Six 120mm fans: a brief sound/noise comparison at 12V

Closely moderated reviews by forum members willing to share their experience and accept stiff peer review. Open only to registered forum members.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee, Devonavar

Post Reply
Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Six 120mm fans: a brief sound/noise comparison at 12V

Post by Felger Carbon » Mon Jan 30, 2006 3:50 pm

Six fans under test (rated RPM and dBA, measured grams, measured RPM at 4100ft after 15 min. warmup), price, U.S. source. The Scythe fans are the new S-Flex models.

Scythe SFF21D 800RPM "8.7dBA" 175g 900RPM $15 Heatsinkfactory
Scythe SFF21E 1200RPM "20.1dBA" 177g 1270RPM $15
Akasa AK183-L2B 1400RPM "18dBA" 101g 1450RPM (not recommended)
Global Win NCB CS 1300RPM "19dBA" 136g 1360RPM $8 Coolerguys
Cool-Master SUV 1220RPM "22dBA" 117g 1230RPM $4 Clubit
Sunbeamtech LED 1200RPM "21.75dBA" 106g 1335RPM $6 SVC

My personal criteria in evaluating fan noise is that I don't like mechanical noises at all. I hate clangs, whines, scrapes, anything at all of that sort. Something continuous, such as air moving or even smooth sleeve bearing noise, doesn't bother me. I believe that, bearing noise aside, one fan will sound as loud as another at the same RPM as far as air-moving sound is concerned. YMMV.

In other words, I make a distinction between noise and sound. Now the comparisons.

Global Win Ceramic vs Sunbeamtech Silent LED: I had made this comparison before, but with a different Su and that time I couldn't tell the difference in sound/noise level, although the character of the sound was distinctly different. This time there was an obvious difference: The Su was louder, and the difference was more bearing sound/noise (although a non-objectionable sound/noise). I was using a different Su fan; the previous one had been mounted in a chassis. Same GW fan unit.

Global Win vs Scythe SFF21E: A repeat of the previous comparison. The 21E was louder, and the difference was the bearing sound/noise. Not as much difference as with the Su. Not an objectionable sound/noise.

C-M SUV vs SFF21E: 21E clearly louder, and again the difference was the bearing sound/noise. C-M bearing noise much less and lower in pitch. What little C-M sound/noise there is is pleasant.

Global Win vs C-M SUV: The two units' sound is very different in character. With that in mind, I couldn't detect a difference in sound/noise level. I could easily live with either fan.

Scythe SFF21D: Obviously nothing here to compare against. Very quiet indeed. Holding the fan 2 inches from my ear, the bearings made a very low, pleasant noise/sound.

Akasa: No point in comparing it. First, because of the higher RPM. Second, because of the annoying whine that's typical of ball-bearing fans. The spec sez the Akasa is 4dBA quieter than the C-M SUV. My ear sez the relationship is just the opposite. At least.

Finally, Su LED vs SFF21E: Too close to call. The character of the sound is even reasonably similar.
-----------------

The Su runs 105RPM faster than the C-M; that's about a 2dBA difference. This may account for most if not all of the difference I was hearing this time as opposed to the last time. The GW ceramic bearing fan is running fast as well (it's supposed to run fast) which makes its sound/noise vs. the other fans all the more surprising. It's a clear winner. I'm gonna buy a few more from CoolerGuys. If you like slow fans, you'll love the SFF21D; I have a spot for some of those. The Akasa is for ball-bearing lovers only (there are some of those!).

Remember we're only looking at a single unit of each fan here (2 of the Su counting my last comparison). How much variance is yet unknown and your reports can help in this regard.

I've made no effort to evaluate the various fans at lower voltages. I wanted to get this brief report to you this year. :)

Addendum next day:

I decided to see how the other fans would compare to the "800RPM" S-Flex fan. My sample turns exactly 900RPM here at 4100ft, so I set the C-M, Su, and GW fans one at a time to 900 with a Spire fan controller and compared them to the S-Flex 21D. In terms of increasing noise: GW, C-M, Su, 21D. But the 21D starts at 900RPM, 12V so if you want to go down to 500Hz, the 21D is the unit to have. At a distance of 1 foot, I could barely hear any of them and at 1 meter closed up in a case, forget it!

I also compared the GW, C-M, Su and 21E at exactly 1028.57RPM. That's the order they finished in, quietest to noisiest. None of these fans had any objectionable noise/sounds, it's just that some had more, some had less.

The GW has one issue: it does not like to shove air toward the center of the earth. A Nexus Silent 92mm has the opposite problem, it doesn't like to shove air away from the center. Reviews of the XP-90 sometimes find suck mode works best (air going up on the test bed), sometimes push mode (air going down on the test bed). In a tower case, the XP-90 pushes air sideways no matter how you mount the fan, so this is not an issue.

The Su is different from the rest because of the shape of its blade - it curves more than the others. Looking at it from the back, there's more of a 'spoon' effect. This may be why the Su seems to push more air at a given RPM than other fans I've used. And we use fans not to get a particular RPM, but to push air. For this reason, I'm keeping the Su on my list. Right now I like the $8 GW, the $6 Su, and the $4 C-M. I like the rated 80,000 hour life expectancy of the GW.

I have no doubt that the future (perhaps the near future) will bring us more quiet 120mm fans. There's a market for them. Competition will push prices down. SPCR seems to have a very conservative following, so I'm sure $18-$27 fans will continue to sell for awhile to SPCR folk.

Edit: added weight of C-M and Su; price and US source; and an addendum.
Last edited by Felger Carbon on Tue Jan 31, 2006 3:31 pm, edited 3 times in total.

GHz
Posts: 190
Joined: Fri Jun 24, 2005 12:54 pm

Post by GHz » Mon Jan 30, 2006 6:02 pm

Thanks for the quick review. I placed an order for one of these... I'm really interested in the bearing/motor noise characteristics at lowered voltages.

Weldingheart
Posts: 192
Joined: Wed Dec 07, 2005 6:04 am

Post by Weldingheart » Mon Jan 30, 2006 11:26 pm

Thanks, following your posts is interesting due to scientific contents/data.
I hope you would include the Nexus later in comparison so people can start having alternative than that fan(thanks there are anything to chose than orange), also how bout including the purchased price :wink:

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Sun Feb 05, 2006 8:12 am

Weldingheart wrote:I hope you would include the Nexus later in comparison so people can start having alternative than that fan(thanks there are anything to chose than orange). :wink:
Actually, GHz has twice posted that the Global Win is quieter than the Nexus. If he's right, then the much-cherished and seriously overpriced Nexus loses out to the $8 GW. Such a conclusion is, I suspect, unwelcome in this forum (to understate the matter!).

Since I am famous for avoiding controversy, I think I'll pass the GW-Nexus comparison on the excuse that the Nexus is out of my price range. :D

thetoad30
*Lifetime Patron*
Posts: 368
Joined: Fri Jul 23, 2004 11:33 am
Location: King of Prussia, PA

Post by thetoad30 » Wed Feb 08, 2006 4:20 am

Carbon,

Just a little critisism on your review:
Include some idea of spec airflow versus actual airflow, so we can compare values a little more easily. :) Otherwise, great review. I also saw another one you did on the 80mm version of the NCB and I'll be checking that one out too.

You say that the GW doesn't like to shove air toward the center of the earth. In what respects were you talking about? Noise? Airflow?

As for the Su fan that you keep mentioning, does that fan hold up to it's advertised 21.75dBA? I can't find the link to this fan again... do they make one without LEDs?

Thank you

NeilBlanchard
Moderator
Posts: 7681
Joined: Mon Dec 09, 2002 7:11 pm
Location: Maynard, MA, Eaarth
Contact:

GlobalWin NCB vs Yate Loon low speed

Post by NeilBlanchard » Fri Feb 10, 2006 9:28 pm

Greetings,

The GlobalWin NCB is a touch quieter than the Yate Loon low speed; at least comparing one of each fan. This is true at 12v, at 7v, and at 5v. The NCB barely starts at 5v, but it does get going -- and it is very quiet. Though I doubt it could exhaust a case very well -- it would be good for the front of a case at 5v because it is so quiet. I think the 19dBA rating is plausible at 5v.

The GlobalWin seems to run a bit slower, blow a bit less air, and it has occasional (muted) bearing noises while handling it and/or depending on orientation.

I don't have a Nexus at the moment, but between it and the NCB, it would be a very close call, in absolute terms -- but when cost is considered, the NCB is a much better value. Since it costs less than half as much...and if it lasts 80,000 hours then it is quite a bargain.

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Re: GlobalWin NCB vs Yate Loon low speed

Post by Felger Carbon » Sat Feb 11, 2006 4:08 am

NeilBlanchard wrote:The GlobalWin... has occasional (muted) bearing noises while handling it and/or depending on orientation.
GHz has also reported handling noises. I have a theory on that: if the noises only happen while changing the angle of the fan in some plane, but stop as soon as the angle-changing stops, then the problem might be gyroscopic precession.

A spinning fan has angular momentum in the plane in which it's spinning. Changing the angle at which you're holding the fan means the angular momentum in one plane has to be dumped. Gyroscopic precession rotates the angular shift 90 degrees and places a fairly fierce load on the fan bearing, which must dissipate the energy related to the angular momentum. [The fan bearing is the only thing that can dissipate this energy, as it is the only physical connection to the outside world that the spinning fan has!]

Mounted in a computer case, the fan is not going to change angle while spinning. This angular change is an artifact of hand-held fan evaluation.

Summary: angular momentum times mass equals energy. This energy is tied to the plane of rotation of the mass. Change the plane of rotation and the energy has to go somewhere.

Recommendation: when hand-holding spinning fans, change the angle of rotation slowly, and expect some bearing distress when doing so.
----------------------

Bearing noise "depending on orientation" is another matter. My GW NCB fans don't like to push air toward the center of the earth. The RPM drops by a large amount when the fan is placed in that position, indicating greatly increased bearing friction. The 92mm Nexus Silent fan has the opposite problem: it doesn't like to push air away from the center of the earth, also evidenced by a large drop in RPM when placed in this position. Some (most?) fans don't have this problem, but the GW and Nexus 92 do.

I use the term 'center of the earth' to distinguish the gravity problem from "down towards the motherboard", for instance. :)

NeilBlanchard
Moderator
Posts: 7681
Joined: Mon Dec 09, 2002 7:11 pm
Location: Maynard, MA, Eaarth
Contact:

Post by NeilBlanchard » Sat Feb 11, 2006 10:28 am

Greetings,

Yes, I can confirm the slowdown when the NCB fan is blowing down towards the ground, though I'd guess it is a ~100-150RPM decrease? This means it is not a good fan for a HS in a desktop case...

frankgehry
Posts: 1424
Joined: Sun Dec 12, 2004 5:00 am
Location: New York, NY

Post by frankgehry » Sat Feb 11, 2006 10:48 am

The best position for a sleeve/ceramic bearing fan is vertical. The sleeve bearing, a type of line contact bearing, works best when the weight of the fan assembly is distributed along the length of the shaft and has proper lubrication. When horizontally oriented, the weight of the fan blade assembly is supported by the end of the shaft. Ball bearings are point contact bearings and are preloaded during the manufacturing process.

http://synergy.sager.com/ProductPDFs/Fa ... Sleeve.pdf
Last edited by frankgehry on Sat Feb 11, 2006 3:54 pm, edited 1 time in total.

nici
Posts: 3011
Joined: Thu Dec 16, 2004 8:49 am
Location: Suomi Finland Perkele

Post by nici » Sat Feb 11, 2006 1:41 pm

Felger Carbon wrote:If he's right, then the much-cherished and seriously overpriced Nexus loses out to the $8 GW. Such a conclusion is, I suspect, unwelcome in this forum (to understate the matter!).
I would be very happy if i found a fan thats better than the nexus and not bloody ORANGE! :)

StarfishChris
Posts: 968
Joined: Fri Jan 07, 2005 7:13 pm
Location: Bristol, UK
Contact:

Post by StarfishChris » Sat Feb 11, 2006 3:54 pm

Felger Carbon wrote:If he's right, then the much-cherished and seriously overpriced Nexus loses out to the $8 GW. Such a conclusion is, I suspect, unwelcome in this forum (to understate the matter!).
We're not zealots (well, not many of us anyway ;)). Brand loyalty takes a back seat to the greater goal of silencing. FWIW a Yate Loon would be a good comparison, and it's not out of your price range either ;)

edlight1
Posts: 36
Joined: Tue Mar 08, 2005 8:33 am

Post by edlight1 » Tue Feb 14, 2006 5:49 pm

I've got a Yate Loon 120mm that I'm about to put in my psu blowing up, but I'm wondering if any bearing experts think it would be better to get something with ball bearings? Maybe a Vantec Stealth?

frankgehry
Posts: 1424
Joined: Sun Dec 12, 2004 5:00 am
Location: New York, NY

Post by frankgehry » Tue Feb 14, 2006 6:13 pm

What kind of psu do you have? Are you planning to use the original fan controller?

edlight1
Posts: 36
Joined: Tue Mar 08, 2005 8:33 am

Post by edlight1 » Tue Feb 14, 2006 6:32 pm

I have a Coolmax 550 with a bottom 120mm fan. The onboard controller has a switch for low, auto, and max. At low and auto it's too slow and gets hot, at max it's too loud. Whatever I put in there will go to a Zalman 5 1/4" bay controller on a channel that's monitored for speed by the motherboard's software app, so I can tweak it to where I like it for different conditions and note the rpm to set back there later. I'll get an alarm if it quits or slows down too much.

frankgehry
Posts: 1424
Joined: Sun Dec 12, 2004 5:00 am
Location: New York, NY

Post by frankgehry » Tue Feb 14, 2006 7:06 pm

I would use a papst 4412F/2GL (1600rpm) or 4412/2GLL (1200rpm). It uses a sintec bearing and will last longer in a psu than a simple sleeve bearing. You can buy the 2GLL here - http://www.sharkacomputers.com/paulqulowspc.html

It's on the expensive side, but well made.

I wasn't able to locate a 1600rpm version.

edlight1
Posts: 36
Joined: Tue Mar 08, 2005 8:33 am

Post by edlight1 » Tue Feb 14, 2006 7:31 pm

frankgehry wrote:I would use a papst 4412F/2GL (1600rpm) or 4412/2GLL (1200rpm). It uses a sintec bearing and will last longer in a psu than a simple sleeve bearing. You can buy the 2GLL here - http://www.sharkacomputers.com/paulqulowspc.html

It's on the expensive side, but well made.

I wasn't able to locate a 1600rpm version.
Thanks!

Unfortunately it says it won't run below 8v, and I'll want to vary the speed with a Zalman controller, which will lower its top speed too, since it tops out at 11v.

frankgehry
Posts: 1424
Joined: Sun Dec 12, 2004 5:00 am
Location: New York, NY

Post by frankgehry » Tue Feb 14, 2006 7:42 pm

use froogle

www.froogle.com

search for papst 4412

edlight1
Posts: 36
Joined: Tue Mar 08, 2005 8:33 am

Post by edlight1 » Tue Feb 14, 2006 7:53 pm

frankgehry wrote:use froogle

www.froogle.com

search for papst 4412
They don't have rpm sensing. Darn, that was important as I'll be monitoring them while varying the rpm.

I'm beginning to see myself putting in the Yate Loon and if the bearing goes poof, going into town and getting a Vantec Stealth.

frankgehry
Posts: 1424
Joined: Sun Dec 12, 2004 5:00 am
Location: New York, NY

Post by frankgehry » Tue Feb 14, 2006 7:59 pm

Seasonics use to have sleeve bearing yate loons, so they will probably be ok, but would not be my first choice.

edlight1
Posts: 36
Joined: Tue Mar 08, 2005 8:33 am

Post by edlight1 » Tue Feb 14, 2006 8:05 pm

frankgehry wrote:Seasonics use to have sleeve bearing yate loons, so they will probably be ok, but would not be my first choice.
That's great to know!

NeilBlanchard
Moderator
Posts: 7681
Joined: Mon Dec 09, 2002 7:11 pm
Location: Maynard, MA, Eaarth
Contact:

Post by NeilBlanchard » Tue Feb 14, 2006 8:15 pm

Hello,

The older S12's and the previous SeaSonic's, as well as the many of the Fortron Source PS's used the medium speed sleeve bearing Yate Loon 120mm fans. They are very quiet and mine have lasted 3+ years running 24/7. They are as quiet as they ever were.

It's the ceramic bearing GlobalWin NCB that slows down when horizontally oriented...

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Tue Feb 14, 2006 9:55 pm

NeilBlanchard wrote:The older S12's and the previous SeaSonic's, as well as the many of the Fortron Source PS's used the medium speed sleeve bearing Yate Loon 120mm fans. They are very quiet and mine have lasted 3+ years running 24/7. They are as quiet as they ever were.

It's the ceramic bearing GlobalWin NCB that slows down when horizontally oriented...
Neil, the GW NCB slows down only when blowing air down. It's fine when blowing up, as in a PSU. In fact, as soon as I get over the lazies, I'm gonna (re)mod a PSU or two to use GWs instead of C-M SUFs. :)

For the record, the 92mm Nexus Silent does have problems blowing air upward, as in a PSU in a conventional tower case.

dfrost
Posts: 525
Joined: Mon Feb 07, 2005 11:57 pm
Location: Seattle, WA

Post by dfrost » Wed Feb 15, 2006 11:54 am

Regarding Papst fans, edlight1 wrote:Unfortunately it says it won't run below 8v
My Papst 4412 F/2GL starts and runs just fine at 5V.
and later wrote:They don't have rpm sensing.
Look for the F/2GL or F/2GLL models. The "/2" indicates speed sensing.

I purchased mine at a local shop in Delft while vacationing in Holland last summer, so no help with your possible purchase.

My Yate Loon'd S12-430 also continues to run quietly after almost a year.

Edit:
edlight also wrote:Maybe a Vantec Stealth?
No! Those are not quiet or remotely stealthy.

edlight1
Posts: 36
Joined: Tue Mar 08, 2005 8:33 am

Post by edlight1 » Fri Feb 17, 2006 8:37 pm

I put the Yate Loon into the psu today. I used Vantec rubber gaskets on the fan and the power supply too. I nibbled a hole for the wire to come out. At first I hooked it to my Zalman 5 1/4" bay fan controller. Amazingly, I couldn't hear a difference between full-on and all the way down. There's a bit of ambient noise from a freeway, though.

At full-on the air coming out of the psu was a little warmer than I'd like, so I unhooked it from the controller, which only gives out 11v, and hooked it into the motherboard directly. It went up to around 1275 rpm (fluctuates a bit). Here it's at the perfect compromise speed.

That's how it turned out. So much quieter! 8)

Tyrven
Posts: 42
Joined: Tue Feb 21, 2006 2:45 am
Location: Seattle, WA, USA
Contact:

Cooler Master Performance?

Post by Tyrven » Sun Feb 26, 2006 10:27 pm

You rated the Cooler Master amongst your select picks (presumably for the price vs. noise ratio). However, ClubIT reports that it's airflow is a mere 11.27CFM (compared to 33-48 for the other fans).

Did you find that its airflow was actually limited or was Cooler Master just being conservative with their numbers?

P.S., Thank you for this review; this covers several of the fans I was considering (in particular the Scythe, which is one of the few quiet fans carried by NewEgg).

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Re: Cooler Master Performance?

Post by Felger Carbon » Mon Feb 27, 2006 1:05 am

Tyrven wrote:You rated the Cooler Master amongst your select picks (presumably for the price vs. noise ratio). However, ClubIT reports that it's airflow is a mere 11.27CFM (compared to 33-48 for the other fans).
First, the fan in question is currently sold out. Aside from that, C-M does make a 720RPM very quiet fan that only pushes 11.27CFM. Clubit has mistakenly posted that fan's specs on the SUV, but then continuing on has the correct specs for the 1200RPM SUF fan. Others carry this fan, including SVC.com, but at a higher price. At the higher price, I'd suggest the $8 Global Win NCB fan at CoolerGuys.com instead.

Tyrven
Posts: 42
Joined: Tue Feb 21, 2006 2:45 am
Location: Seattle, WA, USA
Contact:

Post by Tyrven » Mon Feb 27, 2006 1:12 am

Ah, OK that makes sense. Thank you. After reading this and a number of other reviews, I'm planning on getting the GlobalWin. I wish they were better distributed; I can only find them at a couple of retailers. Fortunately, CoolerGuys have a good rep.

Quick question: Do the GlobalWins have open screw mounts, such that they can be used with rubber grommits (or attached to the Thermaltake SI-120)?

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Mon Feb 27, 2006 3:14 am

Tyrven wrote:Do the GlobalWins have open screw mounts, such that they can be used with rubber grommits (or attached to the Thermaltake SI-120)?
No, the GWs have bridged ears exactly like the Yate Loon/Nexus fans. But Jab-Tech sells this "top-rank" wire clip, which also fits the SI-120, to mount fans using bridged fans.

If you want to mount the GW as an exhaust fan, I recently posted in the Fans and Control forum (look for "penguins, nasty ogre" :) ) how you can do that, and Jab-Tech is also one of the sources I suggested for rubber fan mounts.

edit: corrected forum

Post Reply