 780 RPM (excluding the rumbling model, of course). And as for power efficiency, only the junior solution could be considered economical. The fans startup voltages are 3.6 V, 2.1 V and 4.2 V, respectively. But still, why wasn’t there an 800 RPM fan among them?

SilenX iXtrema Pro IXP-74-09/11

SilenX fans are not as well-known as SilverStone or Evercool solutions, but that made testing them even more interesting. We are going to introduce to you four models from this manufacturer, which were split in two pairs according to their dimensions. First come 120x120x25 mm fans.

Their packaging is of most common design. It consists of a cardboard sheet covered with clear plastic casing:

The cardboard sheet tells you about the “revolutionary” shape of iXtrema Pro fan blades and lists its brief technical specifications. Besides the fan, potential customers also get a PATA power connector, four self-tapping screws and four silicone mounting spindles as well as a brief installation guide:

Just like most fans tested today, SilenX iXtrema Pro IXP-74-09/11 fans are made in China.

We have already mentioned the fans dimensions. As for their weight, the manufacturer doesn’t reveal it, but subjectively, they are the lightest fans of all tested today (less than 80 g). Therefore, if this parameter is important for you, then SilenX iXtrema Pro is the way to go. Each fan has 7 blades and a 33 mm rotor. They look very elegant and attractive:

The fans design is extremely interesting: seven crescent-shaped blades of relatively big size come out of the small rotor. The blades edges are pretty sharp, the gap between the tip of the fan blade and the inside frame surface is less than 2 mm and the surface itself is very even and without any edges. Three supporting rods holding the motor and the fan are very thin and have semicircular cross-section. The fourth one serves as cable housing. According to the manufacturer, “iXtrema Pro fans offer ultimate airflow to noise ratio and no other fans on the market come even close to them in performance”. It is a very ambitious statement and we are going to confirm or deny it, since we have plenty of opportunities to do so today.

 

IXP-74-09 and IXP-74-11 models differ from one another by their rotation speed and other characteriatics connected with it. The junior fan model works at 800 RPM (±15%), the mainstream model – at 1100 RPM (±15%). There is also the third top model in this series called IXP-74-14 that features 1400 RPM (±15%) rotation speed, but unfortunately, we couldn’t get a sample for our today’s roundup. IXP-74-09 fan is claimed to create 38 CFM airflow, while IXP-74-11 should blow the air at 46 CFM. The manufacturer doesn’t mention the fans static pressure.

And we will talk about the most interesting feature of SilenX iXtrema Pro IXP-74-09/11 fans: their noise level. According to their technical specifications, these fans’ acoustic readings shouldn’t exceed 9 and 11 dBA respectively. These are truly impressive numbers; however, I would still suggest that we should wait for the actual results before making any conclusions.

The fan spindle is covered with a holographic sticker with the company logo on it. Looks like the manufacturer tried to make sure that no one will attempt to imitate these unique solutions. Besides that, the sticker also bears some information on it, such as fan model and its major specifications:

 

These fans shouldn’t consume more than 1.44 and 1.92 W at 0.12 and 0.16 A current respectively. They are built with some dynamic bearings of unknown MTBF.

Well, let’s see how good these “silent wind blowers” actually are:

I have to say right away that the fan rotation speeds above have been interpolated from their maximum speed as stated in the technical specs depending on the voltage. We had to do it in order to make the graph readable, because the fans tachometers often reported something absolutely weird. In most cases they reported twice as high rotation speed, rarer – twice as low, but even rarer it was close to reality. The same thing happened with the other SilenX fans. As for the actual readings, the only thing iXtrema Pro IXP-74 can actually boast, is low noise of the junior model. The higher-end model starts rumbling at 750 RPM already adding some obvious dissonance to the silence of the night. The created airflow is pretty weak, so there is no unprecedented leadership in offering the best combination of acoustics and airflow. The startup voltages equal 4.5 V for the model 09 and 3.3 V for model 11.

SilenX iXtrema Pro IXP-76-14/18

The next two fan models also come from SilenX and have a lot of similarities with the two fans we have just discussed. However, they are of different size and measure 120x120x38 mm. The design of the fans package and the bundled accessories are exactly the same as those of iXtrema Pro 25 mm fans:

Moreover, the fans themselves look very similar to the other two, except for the depth, if course:

Seven powerful blades are attached to a rotor 41 mm in diameter. Four supporting rods are a little thicker here, because the blades load is higher. Nevertheless, 38 mm iXtrema Pro fans weigh about the same as the 25 mm models: they are unusually light. The inside of the fan frame is very smooth, the gap between the blade tip and the frame is 2 mm, i.e. the same as by the two fans discussed above:

 

These two fans rotate with different speeds. The models with IXP-76-14 index should work at 1200 RPM, while the IXP-76-18 models – at 1400 RPM. They also create different airflow: 72 and 90 CFM respectively. Again, there is no mention of the static pressure readings, but the noise declared readings are right there in the open: SilenX fans are claimed to generate very little noise: 14 and 18 dBA respectively.

I use the word “should” here on purpose, because the real state of things is a little different and we are going to talk more about it in a little bit. 38 mm IXP-76-14 and IXP-76-18 fans use the same bearing type as their 25 mm brothers. We don’t know how long they will last, because the manufacturer doesn’t mention the bearings MTBF.

 

As for the fans power consumption, I have to add that the model with 1200 RPM passport rotation speed should consume no more than 2.4 W at 0.2 A current, while the 1400 RPM model will need 2.88 W at 0.24 A current.

Let’s see how much closer to life and more efficient the actual readings taken off the 38 mm iXtrema Pro fan models will turn out compared with the 25 mm ones:

True, bigger fans create greater airflow than 25 mm fans; however, it is accompanied with higher noise level. So, if we level out the noise level of both models, then the airflow from IXP-76 and IXP-74 will be the same (you will see it at the summary diagrams later on). The mess with rotation speeds is also present here, though not so often anymore and we did manage to get the readings off the device, so there was no need for interpolation. IXP-76 can be considered acoustically comfortable at 810-840 RPM. In terms of power consumption, 38 mm SilenX fans require more, which is actually quite logical. The fans start at 2.7 and 2.4 V respectively.

SilverStone Suscool 121

The next fan we will talk about was provided by the well-known SilverStone Company using a lot of Everflow solutions for their products. The fan is called Suscool 121 and is shipped in clear plastic casing:

 

You can see the entire fan through the front of the package and get detailed information about its features and specifications and even a performance diagram on the back of it. The front side of the box claims that Suscool 121 is the quietest thermal control fan in the world. It could surely be true, because there are very few 120 mm fans out there featuring thermal control diode.

The fan is bundled with a PATA power connector/cable, a cable with a thermal diode, four silicone mounting spindles and four self-tapping screws:

The fan is made in China.

If you remember what we said about other SilverStone fans that we reviewed before, you may get a déjà vu feeling, because Suscool 121 looks very similar to SilverStone FM121 and FM122 fans, only its blades are of light-blue color instead of white:

The fan measures 120x120x25 mm, its weight is not specified. It has 9 blades and a rotor with 43 mm diameter:

 

The fan rotation speed is controlled by a thermal diode in the interval from 400 RPM at 31 °C up to 950 RPM at 60 °C. Aggressively curved crescent-shaped blades create 17.75 CFM and 42.15 CFM airflow at these respective speeds at a very modest static pressure of 0.1 and 0.57 mmH2O. The declared noise level is really low: 15 and 18 DBA respectively. Actually, Suscool 121 is designed solely for quiet operation that is why we shouldn’t expect it to work any performance wonders.

This fan uses what is known as “Enter Bearing” and in reality is a common sleeve bearing with somehow extended lifespan. Although the specifications claim that its MTBF is only 30,000 hours, which is pretty typical of common sleeve bearings (without any optimizations).

The fan should start at 8 V – it is the highest startup voltage of all today’s testing participants. At the minimal rotation speed the fan should do with only 0.08 A and 0.96 W of power, while at maximum speed – with 0.2 A and 2.4 W of power.

As you can see from the marking on the fan spindle, this Everflow model is called F121225SL. Suscool 121 is priced at about $12.

Since we manually adjusted the rotation speeds of our tested fans, SilverStone Suscool 121 thermal diode has been constantly kept at 60 °C. Let’s check out the obtained results:

The results speak for themselves. Suscool 121 is a very quiet fan with low rotation speed and moderate airflow. I could only point out sharp rotation speed increase from 180 to 720 RPM when we raise the voltage by only 0.9 V. the fan consumes very little power, but can start only at 9.6 V.

Thermalright TR-FDB

Users have been complaining quite often that although the coolers for CPUs and graphics accelerators manufactured by Thermalright, Inc. have always been pretty good, they are expensive and do not come bundled with any fans. And what do you think has happened? Thermalright cut the Gordian knot and launched their own 120 mm fan production line (or bought the already existing line, it doesn’t really matter). Now you can buy four Thermalright branded fans of the same model, but with different rotation speeds.

These fans are shipped in cardboard boxes with a small cut-out window in the front of the box that reveals the sticker on the fan spindle. These boxes are so small that it is really tricky to get the fans out, because they sit there very tightly.

However, the boxes are covered with all sorts of info: anything you might want to know about the fans inside (except the benchmark results, of course) is there. The fans do not have any accessories with them – just the fan and the box. Thermalright TR-FDB fans are made in China and come with 3-year warranty.

Thermalright fans look very plain, just like their packaging: black frame and flack fan blades with a little sticker in the middle:

The fan measures 120x120x25 mm and weighs 156 g. Each fan has 7 blades and a rotor with 48 mm diameter. The supporting rods holding the fan and the motor are oval-shaped and pretty thin (minimal diameter is only 4 mm). The only exception is the rod with a hollow groove for the cable inside it:

 

The gap between the tip of the fan blade and the inside of the frame is less than 1.5 mm. The blades are relatively thick and wide and look somewhat like shovel heads:

According to the specifications, Thermalright TR-FDB fans work at 1000, 1300, 1600 and 2000 RPM (±10 %). They should create airflow of 38.9 (just like Nexus Basic), 50, 63.7 and 80.5 CFM. Fans static pressure is not mentioned and the level of noise is declared not to exceed 15.8, 24.1, 28 and 38 dBA.

Thermalright TR-FDB fans are built on the same type of bearings as Scythe S-FLEX. They are called FDB (Fluid Dynamic Bearing). This suggests that they should be pretty quiet. At the same time, the bearings in Thermalright fans should last only 60,000 hours, which is 2.5 times less than the bearings by S-FLEX can do. On the other hand, these bearings will serve you well for over 7 years, which in my opinion is more than enough.

The power consumption of all four fan models is the same – they require 10.8-13.2 V voltage, 0.13 A current and 1.56 W of power. However, according to the information on the rotor stickers, this current value is only valid for two junior models with 1000 and 1300 RPM rotation speeds:

 

 

If we calculate the power consumption for two top fan models, it will make about 2.4 W for 1600 RPM solution and 4.44 W for 2000 RPM solution. The startup voltage for these fans is declared at 7 V. Their recommended retail price is between $11-$15 depending on the model.

Now let’s check out the results:

First of all, I have to say that Thermalright TR-FDB fans turned out different levels of noise and overall unstable solutions. For example, 1300 and 1600 RPM models working at the same speed sounded almost equally (1300 RPM one was a little quieter), while the 1000 RPM model was slightly louder. As for the 2000 RPM model, it was the noisiest of all Thermalright solutions because its bearing produced a distinct crackling sound. Note that none of the other Thermalright fans suffered from crackling of any kind. And when we increased the fan rotation speed, the noise levels changed abruptly, which indicated instability (true for the top models only). Nevertheless, this didn’t prevent Thermalright TR-FDB fans from becoming ones of the best in airflow despite modest exterior and simplistic-looking blades. In addition, these fans consume very little power and start at 3.6 V, except the 1000 RPM model that has a startup voltage of 8.1 V.

Thermaltake ISGC 12 (AF0018)

The new Thermaltake solution – ISGC 12 (AF0018) fan – was developed first of all for the cooler lineup of the same name, which we are going to introduce to you a little later. Nevertheless, you will also be able to purchase this fan separately. It ships in small white boxes made of thick sturdy cardboard. There is a petal-shaped cut out window in the front of the box revealing almost entire fan to the potential buyer’s eyes:

The package also promises “pure silence” – a pretty unique slogan that must have been suggested by the marketing guys, but not the developers of this fan mode. The back of the box shows a photo of the fan, lists its detailed technical specifications and briefly describes its main distinguishing features and peculiarities. The fan is made in China. No warranty is mentioned anywhere on the package.

Thermaltake ISGC 12 comes with four silicone mounting spindles, four self-tapping screws and a booklet with instructions in several languages:

Besides, there is a small rotation speed regulator attached to the fan with a short cable. It is exactly the same as the one in Thermaltake BigTyph14 Pro cooler, for example:

The fan measures 120x120x25 mm and weighs 157 g. It has 9 blades and a rotor with 41 mm diameter:

The fan looks pretty unusual due to glossy white petal-shaped blades. According to the developers, this particular blade shape with a small cutout on the tip of the blade should increase the airflow by 15% and lower the noise by 3%. While the first claim is actually very easy to check out, the second improvement can only be registered by high-precision measuring tools, but by no means by a human ear.

 

The supporting rods of the Thermaltake fan look pretty bulky, especially after the thin rods we have just seen in Noiseblocker products. Wide rods with rectangular cross-section will create an additional obstacle for weak petal-shaped fan blades with small effective surface. The fan frame is quite common with two standing out edges. In other words, we didn’t notice any optimizations here.

Thermaltake ISGC 12 rotation speed can be adjusted manually from 800 to 1300 RPM creating maximum airflow of 58.3 CFM. They claim 1.422 mmH2O static pressure at 16 dBA of noise (it must be the minimal value).

The fans use a new fluid dynamic bearing designed specifically for them. According to Thermaltake, it is not just extremely quiet, but also can work at 50,000 without failing. As you have already understood after checking out a few previous fans in this roundup, Thermaltake’s claimed bearing MTBF is nothing special any more. But maybe it will please us with its extremely low noise?

Speaking of the fan’s power circuit, we should point out that the startup voltage is claimed to be 7 V, current – 0.1-033 A and power consumption – 1.2-3.96 W.

Thermaltake ISGC 12 fans are priced at $15. Judging by the UL number on the rotor sticker, these fans are made by well-known YS Tech Company.

Now let’s continue to the results of our practical experiments:

The new Thermaltake fan turned out quite mediocre and I would even say strange. Up until 660 RPM it worked extremely quietly, practically noiselessly. And then as the rotation speed increases, the noise level jumps up dramatically, which is also accompanied with rumbling of the bearing/motor. However, this fan can still be considered acoustically at 810 RPM, although it creates pretty weak airflow at this speed. Thermaltake ISGC 12 can start at 4.2 V and consumes very little power.

Thermaltake TurboFan (A2491, A2492)

Thermaltake Company will be represented in our today’s roundup with a few other fans – TurboFan model from 2008. This lineup includes two fans: 120x120x25 mm and 140x140x25 mm. They are shipped in plastic boxes with cardboard inserts inside:

The insert contains detailed info about the fans and their technical specifications. Each fan is bundled with only four screws, and the power cable ends not with a traditional three- or four-pin mainboard connector, but with a common PATA-connector, which may be extremely inconvenient for those users who prefer automatic (performed by the mainboard) or manual (performed by special controllers) fan management:

The fans are made in China.

Each 120 mm fan weighs 127 g, and 140 mm fan – 140 g. Both models have 9 blades each. The rotor of the 120 mm model measures 40.5 mm in diameter, and the rotor of the 140 mm model – 45 mm. The fans have eye-catching bright-red blades:

You can easily notice that Thermaltake TurboFan fans boast uniquely shaped blades that originate from the cone-shaped center and curve like the blades of the WWII airplane propellers:

 

According to the manufacturer, the blades of this particular shape attached to a cone-shaped rotor increase the airflow while retain a considerably low level of noise. In fact, all our today’s testing participants try to create an optimal combination of these two fan parameters, and in a few rarer cases also increase the static pressure.

Four rectangular supporting rods holding the fan inside the frame are not that compact. Besides, there is a groove for the cable inside each of them (except the rod that hold the primary power cable):

 

Looks like Thermaltake were going to release a TurboFan fan with LED lighting that’s why they made these grooves. However, so far we know nothing about a fan like that.

120 mm TurboFan model rotates at a constant speed of 1400 RPM (±140 RPM), while the 140 mm model rotates at a constant speed of 1000 RPM (±200 RPM). Both models create the same airflow if 50 CFM and general very similar noise of 16 and 17 dBA respectively. However, their static pressure is more than twice as different: it is claimed to be at 0.7 mmH2O for the 120 mm fan and only 0.3 mmH2O for the 140 mm fan. You already know that even the modest GELID Silent working at 1500 RPM boasts 1.6 mmH2O static pressure. It indicates clearly that Thermaltake TurboFan fans weren’t designed with high static pressure in mind. In fact, you can tell that the blades of t particular shape can’t support high static pressure:

Both models use sleeve bearing with 30,000 hours declared MTBF. There are no innovations of any kind introduced in these bearings, so there is nothing to tell here.

 

Both fan models have the same electrical specifications: 6 V startup voltage, 0.3 A current, 3.6 W power consumption. The recommended price of Thermaltake TurboFan is $9 for a 120 mm fan and $12 for a 140 mm fan.

Now let’s check out the test results:

Strange as it might seem, but old Thermaltake fans performed much better than the new ISGC 12. Although we can’t say that about the 140 mm model, because its creates even lower airflow than a 120 mm model, but consumes more power and generates more noise. Too bad that TurboFan 120 mm turned out an overall noisy fan, because it creates relatively high airflow even at pretty low rotation speed of 700 RPM, which is considered to be its acoustical comfort zone. The 120 mm model starts up at 3.6 V, while 140 mm model powers up at even lower 2.1 V.

XILENCE PWM

The next fan we are going to introduce to you comes from XILENCE Company. The fan comes in a very simple box made of clear plastic with a molded section for the fan and accessories covered with a cardboard insert:

 

The back of the package lists the fan technical specifications in eight different languages. The fan is made in China and there is no mention of its recommended retail price or manufacturer warranty. The fan comes bundled with only four self-tapping screws that will secure it once and for all in the system case panel.

Unlike all other fans we have already talked about today, XILENCE claims no high-tech innovations or know-how of any kind were used for designing their solution. There is no description of the bearing structure, fan blades shape, or any other unique distinguishing features. The fan measures 120x120x25 mm and weighs 175 g. It has 7 blades and a rotor with 48 mm diameter:

The blades are unusually wide, I would even say they are the widest blades of all fans reviewed today. Bright red fan color poses good contrast to the black frame, which makes XILENCE PWM solution look quite attractive:

 

There is no sticker with information on the fan spindle, and the one on the fan itself has minimum info on it. Its rotation speed is PWM controlled with the maximum at 1500 RPM. The manufacturer indicated no minimum speed for their solution. At maximum rotation speed, XILENCE PWM creates 57.9 CFM airflow and generated 21.2 dBA of noise, which must be claimed lower than it actually is. No mention of the static pressure, either.

The fan uses a fluid dynamic bearing with unknown MTBF. At 0.09 A current the declared power consumption should be 1.08 W. The fan has a pretty short 290 mm cable, which is definitely not enough.

Unfortunately, XILENCE PWM turned out to be the worst fan in our today’s test session:

Everything seems to be fine with the XILENCE PWM graph: the airflow is normal, the power consumption is low enough, and the noise curve doesn’t grow too aggressively… However, all this was ruined by unpleasant loud crackling of the bearing that also doesn’t crackle evenly, but changes the tone of it gradually. Overall, I can only recommend this fan for those systems where background noise is already high enough and will drown the crackling of the fan bearing. The fan starts at 6 V. Speaking of acoustical comfort, well, if we can speak about any acoustical comfort at all here, this fan is bearable only at 720 RPM.

Zaward Golf Fan (AFNS-C025L-R420)

Zaward is not very well-known and they appeared to be making cooling fans as well. To be more exact, they sell fans under their brand name, because in reality they are a daughter venture of a much better known manufacturer – Globefan Technology Co., Ltd. The solution we are going to talk about today is called Golf Fan. Yes, it is indeed named after a popular game for the wealthy. There is even a golf club on the fan packaging:

According to the information on the box, this fan uses blades of unique patented shape that ensures 15% lowering of the noise level without losing any of the rotation speed or airflow. The technical specifications of the fan and its special features are listed on the back of the box.

Zaward Golf Fan is bundled with four convenient silicone mounts with rubber rings, PATA power connector and four self-tapping screws:

The fan is made in China, just like most of the other fans participating in our today’s roundup.

The fan measures 120x120x25 mm and weighs 142 g. each fan has 7 blades and a rotor of 45 mm diameter. Zaward Golf Fan looks very unusual due to unique blades surface:

The external and internal blade surfaces are covered in small dimples, no bigger than 3 mm in diameter that is why they look so much like the gold ball. By employing these dimpled blades engineers tried to lower the level of generated noise by 15% without losing any of the airflow intensity. Tests will show how successful this attempt was, and we would like to proceed with checking out the fan:

 

The fan rotates with constant speed of 1200 RPM (±10%). Zaward Golf Fan creates 48.5 CFM airflow at this speed. If you remember reading about Scythe Gentle Typhoon fans, they need 1450 RPM of speed to create the same airflow. The manufacturer didn’t mention the static pressure, but the claimed noise is only 17 dBA, which is fairly low for a 1200 RPM fan.

The fan uses a ceramic bearing. However, its promised MTBF is very low: only 26,000 hours.

Zaward didn’t even try to hide the name of the original fan maker: the fan has Globalfan’s own original sticker with the fan model printed on it (“RL4Z…”). I also have to say that the dimples on the fan blades are also not Zaward’s exclusive invention: we have already come across Globefan solutions with exact same blades, though not in retail but as part of power supply units we tested.

Fan motor shouldn’t consume more than 3 W of power at 0.26 A current. The fan has a 400 mm cable. That’s pretty much all we can tell you about this golfer fan. Let’s check it out in action:

Zaward Golf Fan turned out surprisingly efficient. I said “surprisingly”, because I personally didn’t expect a cartoon fan from a barely known manufacturer to be that good. And I proved to have been wrong. The fan is fairly quiet. No, not so impressively quiet as Noiseblocker, S-FLEX or Gentle Typhoon, but up until 1020 RPM I experienced no acoustic discomfort in the quiet of the night. Moreover, Golf Fan generates excellent airflow and consumes very little power. It is a pity that its’ startup voltage is pretty high: 8.4 V.

Comparative Results Charts

Here we are going to offer you some diagrams with results for all tested fans. This way you will be able to compare their airflow, noise and combination of these two parameters, as well as power consumption and price. Besides the Noctua NF-P12 fan that we have already added to the charts, we would also like to include the results for the following solutions tested before: three Scythe Ultra Kaze models, two top Scythe Slip Stream 120 models, and Scythe Slip Stream 120 PWM taken off Scythe Mugen 2 CPU cooler. Here are their individual results:

The first summary diagram shows the results of the airflow and noise level tests at their maximum rotation speed. The results are sorted in order of airflow reduction:

Click to enlarge

Click to enlarge

We have also built the same diagram with results arranged in a different way – according to the fans noise levels from low to high (click here to view this diagram). The indisputable leader in airflow is Scythe Ultra Kaze 3000 working at 2760 RPM. It is the only fan in our today’s test session, which airflow readings are higher than the level of generated noise (56 dBA). Far behind it we can see high-speed Thermalright solutions, Scythe Slip Stream 120, Nanoxia, Nexus, etc. Then we see the airflow levels gradually decrease and that is why the noise readings come in very handy, because the smaller is the difference between airflow and acoustics of each specific fan, the more efficient it actually is. For example, the following fans offer very good combination of these parameters: Scythe Ultra Kaze 2000, Thermalright TR-FDB-1300, Scythe Gentle Typhoon, Scythe S-FLEX, all Noiseblocker fans and Nexus Basic.

But all these results were obtained at maximum fan rotation speed and consequently at the maximum noise level for each fan. And what if we had all fans working with the same acoustics? What airflow will they be able to produce in this case? To accomplish this goal we took a subjectively comfortable noise level of 33 dBA and made each fan work at this particular noise level using anemometer readings and rotation speed controller. The diagram below shows our results:

 

Note: SilverStone Suscool 121, Noiseblocker NB-Multiframe MF-12-S1, Scythe Gentle Typhoon B2AP-12 and B1AP-11, Noctua NF-S12B ULN and Scythe Kaze Maru 12L worked at their maximum rotation speed and still generated less than 33 dBA of noise. For exact noise readings, please see the corresponding charts.

Very interesting, don’t you think so? The five leaders are exactly the solutions that proved to be the quietest so far, which is quite logical. They are Noiseblocker, Scythe Gentle Typhoon and S-FLEX, Thermalright TR-FDB. Nexus Basic is a little behind them, and then comes a tight pack of fans starting with Enermax Magma and ending with Nanoxia FX12-1250. The last ones on this diagram are rumbling, rustling, or not blowing fans, including the 140 mm models.

Some of you may find a power consumption chart useful. The fans here are sorted according to their power consumption needs from low to high:

Click to enlarge

Click to enlarge

In conclusion to this part of our review we would like to offer you a diagram with recommended retail prices (rounded to $1):

But that’s not all yet. The next part of our article will discuss a small experiment on select 120 mm fan models that you might find interesting.

Select Models Tested on CPU Cooler

Now let’s try using the obtained results for building real systems. Of course, it doesn’t make sense to test all fans on a CPU cooler. Moreover, there is simply not enough time for that. That is why I decided to take one fan from each segment of this summary diagram where all testing participants are running with the same level of noise. I picked Noctua NF-S12B ULN from the group of fans with the lowest airflow: at 33 dBA it worked at 720 RPM and created 15.05 CFM airflow. From the middle group I picked Akasa White LED working at 900 RPM with the same acoustics but creating 20.20 CFM airflow. And from the leading group I picked Noiseblocker NB-Multiframe MF12-S3HS at 1080 RPM creating 24 CFM airflow. I would like to remind you once again that the fans worked with the same level of noise maintained at 33 dBA.

Since we didn’t test the fans static pressure in our today’s test session, it wasn’t a good idea to check them out on a cooler with a densely packed heatsink array. That is why I decided to choose a cooler that would be optimized for the airflow and would depend on it the most. It is going to be ThermoLab BARAM heatsink that we have already tested before. The fans were installed onto the heatsink with the airflow directed towards the top of the system case.

We ran all tests inside a system case with the removed side panel. The configuration of our testbed was the following:

During this short test session we managed to overclock our 45 nm quad-core processor with the multiplier set at 21 and “Load-Line Calibration” enabled to 3.8 GHz (+42.5%) using the weakest cooling system of the testing participants. The nominal processor Vcore was increased to 1.30625 V in the mainboard BIOS.

 The memory voltage was at 1.62 V and its frequency was around 1448 MHz (7-7-7-14_1T timings). All other parameters available in the mainboard BIOS and connected with CPU or memory overclocking remained unchanged (set to Auto).

You can check any of our latest cooling solutions reviews for the detailed description of the testing methodology (the only difference is that we used latest available software versions). The room temperature at the time of tests was around 26.5-27.0 °C.

Let’s see how greatly the fans with dramatically different airflow readings can affect the CPU cooling. First let’s see the results taken with only one fan attached to the heatsink:

Noctua NF-S12B ULN
(720 RPM)

Akasa White LED
(900 RPM)

Noiseblocker
NB-Multiframe
MF12-S3HS
(1080 RPM)

Next we will check out the results obtained with two identical fans:

2 x Noctua NF-S12B ULN
(720 RPM)

2 x Akasa White LED
(900 RPM)

2 x Noiseblocker
NB-Multiframe
MF12-S3HS
(1080 RPM)

And finally the summary chart for more convenient comparison:

Very impressive and illustrative difference, isn’t it? And note that all fans were working at the same level of noise, in other words, by choosing a proper fan you will be able to significantly lower the CPU temperature without losing any of the acoustic comfort. If we go back to the airflow readings, we will recall that Akasa fan generated about 34.2% higher airflow than Noctua at the same level of noise. And as you can see from the obtained results, it greatly affects the CPU temperature. The airflow from Noiseblocker fan, in its turn, is only 18.8% higher than that from Akasa, which translates into smaller CPU temperature difference. In other words, we see evident practical continuity of the results taken with our thermal anemometer. I would like to add that Nexus Basic fan that is about 4% behind Noiseblocker in airflow taken at the same noise level lost only 1 °C (when tested with two fans on a heatsink):

Conclusion

Summing up the results of our today’ test session we should first of all say that choosing the right fan for your particular needs is not an easy task and it is not limited only to the noise and airflow specs. Even though these two parameters are undoubtedly the key ones, you should also consider such important characteristics as static pressure and fan MTBF. And while we are going to come back to studying the fans static pressure in the near future, we won’t be able to verify if the claimed MTBF promises stick for natural reasons. That is why here we will have to trust the experience of users working with specific fans for a considerable period of time (sometimes, very considerable). Theoretically, fans using ball bearings should last longer than the ones using sleeve bearings or their modifications. However, this is not always the case in real life. Besides, the buying decision also includes personal user preferences, including the aesthetics factor, as well as pricing and availability in the market.

Nevertheless, the results of our today’s tests allow us to single out three groups of fans. The first group includes the best models with excellent combination of acoustics and airflow rates. These are first of all Noiseblocker fans from NB-Multiframe and NB-BlackSilent series, Scythe Gentle Typhoon and S-FLEX fans, Thermalright TR-FDB (the specific model we tested) and Nexus Basic. The former fans offer not only the best combination of airflow and noise, but also look stylish and are equipped with unique shock-absorbing triangles. We should absolutely give due credit to Scythe Gentle Typhoon for remarkable bearing design and ridiculously low power consumption. And Thermalright TR-FDB and Nexus Basic together with the junior Noiseblocker NB-BlackSilent fans are the most affordable solutions among the fans in the leading category. If you would like to know my personal preferences in this group of products, they will be with Nexus Basic fan, as it is a practically noiseless model.

The second group of fans is the largest one. Here we have solutions from such makers as Scythe, GlacialTech (EDLA1), Zaward, Enermax, Noctua, Thermaltake, Akasa, Silverstone, SilenX, GELID Solutions and Nanoxia. The fans from the latter two stand out due to their unique looks, fluorescent blades and the richest accessories bundles, although they come at a pretty high price. Among Enermax fans we should point out their top model called Magma for offering the best noise-to-airflow ratio among all Enermax fans. GlacialTech GT 12025 fan proveddthat you don’t need to pay a lot for good fans; the most important thing is to choose the right type of bearing (we are talking about EDLA1). The biggest disappointment in this group was SilenX fans: we expected much more from 38 mm models with small rotor, uniquely-shaped blades and promising slogans on the packaging. Besides, the “cheating” with the iXtrema IXP-74 rotation speed also upset us a little. Unfortunately, Noctua fans didn’t prove up to our expectations because they yielded even to the previous NF-P12 model. However, all these solutions are pretty good anyway, they are simply average overall.

And the third group, as you may have already guessed, is today’s outsiders. First of all here belong crackling Akasa UltraQuiet Amber and Xilence PWM – we can’t recommend any of these models because of very unpleasant crackling sound that they make during work. The two remaining GlacialTech fans also make some additional noises, besides, a lot of users reported that they don’t last too long. The new Thermaltake ISGC 12 fans create very weak airflow, and so do 140 mm fan models, which airflow is also very unfocused. Enermax Apollo produces very high CFM at maximum speed, but also runs very noisy. However, this fan is primarily intended to create lighting of inimitable beauty inside your system case, because no other fan has outstanding lighting like Enermax Apollo.

Well, that’s about all we wanted to tell you about the fans today. Hopefully, this article will give you some food for thought and will help make the right choice or confirm that you have already made the right one. Good luck!