120mm Fan Testing on an MCR120 Radiator Round 6

[Martinm210]

I'll see if I can dig up those videos and at least clip a screenshot of the Kamaflow 2 vs GT.

[tistou77]

In the video, the 2nd unit from the right (right multimeter for voltage), what is it?

Thanks

Someone to tell me

Not really a good idea to put data from 2 different test beds head to head like that as there can be some subtle (or major) differences in how the test bed is made which might produce different results. Having said that, note that the GT produces those numbers at a slower RPM which usually indicates better static pressure which is needed to push air through the fins of a rad.

Yes I know, I say that it was not good to compare, it was an example

I'll see if I can dig up those videos and at least clip a screenshot of the Kamaflow 2 vs GT.

Thank you very much

[Martinm210]

Here is a quick look at a few of the fans I tested in Round 11 at similar (550-590 air flow speeds), but haven't published yet. I liked the noise quality coming from the Kama Flow 2, Be Quiet, & the Phobya fans, but the GT was still prevailing on noise levels when radiator mounted. They all had good smooth air like smooth sound, but the GT is practically absent of air noise from it's blade design giving it the lowest noise level. This does however leave the bearing noises and other not so pleasant noises to be more apparent so some users don't like the GT sound quality and I can understand that.

The meters are left to right (Exhaust Chamber Air Flow in Feet Per Minute | Noise Level dbA | Voltage at Fan Plug | Speed in Hz (typically Hzx30=RPM))


The Be Quiet does require a bit more RPM to reach the same air flow levels. The Kama Flow 2 was pretty good with that, not quite to GT CFM/RPM levels but decent.

The Phobya fans have really good noise quality and very similar to that of Noise Blocker fans, but just not keeping up with GTs on noise level either.

[tistou77]

Here is a quick look at a few of the fans I tested in Round 11 at similar (550-590 air flow speeds), but haven't published yet. I liked the noise quality coming from the Kama Flow 2, Be Quiet, & the Phobya fans, but the GT was still prevailing on noise levels when radiator mounted. They all had good smooth air like smooth sound, but the GT is practically absent of air noise from it's blade design giving it the lowest noise level. This does however leave the bearing noises and other not so pleasant noises to be more apparent so some users don't like the GT sound quality and I can understand that.

The meters are left to right (Exhaust Chamber Air Flow in Feet Per Minute | Noise Level dbA | Voltage at Fan Plug | Speed in Hz (typically Hzx30=RPM))


The Be Quiet does require a bit more RPM to reach the same air flow levels. The Kama Flow 2 was pretty good with that, not quite to GT CFM/RPM levels but decent.

The Phobya fans have really good noise quality and very similar to that of Noise Blocker fans, but just not keeping up with GTs on noise level either.

thank you very much for your excellent work
the GT (AP-13?) is "better" than the Be Quiet and Kama Flow 2

below 1000 rpm, the same conclusion I presume

[CryptiK]

Brilliant review, pity the videos dont exist anymore. I picked up some Gentle Typhoon AP-15's (new revision) and when tested individually in free air they sound almost silent. Great I thought, loomed them up and bolted all 4 to my rad and oh god, the resonance is terrible. All 4 are running off of one channel of my Lamptron FC5, and all seeing the same voltage. They resonate and whine at all rpm's over ~800. It's intolerable. I decoupled them off the rad using rubber gaskets and they are about the same, I can't explain it. Sounds like I'm sitting in the next room from a fan doing about 6K rpm.

Since these cannot be used, what's the next best CFM per noise? The noiseblockers? I didn't think the static pressure was much good on these and I can't buy them locally to test one (no one I can find stocks these in Australia). If I can't get them what's next? I have yet to try the Corsair SP Quiet editions, has anyone used these on a rad and can comment?

I am coming from 4 x Scythe S-Flex F's at ~1150rpm which produced a nice smooth sound but didn't cool quite as well as I needed.

[Martinm210]

Sorry a out the videos, I dumped my whole youtube collection a while back.

Not sure on your resonance issue, but perhaps there hope in purposely altering the voltage in each fan just enough to avoid that Multiple like frequency harmonic. While I have not tested that much in fans, I have noticed it with multiple pumps. In those examples a minor adjustment to very slightly make the RPM different was enough to fix the problem.

I would suggest giving that a try. Put each fan on a separate channel and adjust them slightly differently.

[CryptiK]

There was definitley a multiple harmonic effect going on, but I think there will still be a whine present at any rpm as when the fan speed is turned up or down, a new whine starts as one stops, there's like multilpe overlayed whines at all speeds until I hit ~800rpm. They move a lot of air but the noise is awful. I will try them on separate channels though thanks for the suggestion.

How do the Noctua NF-P12 go on rads? Their noise and static pressure ratings are appealing.

edit - strike that I see it went horribly (didnt think you tested that model!)

[CryptiK]

If runnign the GT's individually doesnt correct the issue I think I'm going to go with the Noiseblocker M12-S3HS's they look to be the next best thing and don't suffer from the harmonics and odd sound signature that the GT's do.

[Martinm210]

GTs do seem to have some harmonics at specific RPMs as well where 900rpm could very well be better than 800..etc I think I counted upwards of 5 or more bumps on various samples so it is worth experimenting with RPM levels as well. Interesting report though, I never have done much in the ways of testing multiple fans, always just did the single. It would be interesting to see if things changed on a multi fan rad bench.

As far as other fans, under about 1000RPM or so, many fans are excellent performers and GTs are really nothing special down that slow either. I personally still use yate loons at 800 rpm as the difference is too small at those very slow fan speed to warrant buying more expensive fans.

That and I am just cheap..

[CryptiK]

Ok did some more testing running the fans on individual channels. You can spin one fan up to target, say 1350 rpm and listen and it's basically silent just some slight motor and air noise. However spin a second fan up and when it gets within ~120 rpm of the first fan (when it reached about 1230 rpm) that heterodyne effect starts and you can hear the harmonic oscillation. With a 3rd and 4th fan also brought up to the same speed the effect was additive but not as noticeable (the most noticeable was going from normal fan sound to the harmonic appearing). I have read in a few places now people claim that the new versions or at some point in recent history these fans started making this sound, and early production samples were quiet. Someone even emailed asking if they'd redesigned the motor but received no response.

I then did some testing bringing each fan up and down with the fan controller or stopping or slowing individual fans by hand and I identified 2 fans that would make the tone that caused the harmonic effect more strongly than the others. I have a quad rad but bought 7 GT's thinking I might add another single rad or even throw a couple on the other side for push pull to help get the air out of the care (this is running in the base of a TJ07). So I swapped out the noisier 2 fans and installed the other 2 and the effect was immediate. I could still here a little harmonic effect but at perhaps a third or less of the volume of the initial incidence.

I wasn't sure what was going on, but since the effect is not audible when the fans are tested sitting on carpet but becomes audible when mounted to a rad I thought perhaps the fact the fans are bolted hard against the rad which is on metal L brackets riveted to a solid aluminium case there was some transmission of the resonance into the case. I grabbed some TFC decouplersa that came with my rad (bad design so I just cut the corners off) and used them to decouple the fans and took care to ensure the fan frames weren't touching one another. The result was instant almost complete mitigation of the harmonics. There is still a mid-bass motor noise hum which is made up of 2 notes an octave apart but it's very quiet and not unpleasant, with hardly any moving air noise even at 1150-1200 rpm. Similar to the motor noise of san ace's, they sound like a 'real mans fan'. I would now rate the sound quality of these fans as very good.

All up, if you are having issues with resonance with GT's, try and identify which fans are doing it (it's only subtle but you can hear some are louder than others and that's all it takes) and swap them out for the quieter ones. Out of 7 purchased I have 5 that are ok for use. Then decouple them using rubber mounts and ensure the fan frames aren't touching when mounted and you should enjoy the GT's.

Thanks Martin, without your suggestion I would never have done that testing. I owe you a beer I am still considering whether or not to try the noiseblockers for a more silent experience, but I think if I really want quiet I should switch to a larger external rad like the MO-RA3 with some 180mm phobyas at 700 rpm or similar.

[Martinm210]

Interesting, thanks for reporting back on the results of the several trials. I guess as with most things, a little tinkering can go a long ways in solving big problems or at least making big problems small.
Cheers!
Martin

[Church]

CryptiK: My guess is that for larger fans you go, worse results you'll get, with reasoning behind: 1) the larger the fans, worse/smaller fan selection is (mirriads of 120, acceptable count of different 140, but anything above - 2-10 at most), thus less chance for some model to be exceptionally good (and also less chance for something to be independently be tested); 2) larger the fans, lower the rpms at which noise is within acceptable level (my guess - because with fan blades of big size you get sooner less rigid at that size blade vibration issues), thus very probably hurting a lot fan's produced pressure (=acceptable for case fans, not so for rads). I doubt that large open areas between large slow rotating blades can bring out reasonable pressure. So imho is best noise/performance wise to limit oneself to fan choice upscaling just to 140mm fan size (eg. 9x140 mora3). About the only exception in my book would be some specific rads well fit for specific cases/fans - eg. 3x180 rad for FT02/RV02 cases, where that allows for stock fans be used and also rad overall wider frontal area might somewhat compensate worse fan performance. With external square rads with same frontal area just different count of different sized fans imho you'd be better of sticking to 120/140.

[CryptiK]

This is a video of the harmonics with the GT 1850's. Crank the sound up, don't worry there's no loud noises on the video:




Church - yeah I know what you mean, I've heard a few reports though of basically silence when using a large rad like the MO-RA3 and ~700 rpm fans either the silverstones or Phobyas. A local guy ran that setup and said with the rad 3 feet from him he could hear nothing from his system other than hard drive searching. Rads that size I think just have so much surface area vs little rads we fit in cases and with a low fin density and slow fans work exceptionally well. Maybe not as good in terms of power dissipated per square cm as say a good 480 rad with GT's on it at reasonable RPM, but enough to easily handle an overclocked CPU and SLI setup and keep temps low. I'm hopeful anyway! Not sure which way I'll go yet, I have a set of 4 noisblockers in my shopping cart, think I might try that path first. I can only get the PL-2's in Australia though (we have a small selection of stuff here but I can't bring myself to pay $90-100 shipping from frozenpc etc for fans.

[MengNa]

Hey Martin, I was wondering if you're planning on testing the Swiftech Helix fans (PWM or regular) and the Corsair high static pressure fans in the future?
Thanks for all the testing you've done so far, it's all been a great help!

[Martinm210]

I am going to be doing kit reviews in the near term one of which is the H220. Part of that will include some modified type tests to evaluate the stock fans vs other fans and include noise efforts.

It will be different than this test bench but should still provide some helix based information. I may follow up with a fan only comparision later, but my focus and intent is to test in a case with multiple fans. I have had a few people report some fans don't work well in multiples and that is something missing from this single fan test method I used here. I also want to include the grill and case condition as part of the test. Will be tricky to isolate a bit, but should be fun and something different and I hope more real world use.

[MengNa]

That's great! I like that you test on Swiftech rads as I use them myself so it's very relevant for me.
I do hope that some of the other fans you compare the Helixes to are GTs and Corsair SP120s as I'm trying to decide between them for my next upgrade. I really want to move to PWM so hoping that the Helix fans turn out comparable to GTs.
Edit: I also wanted to say that I'm happy that your site is up and running again. It was saddening to see such a great source of information being shut down. I haven't been following the forums very closely lately, so I hope you don't mind me asking, but what happened before and what changed now?

[Martinm210]

My website hosting was expiring, and figured on hanging up the hat. Had several people ask that I keep it going and contributed, then winter set in and it was too cold to work in the garage. I am also moving so I may be without the garage and atv hobby for a few months here so I figured why not, something to do in my new apartment until I buy a new house at the end of summer.

[OldChap]

Church: I suspect that you are right about large slow fans being poor on a rad.

There must be a reason why some air-con makers are using a more extreme blade geometry

[Church]

Martinm210: if you ever touch again fan testbed, apart from Swiftech's Helixes with GT alike blades i'm a bit interested in NB's eLoop fans noise/performance, by vendor specs supposedly ~ 3db quieter then multiframe series.

OldChap: Take a note, that for most aircons having heatspreaders with fans mostly mounted outside building, noise matters way less, thus they usually go simplest way - ramp fan power/speed up. If you can ignore noise, rpms are by far cheapest and simplest way to get airflow & pressure and thus performance up, no matter what blades fan has. I've seen many with simple metalic small count blades .. but oh noes, that god awful rattle

[CryptiK]

Martin, on the subject of the qualitative assessment of fan noise, have you seen this: http://ixbtlabs.com/articles2/cpu/fa...k7-noise2.html

Quite interesting, I think it's much more telling of how a fan 'sounds' rather than verbal description alone. Any chance something like that could be done?


edit - ive tried many voltages, tried decoupling, tried just getting used to it, but I can't handle the GT's. That soft warbling harmonic just gets on my nerves when I'm browsing or working. When the fans are at ~1000 rpm the harmonic tone is ~577 Hz, and by 1300 rpm this falls to ~540 Hz. I find anything in this tonal region very irritating especially with any oscillation. One of the reasons I'd like to see spectral analysis of fan noise - might not be loud but damn can it drive you nuts. I've gone ahead and ordered the noiseblocker M12-S3HS, hopefully they sound real nice by comparison!

I was using this to match the tone: http://onlinetonegenerator.com/

Just move it around until you hear it match the tone of the fans and begin to oscillate with them.

[CryptiK]

Just an update on my situation. I installed 4 x noiseblocker M12-S3HS fans on the rad. They do make more of a typical 'air movement' sound than the GT's but they have no chop or buffeting that the Scythe S-Flex's used to have. At 1000rpm they are basically inaudible, I have to unplug my ram fan to really hear them. At 1200 rpm they are audible, but produce that low air noise that's quite pleasant, definitely a high quality exceptional sounding fan. And the best part, no harmonic oscillations! They move a lot of air, I think more than the Scythe S-Flex's used to at the same rpm. I was almost used to the GT's TBH, but these are real nice sounding by comparison I think if I put the GT's back in I'd be instantly irritated again. I also grabbed 2 x 92mm Noiseblocker blacksilent pro PE-1's to use as rear case exhaust fans and these sound great as well. Running at 1000rpm I can only barely hear them when I put my ear right up to them.

It's odd though, something about the huge motor and industrial feel of the GT's, even the motor noise they make, is 'tough' and I feel like a bit of a :banana::banana::banana::banana::banana: now with my noiseblockers. Weird I know but just being honest O.o

[tistou77]

Hi

Is it fans will be tested or not?

http://www.kitguru.net/components/co...-140mm-review/

They are excellent at first sight

Thanks

[lowfat]

I don't believe Martin does much for testing these days.

[tistou77]

Ok