Archive for the ‘Fans’ Category

As part of my radiator testing, I did purchase several sets of fans to experiment with.  Since the Gentle Typhoon AP-15 did so well in a noise/cfm ratio in previous testing I had been curious how the higher speed GTs work in comparison.  The AP-29 differs from the AP-15 in that it only has 7 blades as opposed to 9 and it also has a special and unique stiffening ring.  I tried to make contact with Scythe about possibly sponsoring fans but my request was ignored and I could never make any sort of actual contact.  The Titans are just something I found perusing SVC one day and had to have a set and also picked them for my radiator testing since they were cheap, in stock, and had a good RPM range.

Being without any sponsor, I’ll give myself a special thanks to for sponsoring this test…:)  I ordered the GT AP-29s from an ebay seller as it seemed to be the cheapest price shipped I could find.  I also ordered the Titan’s from SVC myself since they were on sale for $10.99 and I was anxiously crossing my fingers I would find a good alternative to the GTs.

Gentle Typhoon AP-29

First thing to catch my attention is this fan comes only with 4 pin molex connectors, so you will need an adapter or to rewire it on your own if you need 3 pin.  The fan doesn’t come with much more than standard self tapping case screws.  It is built well though and heavy and more industrial in quality than your average 25mm fan.

Gentle Typhoon AP-15

For comparison purposes with the previous 1000-1800RPM winner, here is the AP15.  The king of CFM/RPM from my previous rounds.

Zalman ZM-F3

For comparison with a good but average noise level and more conventional 7 bladed fan, here is the Zalman ZM-F3.  It what I consider a good average noise level fan at a great price.  It also has a pretty good speed range so it’s a good one to compare with as well.

Titan Kukri

These are the fans I’m using for the radiator testing that were attempting to be low cost GTs.  They look nice and actually do better than the Zalman at slower speeds, but things fall apart quickly at the higher speed range.  The sample tested seemed to have some debris in the motor causing a bit of rubbing type noises from 1400-1800 and then a harmonic developed from 1800 up to 2300RPM.  The debris would probably work itself over time and the harmonic may or may not be specific to the radiator test rig, but that’s what I got in this run.

Charted results

The charts above include looking at a few variables independently.

Of primary value is the air flow vs noise level chart which is the most representative of radiator performance.  It’s not perfect nor as good as actual thermal based testing, but I find it good “quick” simulation that allows looking at the full spectrum of possible undervolting conditions very quickly.  It does however only represent noise “Level” and not “Quality”.

Chart 2 is just comparing RPM to noise and plots a similar response.  The GTs are very good about keeping noise “Level” down to a minimum when mounted to a radiator and much the same using RPM between the 15 and 29.

Chart 3 is taking a closer look at how much air they push per RPM.  This is where the GT 15s doing very well having a higher blade count without stiffening ring allows them to produce more air per RPM.  General most fans produce the same amount withing about a 200-400RPM window though.  Somewhat surprising though, the Titan despite it’s similar appearance doesn’t have the same flow/rpm advantage as the GT 15s.  There is quite a bit more gap between fan blade and frame and slightly different blade pitch as well as how the back of the blade curves more than the GTs.  The AP-29s do a bit better than average at slower speeds, but seem to loose all of that advantage from about 2000RPM on upwards which I suspect is mostly due to the stiffening ring causing turbulence.


The higher speed AP-29 does seem to follow in a similar trend (CFM per RPM) as the exceptionally low AP-15 did although I would say subjectively when listening to the videos side by side, the 15 is producing less motor noise and smoother in sound.  The stiffening ring does seem to eliminate the resonance spikes that are present as specific RPM levels of the AP-15, so that’s a good thing although the CFM per RPM level is not quite as good as the 15 which is also probably due to this same ring.  The range is a bit more limited on the lower end were 5V is basically starting at 1200RPM, but it makes up for that in it’s near 3000RPM maximum at 12V.

While it’s hard to find fault in the AP-29, I personally find noise levels beyond 2000RPM still to high for my taste despite the noise levels being significantly lower than your average fan and also find the noise quality of the GT15 better than the 29 at slower speeds so it is a bit of a compromise.  I think more people would be better off with the slower speed 15, but the AP-29 does extend the fan power range for those needing similar GT quality in a higher speed flavor.  As with pretty much all fans, you can’t have it all and extremely high speed fans generally struggle at slower speeds and the AP family is much the same there.  The AP-29s are exceptionally good at 1800-3000RPM, but  the AP-15s are still king of 1000-1800RPM levels.

The Titan Kukri’s do pretty well at slower speeds and they do have a more desirable white blade color, but they are not even in the same league regarding build quality as the gentle typhoons.   On my particular test rig with radiator, I also had high speed harmonics and other not so smooth motor type noises as RPMs increase beyond about 1400RPM and particularly beyond 1800RPM.  Not quite sure why but I suspect it’s the blades flexing since the plastic used is softer than that of the GTs and not as reinforced at the hub either.  Overall a great fan for the price when operated at slower speeds, but the Titans are not at all outperforming or coming close in performance to the GT-15 in terms of noise level or noise quality in the 1400-1800 range.  I would say they are better than average in the <1300-1400RPM level, but average to poor at higher speeds.


Welcome to another pump speed controller (CTR-SPD10), and before you pass by thinking you can get the same thing from any fan controller, take a closer look.  This little analog controller is also built with a transformer so you get slight “Overvolt” capabilities as well as undervolting.  I used these controllers in my pump noise testing because of this ability and they are really well suited to driving the PMP-400  or other DDC pumps that otherwise don’t have speed control.

I would also like to thank Tim from for providing this sample.  These controllers gave me the ability to complete my noise testing in complete silence (my test PSU clicks very loud when switching) Thanks!!


The controller housing notes the range of use is 7.5V to 12.7V out at 25W max.  This is really ideal for the PMP-400 or similar fixed pump.


The unit is compact in size and could be installed in a variety of locations.  It comes with a piece of velcro as one option and also a pair of screws where you could solidly mount the face by drilling three holes (one for the control knob and two for the mounting screws).

It is just over 3″ long and about 1-3/4″ wide and just under 1″ tall.


Input is a 4 pin molex, output is a 3 pin header


The control knob is located up front and has a small phillips slot


I figured the most common use of the controller would be controlling a DDC pump such as the Koolance PMP-400 pump, so that’s what I used to test the controller.  I wanted to see what the actual voltage range was when feeding it exactly 12.0V while under load of the pump.  I also wanted to take a look at the pump performance and rpm to see what changed under this control.

Minimum Voltage As Measured to pump = 7.90V

Maximum voltage as measured to pump = 13.28V

Overall, this is a good range of voltage control for this pump with the exception of the maximum being a touch high.  I have found that the PMP-400 pump doesn’t like to start beyond 13.0V, so you will need to be careful to dial down the controller just a bit to ensure starting.  12.8V is about the maximum overvolting that I would consider, so that’s what I used in testing the pump.

I did my normal pump testing to measure dynamic head pressure (outlet – inlet) while increasing restriction and measuring flow rate as shown above.  I also adjusted the pump voltage via the Koolance CTR-SPD10 to several fixed voltage ranges to see how the performance is impacted.

Performance Pressure vs. Flow Rate Summary, up to a 15% pressure head increase

Here is a look at the RPM or speed of the pump for noise reduction.

Speed (RPM) comparison, up to a 35% RPM reduction


This is a great little pump controller particularly for the many pumps which are fixed in speed such as the PMP-400 or other DDC pumps.  This controller essentially gives you variable power and noise control over a pump that is otherwise fixed.  In addition it gives you the option of over-volting the pump due to it’s unique transformer.  While there are many fan controllers that could also reduce speed, this is the only option I know of that can go down to 8V and UP to 13V.  Most fan controller will have a maximum below 12.0V because of losses within the fan controller.


  • 25 watts capacity is ideal for the PMP-400
  • Over-volt (13Volts) capability can increase pump pressure head by up to 15%
  • Under-volt (8 Volts) capability can reduce pump noise for ultra silence
  • Compact size make is fairly easy to install in a variety of locations
  • Simple operation up and down


  • Maximum volts can exceed pump startup limit (user needs to reduce slightly from maximum)
  • All manual control only

Overall I am very happy with this controller, particularly for it’s unique ability to not only provide a solid 12V which most fan controllers can’t do, but also to increase it for an extra bump in performance.  I would suggest installing extra cooling on pumps that you are considering overvolting, as any sort of voltage increase will also increase heat output of the pump.

Where to Buy



The Koolance CTR-SPD24 is a new 24 Volt speed controller which will take your ordinary 12V power supply and allow you feed and control a pump or fan from 10 volts to 24 Volts.  This is really exciting because previous to this, 24V or higher voltage was typically only possible with a separate dedicated power supply.  In addition, the controller is ideal in controller the new stronger PMP-450S pumps.  So unlike your typical fan controller which is either PWM or resistance based (Can only reduce voltage), this one can got both directions.

Before going too far, I would like to thank Tim from for supplying two of these units that I have been using for my PMP-450S performance and noise testing.  Thanks!


Here are some pictures showing the box, packaging, and size of the controller:


Minimum and maximum voltage settings

For testing I am including both the minimum and maximum photos of the PMP-450S pump under load using the controller.  I fed the controller precisely 12.0V using my test power supply, and then measure the pump voltage at the pump plug for the actual voltage provided.

Minimum Voltage = 10.14V (Note: I would suggest 11V as the minimum for the PMP-450S)

Maximum Voltage (Note: I would suggest 20V as the PMP-450S maximum)


To determine the efficiency of the controller I measure the amperage and voltage fed to the combined pump plus controller, then did the same for just the pump.  Using these two sets of data I was able to determine the energy lost through the controller.  Here are those results:

Efficiency Calculation

It turns out the controller is very efficient at over 89%, so the heat-sink and controller is wasting up to about 5 watts at a 48 watt consumption level.  I did also measure the heat-sink with a laser thermometer and it does get warm/hot to the touch.  I measured roughly 45C in a 25C ambient which is warm, but not that hot.  I have actually measured some pumps measuring upwards of 60C, so 45C seems to be relatively good.

What good is 24V you might ask?

It’s what the “STRONG” or PMP-450S NEEDS for full out pumping performance.  The PMP-450S with this controller will easily outperform the most powerfull PMP-450/DDC pump.  It also serves as a “Vario” feature for the PMP-450S because that pump does not come with a variable speed knob.  Basically, the PMP450S and this speed controller should go as a pair for full pump performance.  Here is how the “STRONG” compares to the PMP-400 with top, and I haven’t even completed testing with a top on it.  It is simply one of the most powerful pumps you can buy at that price point.

24Volts gives the PMP-450S it's "STRONG" performance

I also know of a few people that have 24V fans and there are other needs I’m sure.  I would not suggest the controller for the regular PMP-450 pump though as the performance gain on that pump model is fairly minimal.  The pump that needs this is the PMP-450S, she likes the extra volts!!..

I really like this controller, particularly for it’s compact size, simple analog control, and extra legs it gives to the PMP-450S model pumps.  I could see this controller also coming in handy for other 24V needs, no more need to have a  separate dedicated power supply for high voltage, this fits that need nicely at a good price.  I really could not find any faults in this controller, it’s a nice compact unit and has plenty of power for one PMP-450S.  It would even have enough power to feed two PMP-450S pumps at lower voltages/lower flow rates.  I will be using these for a variety of 24V testing needs in the future, extremely nice!!

 Update 7/27/11 Connector Photo

This is how you connect up a molex connector type pump like the PMP-450S:


R8 120mm Fan Testing on an MCR120 continued

Posted: December 16, 2010 in Fans

I’ll update this post as I get all the individual charts cut, but here are the summary results and videos. This batch is released separately from my R6 results because for some reason my anemometer is reading differently and the batch failed my check in/out process to ensure consistency.  As with any large group of tests, it can be a challenge trying to maintain the same exact conditions.  In this example, I was not satisfied, so I broke the results out separately.  I do however have one common fan, the yate loon, so you can do some cross references by comparing to that fan, then comparing to the other round 6.

A special thanks to Old Chap, RatDog (Cisco Systems), Utnorris, & IKIKUINTHENUTZ for sponsoring the fans


In general these were all fairly impressive fans considering most of this batch edged out the very good yate on noise level…at least at higher volts. I also had my first 10 for noise quality (At 500RPM the SlipstreamSL is hard to find fault, it’s practically inaudible)..not moving much air at that speed, but darn quiet too..

I also generally found more of the same, fans designed for high speed have noise quality issues at very low volts. The slipstream PWM vs SL for example, same goes for all the really high speed fans.

Utnorris’s San Ace also did really well at 12V slightly edging out the HE.

Regarding the Deltas….there seems to be a very clear difference between the newer REV 3 with notches in the fan hub vs. the older style. The SHE tested was of an older revision and did not test nearly as well as the HE tested here or the VHE tested in round 6. The older 25mm delta didn’t seem anything particularly above and beyond other than it has a fairly impressive RPM range. Unfortunately the slower speed Delta HE had trouble starting at lower volts. It really wouldn’t kick on until just under 9V, but since there are some controllers that can jump start at higher volts, I ran it back down to a lower than startup voltage just so you could see that possibility.

Slipstream PWM also seems to surprise a little on noise level, it performed really well there although I think it suffers from some motor tick.

I also had fun with my first “Made in the USA” DC-Muffin fan. Like the Papst, it too runs backwards. It was pretty tinny at low volts but had a respectable noise level at higher volts. I do think the 3 bladed design has more blade chop type noise though.

I also gave up on PWM vs Voltage control testing. So far, voltage control has been superior in noise quality, so I’m sticking to that for the videos.

Anyhow….those are my random thoughts…have a look at the videos and see if anything catches your eye.

In the end for me, I’m still eyeballing the sub 1000RPM range, and the yate and slipstream SL were smoothest down there. I think the slipstream SL vs GTAP13 or 12 would be a close match..




Akasa Viper PWM (Voltage Control)

Delta AFB1212HE -S45T 25mm

Delta AFB1212SHE -46T 38mm Old Version

Delta AFB1212HE Rev3 (From SidewinderComputers) 38mm

Evercool Aluminum Frame EC12025M12CA

Muffin-DC Model MD12B2

San Ace 109R1212H1011 38mm Sample Utnorris

Scythe Slipstream SY1225SL12SL 500RPM

Scythe Slip Stream 110CFM/1900RPM PWM (Voltage Control)

There you have it. I would highly recommend the side by side video method for comparing fans, just match up similar FPM levels via the blue anemometer and listen with your headphones at a reasonable level. Personally, I think noise quality is more important than noise level, but everyone has a slightly different take on it and their own preference.


Same as my round 6 testing and methods with the exception of using a Hardware Labs 140mm SR1 radiator. Seems as though we’re getting more and more 140mm radiator options, so it would be nice to get a feel on how a fan performs on one from a noise perspective. This thread is devoted toward more real world like testing of fans physically mounted to the HWlabs SR1 radiator.

Special thanks to the many generous sponsors:

Here is the list of fans, sponsors and results currently complete. I plan to do these 140mm fans as well as a few of the top performing 120mm fans on an adapter for 120mm options.


The 120mm fans are tested on a BGears slim profile 120>140 adapter. Soo…the 120’s will be getting a small shroud benefit from being further from the radiator fins. In order to see how much if any this diffence means, I’m going to retest my top 140 fans with a shroud and include them in the below chart.
The below chart is sort of a mix, not all tested equally, so take that into account when reviewing:

140mm fans only: It’s pretty tight with most of the results within the 3dbA or lower “Barely Perceptible” limit. There are however some fairly notable sound quality issues, particularly with the 7 blade fans at higher speeds including the yates. For low speed, I would tend to favor the Thermalright X-Silent, for high speeds, the Aerocool Shark.

120mm fans on an adapter: This is a bit apples to oranges with the adapter, but it’s fairly surprising just how close 120mm fans perform to 140mm fans on a CFM per dbA ratio perspective. The 140mm fans do have some CFM per RPM advantage and seem to produce a slightly lower frequency tone, but their noise levels were pretty much the same. This makes some of the stronger performing 120mm fans a very viable option on 140mm radiators. I had really hoped the 140mm fans would be a huge benefit over 120s, but I’m just not finding that. There is a good sized advantage to the larger 140mm radiator and reduced restriction, but the gain is in the radiator frontal area, not the 140mm fan itself.

I have a few more 140’s coming but that’s what I think so far…