S-Flex SFF21D 120mm "8.7dB" mini review

Control: management of fans, temp/rpm monitoring via soft/hardware

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
ckolivas
Posts: 393
Joined: Thu Jan 27, 2005 7:16 am
Location: Melbourne, Australia
Contact:

S-Flex SFF21D 120mm "8.7dB" mini review

Post by ckolivas » Wed Nov 22, 2006 5:14 am

Ok here's a summary of the manufacturer's data.
Scythe S-FLEX 120mm Fan SFF21D

Features for Scythe S-FLEXâ„¢ 12cm Case Fan

Equipped with Genuine S-FDB* Bearing

This fan is equipped with genuine S-FDB (SONY Fluid Dynamic Bearing) made by SONY corporation.

Market Leading Low-Noise Level

S-FDB technology drastically eliminates the friction between the bearing and the shaft which makes this fan nearly inaudible to human ears. Additional suction magnet optimizes the magnetic force balance within the rotor to achieve the absolute silence. To further achieve the absolute silence, the fan is equipped with the Linear drive IC for reducing electromagnetic noise.

Outstanding Long Life of 150000 Hours

Patent Pending "S-FDB" structure allows the fan to last up to 150000 hours.

Anti-Shock Ability

Conventional ball bearing carries a point contacting structure for the rotation, which gives point supporting situations at dropping impact to damage ball bearing surface which could cause irregular noises. The S-FDB carries a surface contact mechanism for improved anti-shock structure, which is 5 times stronger than the conventional ball bearing.

Scythe S-FLEX 120mm Fan SFF21D

Features for Scythe S-FLEXâ„¢ 12cm Case Fan

Equipped with Genuine S-FDB* Bearing

This fan is equipped with genuine S-FDB (SONY Fluid Dynamic Bearing) made by SONY corporation.

Market Leading Low-Noise Level

S-FDB technology drastically eliminates the friction between the bearing and the shaft which makes this fan nearly inaudible to human ears. Additional suction magnet optimizes the magnetic force balance within the rotor to achieve the absolute silence. To further achieve the absolute silence, the fan is equipped with the Linear drive IC for reducing electromagnetic noise.

Outstanding Long Life of 150000 Hours

Patent Pending "S-FDB" structure allows the fan to last up to 150000 hours.

Anti-Shock Ability

Conventional ball bearing carries a point contacting structure for the rotation, which gives point supporting situations at dropping impact to damage ball bearing surface which could cause irregular noises. The S-FDB carries a surface contact mechanism for improved anti-shock structure, which is 5 times stronger than the conventional ball bearing.

Specifications for Scythe S-FLEXâ„¢ 12cm Case Fan

Model Number: S-FLEX SFF21D (800rpm)

Dimensions: 120 x 120 x 25mm

Specifications: 800rpm Version: 33.5CFM / 8.7dBA / DC12V / 0.10A

Connector: 3-pin (4-pin adaptor included)

Cable Length: 30cm

Bearing Type: S-FDB (Fluid Dynamic Bearing by SONY Corporation)

MTBF: 150 000 hours

S-FLEXâ„¢ is a registered trademark of SONY Corporation
There have been some interesting reviews on the 80mm fans, and some on the higher speed "E" 120mm fans, and other websites have shown some good figures on these D fans so I thought I'd buy a few and try them. I purchased 4 from pccasegear.com in Australia. They were excellent in delivering them less than 24 hours from my order to my door.

They come packaged in a stiff plastic shell with an outer plastic wrapper.
(most photos have enlargements just click on them)

Image


The detail on the top of the package says the fan appears to be basically made by Sony.

Image


Lots of advertising bullsh... jargon on the back.

Image


The contents are pretty standard 3 pin based, with a 4 pin adapter that nicely includes an rpm monitor you can plug in

Image


The fan has all important partially open corners and looks like so close up:

Image


Finally they mount without a hitch on the Thermalright ultra 120 (one of the uses I had in mind for them)

Image

Ok so on to the review.

On visual inspection there are no outstanding features of this 7 blade fan and it has zero bling value which to me is a good thing. It is about average weight for a fan of this size (unlike say the Aerocool fans which are much lighter).

I'm fortunate in that I live in an extremely quiet area and the downstairs of my house at night is nigh on silent where a Nexus 120mm fan at 12V is downright noisy.

So I plugged it in and fired it up via various means. Many of my colleagues have criticised me at work for giving the bad news last which always downplays things so I'll reverse my usual stance, and do the ugly, the bad and then the good.

Bear in mind that this fan runs at 800rpm at maximum voltage...

This fan buzzes like a blowfly in heat with pwm control so that as an option to undervolt it is basically out.

8.7dB? Not in your wildest dreams. I certainly didn't believe it on the ad alone, and in the flesh I could tell it definitely was not 8.7dB.

Most fans are known to start at 7V. This one? Nope.

So it doesn't really start at low voltage. However it did go down in voltage once started, but I was not really happy that the way it ran at 7V would be reliable. A zalman fan mate was a decent match for it, but could easily drop it to a speed below which it refused to turn.

Ok that's the ugly and the bad out of the way. Now onto the good.

Vibration in the hand test? Very little, and almost nothing transmitted to the heat sink when it was mounted even without the rubber stoppers included in the ultra 120. Actually at equivalent voltage/speed the Nexus felt slightly smoother, but it didn't seem to make an audible difference.

This is one damn quiet fan at full voltage. At 12V I compared its noise signature and apparent loudness to a Nexus 120mm at 7V. They are very similar in loudness, with a higher pitched sound from the s-flex. Neither fans' sound characteristic was unpleasant. Furthermore it didn't make a difference what direction I held the fan at which may be nice for unusual position mountings of PSUs and heatsinks. The s-flex was only really audible in the presence of other components when I got up close to it (less than a foot). Mounted on the heatsink it almost seemed quieter (but that could have been my imagination), however it definitely was not noisier. Once mounted in the box I could not discern the sound of the fan over any other components, the WD5000KS suspended in an Antec Solo being the loudest.

Well the main attraction for me to this fan was that it was going to be probably the right compromise in terms of flow and noise at default voltage. I still feel uncomfortable running too many 7V things off the 12/5V split from the power supply since potentially I could feed back voltage to the PSU since less and less components are using 5V. The zalman fanmate was great apart from the fact it didn't let you turn it up to maximum voltage since there was always a drop. And pwm has failed to impress me on most fans I've used it on so it's something I basically never use except for one unusual fan I got in the past that seems to love it.

So what is my conclusion? Compared to the reference? The Nexus 120mm is a more versatile fan overall since its flow has greater flexibility at different voltages and can reach higher levels but at equivalent revolutions there is barely anything to distinguish them. There is great appeal in a fan that is working at the manufacturer's settings and achieves noise levels I desire. There is also the theoretical longevity advantage of the bearings, and the open corners that don't require cutting. Conclusion is that if you do not want to fiddle with voltage and want something below the noise level of other components, then this is the fan for you. Otherwise get a Nexus. In Australia, at least, these fans are priced almost the same.

andyb
Patron of SPCR
Posts: 3307
Joined: Wed Dec 15, 2004 12:00 pm
Location: Essex, England

Post by andyb » Wed Nov 22, 2006 7:48 am

Nice review, I am looking forward to the SPCR face off of 120mm fans, both the Nexus, and various S-Flex fans will be involved.

I got a 1200rpm S-Flex with my Ninja, and just like the 800rpm model you have got, it buzzes, ticks, and clicks very annoyingly, unfortunatly I dont have a molex ---> 3pin adaptor to test it at a set voltage, so I have never been able to compare it to my Nexus's (or are the called "Nexi"???).

I also find a nexus @ 12v loud, but at 5v, they are whisper quiet (not silent) and very very smooth, I have to concentrate to hear them, and even then i am not sure if its my imagination, I am guessing that they are running @600rpm.


Andy

sjschwinn
Posts: 68
Joined: Wed Jul 05, 2006 1:38 pm

Re: S-Flex SFF21D 120mm "8.7dB" mini review

Post by sjschwinn » Wed Nov 22, 2006 8:07 am

ckolivas wrote: A zalman fan mate was a decent match for it, but could easily drop it to a speed below which it refused to turn.

..

So what is my conclusion? Compared to the reference? The Nexus 120mm ... at equivalent revolutions there is barely anything to distinguish them. ... There is also the theoretical longevity advantage of the bearings
Nice review. I also just purchased the S-Flex D, and have a few additional comments.

I have also tried running it with a Fanmate, and although I haven't checked the actual voltages, I can run the fanmate all the way down to it's lowest setting - even while running 2 other fans in parallel - and although it won't start at that voltage, it runs just fine there.

I do agree with the rest of your review. I find that although it's not quite as smooth sounding as the Nexus or YL, the motor/bearing noise was as non-irritating as any other fan I've tried (and I've tried several of the recommended fans here). It was noticeably better than the GlobalWin 120mm that I've seen recommended around here (which is actually a decent fan), which I bought at the same time.

Lastly, the comment on the longevity is why I decided to purchase this fan. I plan on installing it blowing downward in my P180, and I didn't like the idea of running any of the sleeve bearing fans in that orientation (many of which make more noise in that orientation anyway).

sjschwinn
Posts: 68
Joined: Wed Jul 05, 2006 1:38 pm

Ninja Plus fan?

Post by sjschwinn » Wed Nov 22, 2006 8:10 am

andyb wrote:I got a 1200rpm S-Flex with my Ninja, and just like the 800rpm model you have got, it buzzes, ticks, and clicks very annoyingly, unfortunatly I dont have a molex ---> 3pin adaptor to test it at a set voltage, so I have never been able to compare it to my Nexus's (or are the called "Nexi"???).
Are you referring to the fan included with the Ninja Plus? I have the Ninja Plus, and I'm almost certain the fan that was included with it was NOT the S-Flex. (I'm not using that fan, BTW.)

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

Re: Ninja Plus fan?

Post by Felger Carbon » Wed Nov 22, 2006 10:12 am

sjschwinn wrote:Are you referring to the fan included with the Ninja Plus? I have the Ninja Plus, and I'm almost certain the fan that was included with it was NOT the S-Flex. (I'm not using that fan, BTW.)
I own both fans (SFlex E, NJ+) and I'm sure they aren't the same fan. But amazingly, the motor noise sounds exactly the same, both in volume and in character (frequency characteristics). Blindfolded and just listening, I'd defy you to tell the difference! BTW, this NJ+ is not the rev B fan.

andyb
Patron of SPCR
Posts: 3307
Joined: Wed Dec 15, 2004 12:00 pm
Location: Essex, England

Post by andyb » Wed Nov 22, 2006 10:16 am

It's NOT an S-Flex like I thought, and it did come as part of the Ninja Plus bundle.

Its a scythe branded, AD1212DS-A73GL(6TCL3)
DC 12v @ 0.13A

I got confused because the Plus version of the Ninja came out at about the same time as the S-Flex so I put 2 and 2 together and got a crap fan.


Andy

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

Post by NeilBlanchard » Wed Nov 22, 2006 10:42 am

Hello,

The number of Scythe models is fairly confusing. I have only used the Scythe DF (by Young Lin), which is being discontinued. It is a 1200RPM model, and it is slightly quieter than the GlobalWin NCB (which has/had a very mild bearing noise). Both were much quieter than a low speed Yate Loon.

The "8.7dBA" rating might be a little more believable with a 1 in front of it, but even 18.7dBA (@ 12v) would be a virtually silent fan.

ckolivas
Posts: 393
Joined: Thu Jan 27, 2005 7:16 am
Location: Melbourne, Australia
Contact:

Post by ckolivas » Wed Nov 22, 2006 12:29 pm

Yes I would say 18.7dB is probably close to what this fan sounds like given comparisons with other fans. Probably the best thing I can say about the sflex D is it is the quietest available 120mm fan on the market... but then it is also the lowest flow. As regular SPCRers know, though, it is the only way since the laws of physics still need to be obeyed ;)

nightmorph
Posts: 316
Joined: Thu Aug 10, 2006 11:07 am

Post by nightmorph » Mon Dec 11, 2006 10:15 pm

I have an S-Flex-lookalike, the one that came with my Ninja. However, mine is even louder than yours...possibly because it arrived in very, very small pieces? :)

Here in the US at least, S-Flex prices aren't quite as bad as the typical Nexus price of $14.

ckolivas
Posts: 393
Joined: Thu Jan 27, 2005 7:16 am
Location: Melbourne, Australia
Contact:

Post by ckolivas » Mon Dec 11, 2006 10:29 pm

Heh that can't be good. The fans that come with the heatsinks aren't the same as the s-flex.

Also it's nice that the recent 120mm fan roundup by SPCR agrees with what I've said here about this fan.
http://www.silentpcreview.com/article695-page5.html

nightmorph
Posts: 316
Joined: Thu Aug 10, 2006 11:07 am

Post by nightmorph » Mon Dec 11, 2006 10:37 pm

ckolivas wrote:Heh that can't be good. The fans that come with the heatsinks aren't the same as the s-flex.

Also it's nice that the recent 120mm fan roundup by SPCR agrees with what I've said here about this fan.
http://www.silentpcreview.com/article695-page5.html
Well, I bought my fan/heatsink combo at a discount, knowing ahead of time the fan was damaged in shipping to the e-tailer.

Actually, the Kama Flow really looks like a more attractive fan overall. Too bad it doesn't seem to be available outside of Japan. Still, the Scythe is available, and is only $10US at NewEgg (for the E model). Other online stores such as Jab-Tech sell it at $14 or so, and that's the same price as a Nexus, so there's not as much reason to choose it over the Nexus, in my opinion. In fact, the lower RPM versions are even more expensive than a Nexus, making it much more cost-effective to go with a Nexus and buy some fan controllers or diodes for voltage reduction with the remaining price difference.

(Offtopic: didn't know you hung out here. We quite like ck-sources over at Gentoo. :D)

ckolivas
Posts: 393
Joined: Thu Jan 27, 2005 7:16 am
Location: Melbourne, Australia
Contact:

Post by ckolivas » Tue Dec 12, 2006 2:37 am

nightmorph wrote: (Offtopic: didn't know you hung out here. We quite like ck-sources over at Gentoo. :D)
Thanks :D

(/offtopic)
You might like my recent addition then:
viewtopic.php?t=36474

sjschwinn
Posts: 68
Joined: Wed Jul 05, 2006 1:38 pm

Post by sjschwinn » Tue Dec 12, 2006 10:40 am

nightmorph wrote: so there's not as much reason to choose it over the Nexus, in my opinion.
The reason I chose the S-Flex was for the non-sleeve bearing for use in a "blow-up" or "blow-down" orientation. In my experience with sleeve bearing fans, when placed in that orientation they don't sound nearly as good, and the longevity/durability is very poor. The S-Flex should stay quiet much longer than the Nexus in a horizontal orientation. (Another example of this is when Seasonic went with the Adda ball-bearing fans over the YLs in their PSUs due to durability issues.)

ckolivas
Posts: 393
Joined: Thu Jan 27, 2005 7:16 am
Location: Melbourne, Australia
Contact:

Post by ckolivas » Tue Dec 12, 2006 2:23 pm

I agree. The value of the SFD bearings has not been emphasised enough in most of these comparisons (except by the SPCR crew) where the extra flexibility of possible usage applications and the potentially much longer life expectancy are of great value. Especially considering all of my pcs are on 24/7.

nightmorph I forgot to say I've been thankful for the gentoo community over the years for the extensive testing they've done on my patchset and the feedback I've gotten. If ever there was one failing on their part it was the preoccupation with merging patchsets of early code which then ends up being 1. a bad advertisement for the early code in question 2. provides no direct reliable path for feedback/bug reports to the original coders and 3. because of inappropriate merging the code is not being properly tested. Nonetheless the community's interest in new and exciting code has made for a great testing ground for kernel hackers.

Post Reply