R10- Fan Testing-Bitfenix,Noctua,Silenx,NoiseBlocker,Phobya

[Martinm210]

This test is sponsored by "cpachris" from the OCN forums..

Details & Videos




Scythe (Servo Nidec) Gentle Typhoon AP-15



Bitfenix Spectre Pro All Black 1200RPM



Noctua NF-F12 PWM 1500RPM



Silenx Effizio EFX-12-15 1400RPM



Noiseblocker M12-S2 1250RPM



Phobya Nano-G 12 1500RPM



Results



Since noise is just one of many factors, I will let you decide. I actually like the motor noise better in the noise blocker also the rubber frame. For looks I kind of like the bitfenix.

Cheers!
Martin

[ryan92084]

Great as always martin. AP-15 keeps its crown.

[cmdprompt]

More good stuff, Thanks Martin!

[squick3n]

From Noctua's pricing and marketing materials, you would have thought they had reinvinted the case fan. Pretty unimpressive showing imo

[OldChap]

Is this on the same rig as the round 6 test Martin?

NVM I just read the graph titles

[itznfb]

Those Gentle Typhoons are pretty impressive. When looking at noise vs airflow they just dominate. Is it odd that other companies aren't able to reproduce those results?

Do have links to other tests? I'm sure you've tested Yate Loons?

[Martinm210]

Is this on the same rig as the round 6 test Martin?

NVM I just read the graph titles

Yes, although my anemometer is not to be trusted between rounds. It's been twice now where it just suddenly changes calibration on me, so to counter that I am testing the first fan in each round at the end to make sure there wasn't any change between batches. While the hot wire is superior in sensitivity, warm up, and no noise, it is a bit touchy about dust and temperature. I reset to zero at the beginning of the test, but there is probably bits of dust on the sensor by now causing some inconsistency between rounds. I could probably correct up for it, but I'd rather just do isolated batches.

Those Gentle Typhoons are pretty impressive. When looking at noise vs airflow they just dominate. Is it odd that other companies aren't able to reproduce those results?

Do have links to other tests? I'm sure you've tested Yate Loons?

They are all on my site and also round 6 is in the stickies.

Not sure why others haven't been able to do the same, but I suspect it's the test condition on a radiator that just hasn't been developed. There is something very unique about the GT fan blade geometry that enables it to nearly eliminate all the air noise while mounted to a radiator that others haven't been able to match. Open air or a "case fan" type condition would likely yield completely different results is my guess based on other's detailed testing results.

When mounted to a radiator, the fan is forced to push against quite a pressure load and any flexing of fan blades, etc can result in odd harmonics. In addition any noise created by the blades is probably amplified by the radiator core and shroud area to some extent. I personally don't care much for the GT motor and bearing noise myself, but their blade is superior to anything else I've tested on a radiator. The blade material is very rigid and doesn't flex much and the high number and dense blades probably block much of the noise in the rad area.

With that said, I'm still running yate loons myself. I don't see the difference in sub 1000RPM speeds myself and too cheap to pay for more expensive fans.

[Utnorris]

Wonder how the Helix will stand up to the GT's.

[grasskicker509]

Thanks for the great work...is there a performance difference between push and pull when using gts on a rad?

[Martinm210]

Thanks for the great work...is there a performance difference between push and pull when using gts on a rad?

I haven't figured out a way to measure push vs. pull in a direct apples to apples way so I don't know. If I mount a fan in pull on my test rig the noise meter and the H1 audio recorder are bombarded with wind buffeting error. If I mount the fan behind the radiator to pull through the rad and into the exhaust chamber, then it is obscured and muffled by the radiator not to mention a PITA to do for every fan.

I would say bury the fan as deep as possible, the more radiator or stuff you have between the fan and you the more muffled any noise is. I have yates on my sons system pushing up through a radiator mounted in the ceiling. It is much quieter that way than having the fans on the exterior pulling because the radiator itself acts as a muffler of sorts.

[chaotic]

Good job as always Martin Thanks a lot!

[litoralis]

I just finished reading your review and figured I'd find you lurking about here.

I have what is an annoying question since you just answered it partially in the replies above...

I am working on a new 240mm radiator plus 120mm radiator liquid cooled rig. The whole point of this was to lower the ambient noise of the new PC. Going so far as to install a separate liquid cooling loop for the GPU.

Thinking about doing AP-14->Shroud->Rad->AP-14 , the arrows point the direction of airflow.

So, without having to read your entire website today, can I ask if you've tried these same tests using shrouds at any point in the past?

I'm looking at a purchase of eight(8) GT AP14's or AP 15's, it's a bit of an expense for just the fans, so I'm trying to wrap my head around ways to further silence the darn things before I pull the trigger and buy them today/tomorrow.

[squick3n]

I can at least help you a little as far as noise. What he's saying about obstructing the fan is definitely true. I previously had AP14s in a pull at the top of my case, and they were louder running at full speed than AP15s running at full speed with the fans in a push config. There's just no way you aren't going to hear those fans sitting on top of your case with nothing to suppress the sound. Unless you go with a very low speed fan, at which point you won't be disipating much heat

If you are going for silence, keep the fans inside the case and pushing air out against the rad

[Lanek]

I am maybe wrong... but why the GT AP14-15 are so good performance wise is due to the blade configuration, who create an extremely high pressure... and this is perfect in push for a rad.. But in pull.. it dont work the same at all. Basically your fan will create an aspiration... the pressure dont matter any more.


If you want to use pull only .. , i can believe the surface of aspiration is the more important, go for a rads who can own 140mm fans.

[Endgame]

Is a push / pull setup, would there be something to be gained from using really low rpm fans pull side and push fans that don't kick in until a certain temperature is reached? Maybe something really slow pulling like undervolted YL lows, and then when temps kick up (say like when playing BF3) having some push ap15s kick in. Part of me thinks that would work great, the other part thinks the slower fan would really get in the way when the push fan really kicks into gear.

[Martinm210]

I'm not sure on the shroud questions. My RS120 based testing had me thinking there were good gains to be had on all radiators, then my more recent MCR120QP based retesting gave little to no gains with shrouds leaving me to conclude that it's very radiator specific. It's just to laboring to try and sort out, so your guess is as good as mine.

Push plus pull is always a good gain and I think better for noise too, but more costly and more space required. As far as mixing fans, I personally don't think it really matters. Like pumps the fan produces differential pressure to move air through the radiator, it doesn't really care too much as long as there is a flow spreader like a radiator to help straighten the air back out. I do know that stacking fans does NOT work, the air is too disturbed.

[Church]

Martin: how about these?

[Martinm210]

I'm looking at maybe doing a filter/grill type round if anything in the future. I've probably tested 100 fans now, so something a little different would be good.

[OldChap]

If you do filters Martin please try to include some of the more industrial types of material than we see clipped on to cases just now. I would be really interested to see how pleated and hepa materials fare

[Church]

And another good candidate might be Demciflex's as imho rather popular ones because of their magnetic mount sys & having sets for many popular case models and also for almost each and every rad size except maybe large squared ones like MoRa-s.