120x38mm fans with high Pressure at Lowest dBA

[jonjan]

edited: this thread originally didn't specify 38mm thickness.

I've received the Nexus 120mm D12SL-12, and the Noctua NF-P12... and both are 25mm thick.
The thickness of a 38mm should provide more static pressure.

So the thread is now.... what are the 120x38mm fans that have the highest pressure and are quiet? I'll be dropping the voltage to reach almost silent at 1m db levels (same as SPCR's level of 19dB, or xbitlabs level of 33dB)




edit: this thread started out as a request for help. Now that I've started a list here, the thread now just welcomes posts of new fans not already listed, that have above average pressure, and low dBA.

Above Average Pressure is defined as... I don't know All I see are numbers.. no idea how they compare or what they mean in the real world. What is the mm H2O of an ordinary everyday 120mm?

Low dBA is defined as 20 dBA or less.

CFM needs to be 50 or more.

Fan size - the smaller the better[/b] .


So add any fans to the list that you know qualify (and are available to be purchased)...

And if you have any personal experience with any of the fans, please add that too.



thanks!


-----------------------------------------------------------------
LIST section to be updated----


checked against http://translate.google.com/translate?h ... id%3D29175
where available

Akasa Apache 120mm - AK-FN057
2.64 mm H2O at 16 dBA and 57 CFM - at 12v
20.5 dBA 56 CFM
http://www.akasa.com.tw/update.php?tpl= ... l=AK-FN057

Be Quiet SilentWings USC 80mm - BL011
2.35 mm H2O at 15dBA and 26 CFM - at 12v

http://www.be-quiet.net/be-quiet.net/in ... teLang=en#

120mm SilentWings (for comparison)
2 mm H2O at 17dBA and 50 CFM
23 dBA 52 cfm

Noctura P12 120mm - NF-P12
1.68 mm H2O at 19.8dBA and 54 CFM - at 12v
25 dBA
http://www.noctua.at/main.php?show=prod ... g=en&set=1

Akasa 180mm Clear Fan - AK-F1825SM
.33 MM H2O at 15dBA and 51 CFM
http://www.akasa.com.tw/update.php?tpl= ... AK-F1825SM

Scythe Kama FLEX 135 Series - 135mm - SA1325FDB12L
16Dba and 50 CFM

--------------------
using new measures -xbit labs
(all CFM numbers are roughly half of manufactr #s)
http://www.xbitlabs.com/articles/cooler ... -1_20.html
33 Dba is "acoustic quiet" by their meathods

Noiseblocker NB-Multiframe S-Series MF12-S2 - 1250rpm
1.1 mm H2O and 51 CFM
"acoustic quiet" 16-24 CFM (9.9v)
http://www.xbitlabs.com/articles/cooler ... html#sect0
http://www.xbitlabs.com/misc/picture/?s ... ig.png&1=1

Be Quiet SilentWings PWM BL023
acoustic quiet 17.2 CFM 8.5v
http://www.xbitlabs.com/picture/?src=/ ... bq_xbt.png IMAGE

NoiseBlocker 80mm S2
14 dB 1.4 mm H2O 27 CFM
http://www.frozencpu.com/products/8398/ ... 4_dBA.html

Cooler Master SickleFlow 120 (R4-C2R-20AC-GP)
acoustic quiet 2.3 mm H2O 17 CFM at 5v
http://www.xbitlabs.com/picture/?src=/i ... cm_xbt.png IMAGE
http://www.xbitlabs.com/articles/cooler ... html#sect2

Scythe Ultra Kaze (DFS123812L-2000) 120x38
acoustic quiet 19 CFM 5v

Scythe Kama Flow 2 900rpm (SP1225FDB12L)
acoustic silent (31.5) 17 CFM 12v NO STATIC INFO
http://www.xbitlabs.com/picture/?src=/i ... gr_xbt.png CHART
http://www.xbitlabs.com/articles/cooler ... html#sect0


Scythe Kama Flow 2 120 1400rpm (SP1225FDB12M)
acoustic quiet 17.6 CFM 8.5v
http://www.xbitlabs.com/picture/?src=/i ... gr_xbt.png CHART
http://www.xbitlabs.com/articles/cooler ... html#sect0


Triebwerk TK-121 120x55 (blade is 30 wide)
acoustic quiet 17.7 CFM 8v
http://www.xbitlabs.com/picture/?src=/i ... gr_xbt.png CHART

http://www.xbitlabs.com/articles/cooler ... html#sect0

Nanoxia DX12
acoustic quiet .4 mm H2O 17.4 CFM
http://www.xbitlabs.com/articles/cooler ... html#sect0

[ces]

The dba numbers are not comparable from one tester to another, nor are the pressure numbers.

Theoretically, all things equal, a fan optimized to work under high static pressure should deliver more cfm with less db when facing high static pressure at the cos of delivering less CFM with more db in comparison to a fan optimized to operate under less static pressure.

That theory seems to be irrelevant to the fans commonly used to cool personal computers.

Also db ratings are the same as quietness to the human ear. The db ratings measure a spectrum of sound. Certain spectrums of sound of equal db will present themselves differently as noisy or quiet, pleasant or unpleasant, to the human ear.

The Scythe Gentle Typhoons are probably the best commonly available fans for use under high static pressure. They are very quiet. They are the only commonly available PC fan that uses real ball bearings. That gives them a long life. Unfortunately ball bearings fans don't ship well, so a certain percentage of them are noisier than others. If you really are going for quiet you may need to purchase several in order to assure you get a good one... though even the bad ones are still pretty good.

Noise blockers seem to generally be pretty good under high static pressure and are fairly quiet.

The high static pressure performance of Noctua and Nexus fans seems to be mixed, depending on who is testing them. Looking at SPCR testing it appears that Nexus is better under high static pressure but I have seen contrary results elsewhere.

I own an Akasa but have not yet used it. From what people report it may be a bit nosier than the above fans...though one of the most important factors affecting noise is fan speed... and it runs fairly slow compared to most other PWM fans. Scythe has the slowest running PWM fans. I wouldn't rely on them for high static pressure use... though many people swear by them for use on CPU heatsinks (which is a relatively high static pressure type of application).

What makes you think that you need a high static pressure fan? How do you intend to use it?

[ces]

Just a followup.

One number can't adequately describe high static pressure fan performance.

You need a graph of CFM against static back pressure.

You also need a graph of db against CFM.

Then you need a series of these graphs for a multiple of speeds or voltages.

And you still aren't done. The same number of db for one fan will sound different from the same number of db for another. That can be captured with a graph of sound spectrum... but it is best assessed subjectively by a human being listening to the character of the sound.

Also you shouldn't compare 80mm fans with 120mm fans. All things equal the 80mm fans will operate at higher pressure but will generate lesser cfm.

[jonjan]

The static pressure is to pull air through a 1" thick filter. My system cools acceptably with a single 50+ CFM fan.
The fan size doesn't matter, I'll modify the hole. I just want the pressure, CFM, and quiet.

I used to use a SilenX 120mmx38mm, and it had enough pressure.

I also had some random 60CFM 120mm fan that had enough pressure. I forget which fan it was.

add: So I guess I'm really not needing a super high-pressure fan. Probably a slightly above average pressure would do okay.
If I could find the mm H2O for a common 120mm, I'd just need a fan that matches that number or is higher.

Any idea what the mm H2O is for a common simple 120mm fan?

[ces]

My system cools acceptably with a single 50+ CFM fan.

The CFM was no doubt measured in free air without any static pressure. That is usually how most cfm numbers are measured.

One inch is a pretty thick filter. Do you really need it? Take a look at the modright filterright filters.

In any event, try the Gentle Typhoon D1225C12B5AP-15 (1,850 rpm). Then experiment with different speeds. At 5 volts it should run about 600 rpm or so.

Once you determine that it is generating the cooling you want then focus on whether you got buy a few more to get a quiet one. Again... even the loud ones are fairly quiet.

I would also experiment with a 120mm nexus fan.

[jonjan]

My system cools acceptably with a single 50+ CFM fan.

The CFM was no doubt measured in free air without any static pressure. That is usually how most cfm numbers are measured.

One inch is a pretty thick filter. Do you really need it? Take a look at the modright filterright filters.

In any event, try the Gentle Typhoon D1225C12B5AP-15 (1,850 rpm). Then experiment with different speeds. At 5 volts it should run about 600 rpm or so.

Once you determine that it is generating the cooling you want then focus on whether you got buy a few more to get a quiet one. Again... even the loud ones are fairly quiet.

I would also experiment with a 120mm nexus fan.

True the CFM is measured without pressure. The 60 CFM fan I used provided adequate CFM when used with the filter.

I want the 1" filter to ensure that the air will be 100% clean. I want that to stay the same, and only want to change the fan. I need a replacement for the silenx 120x38. I figure that after 2 years+, there must be a fan that can give the same pressure and cfm, at even less dB. thanks for the idea on the thinner filter tho.

It looks like I'll need to power a fan at lowered voltage in order to get the quiet I want.
Still possible to find an acceptable fan at stock. I just saw the NoiseBlocker 80mm S2: 14dB - 27 CFM - 1.4 mm H2O.

I know the specs aren't neccessarily reliable from sellers. I've been looking for either apparently-non-biased measurements, or personal testimonies on forums.
so far xbits seems helpful
http://www.xbitlabs.com/articles/cooler ... dup-1.html

[Lithium466]

Then try Gentle typhoon 1450rpm (AP14) ! Or Nexus real silent basic fan, like ces said.

I doubt NB M8-S2 will give you the cooling performances you need...

Have a look here : http://www.xbitlabs.com/articles/cooler ... up_18.html

[jonjan]

I checked again and can't find any static pressure numbers for the Gentle Typhoon.


So far, I've checked the xbit lists for all the 120mm and 140mm fans. After comparing numbers I also noticed that they tend to give a CFM result of half what the manufacturer gives, at 12v. I don't think manufacturers fudge that much? The industry should standardize the CFM and dB measurements!

I've revised my goals also.. if a fan can give at least 5 CFM, after sucking through the 1" filter, that'll be enough. Of course with the margin of error of guessing, I'd rather shoot for at least 10 CFM.
I figure a fan with a high enough static pressure should be able to retain most of it's CFM after going thru the filter. This is 80% guess and 20% guesstimate.


So far, here's the best of all those lists...

xbitlabs gives a distinction of "acoustic quiet" for fans that come in at 33dB measured by their technique. I'm assuming this is the same as an acoustic quiet room level.

note- for undervolting fans, the effect on static pressure (SP) is:
SP new= ( (RPM new / RPM old) squared ) * SP old
http://docs.google.com/viewer?a=v&q=cac ... rMO5HCUiyA

sorted by combo of highest static pressure, highest CFM, and room to reduce the rpm further...


Triebwerk TK-121 - 120x55 (blade is 30 wide) - 60CFM - 1.6 mmH20 - $25
dB=acoustic quiet - 17.7 CFM - 8v (920rpm in tube – of 1200) = .9 mmH2O
http://www.xbitlabs.com/articles/cooler ... html#sect0

Be Quiet SilentWings PWM BL023 – 50 CFM - 1.6 mmH2O - $22
acoustic quiet - 17.2 CFM - 8.5v (1000rpm in tube – of 1500) = .7 mmH2O
http://www.xbitlabs.com/articles/cooler ... html#sect0

Noiseblocker NB-Multiframe S-Series MF12-S2 - 51 CFM - 1 mmH2O
acoustic quiet - 16 CFM - 9.9v (1000rpm in tube - of 1250) = .6 mmH2O
http://www.xbitlabs.com/articles/cooler ... html#sect0

Cooler Master SickleFlow (R4-C2R-20AC-GP) - 69 CFM - 2.3 mmH2O - $10
acoustic quiet - 17 CFM - 5v (1000rpm in tube – of 2000)) = .6 mmH2O
http://www.xbitlabs.com/articles/cooler ... html#sect2

[ces]

I checked again and can't find any static pressure numbers for the Gentle Typhoon.

There are no published numbers that I know of. Take a close look at the fan blades. Notice that there is not much gap for air under pressure to slip backwards. Notice the shallow attack angle. They have to rev at higher RPMs to deliver the same free air cfm as the other more conventional fan blade designs.

But their CFMs should not degrade so quickly on pressure. And the blades will be more resistant to luffing as increased back pressure fights the pressure being generated by the blades.

After comparing numbers I also noticed that they tend to give a CFM result of half what the manufacturer gives, at 12v. I don't think manufacturers fudge that much? The industry should standardize the CFM and dB measurements!

There are standards for db measurements I believe. Nexus follows them. No one else does. You can call it fudging or whatever but the db measurements are not standardized and therefor can't be compared. Even if they could be, one fan with lower db can have a more offensive and irritating sound than another fan with a higher db... as not all dbs are equal... at least to the human ear.

Try the Gentle Typhoon and a Nexus.

My understanding is that in real life
the Triebwerk TK-121 doesn't sound so good
the Cooler Master SickleFlow doesn't sound so good (the Akasa might be a better bet)
the Noiseblockers are generally pretty good
Be Quiet SilentWings have a good reputation... but do not seem to have any distribution in the US

The reviews will only take you so far, You are going to have to experiment, expecially given the unusual and extreme nature of the problem you are attempting to solve. Try the Gentle Typhoon and a Nexus and tell us what you find.

[ame]

Its unlikely anyone can draw conclusions from comparing information from a few review sites that use different test methodology.

Still, ignoring (for a moment) what I know + what SPCR review knowledge base has to offer, and focusing on just the x-bit roundup - its clear that when running most fans at about 700-800 RPM you get ~31db that traslates to little effect on the room noise floor level and might be considered a quiet range for their test. In that range there are a few fans that do better. For standard thickness of 25cm I can see Kama Flow2 has good 17 CFM at ~31db and Nexus also has about 13.5 CFM at that sound level with Noiseblocker testing similar to Nexus (on paper).

If I now add what I already know from experience about these fans and what SPCR review tested in a proper chamber its safe to say Kama Flow and Nexux (though I tent to prefer the black version)fans are among the best for quietly cooling heat sinks along side Slipstream (not slim) and Noctua NF-P12. To top it off these 4 fans I mentioned also sound subjectively very smooth and practically inaudible at 800 or lower even in very quiet rooms.

[ces]

f I now add what I already know from experience about these fans and what SPCR review tested in a proper chamber its safe to say Kama Flow and Nexux (though I tent to prefer the black version)fans are among the best for quietly cooling heat sinks along side Slipstream (not slim) and Noctua NF-P12. To top it off these 4 fans I mentioned also sound subjectively very smooth and practically inaudible at 800 or lower even in very quiet rooms.

The additional criteria of handling the high static pressure from a 1 inch thick filter makes it worth testing the Gentle Typhoon, which by all appearances seems like it should do well in such circumstances. This assumption would tend to be supported by the fact that these fans are particularly popular among the water cooler crowd.

[ces]

For standard thickness of 25cm I can see Kama Flow2 has good 17 CFM at ~31db and Nexus also has about 13.5 CFM at that sound level with Noiseblocker testing similar to Nexus (on paper).

Don't forget those are free air CFMs... which would make me tend to trust the Noiseblocker and the Nexus to do better under high static pressure... just because they are good fans and they are doing relatively poorer in free air.

[jonjan]

I found a few SPCR measurements, and found the sound sample page for the Nexus... and that makes me realize more than yes, the only way to really know what a fan sounds like in person, is to hear it in person.
If I'm able I might test a few fans, though I'd rather try to find a keeper and buy only that one. My guess is that with the help of this thread and the net the first fan will suit my fancy amicably.

I checked out the Noctua P12 and the Nexus. Both have some good numbers agreed by Noctua, xbitslab and spcr.


As a general benchmark, I'm using the sound level that is inaudible in a quiet room in a quiet house in a quiet neighborhood. Most sources online mark this at 20 dB.
xbitlabs seems to have a distinction for this.. "acoustic quiet", for their measurement of 33 dB. For SPCR, 19 dB seems to be a match for that same level, measured from 1m away.
(How SPCR defined the 19dB level (when testing Noctua NF-P12):
"Fan @ 7V: The fan was almost silent and very smooth. Close-up it was still audible, and the same airplane-type noise persisted though to a lesser degree.")

That dB level is the first criteria. Then the static pressure is next. Without really knowing what they mean, or what I'd like, I just look for the higher numbers.


Here is the condensed data & info...



Noctua NF-P12 (1300rpm)(12v)
L.N.A. ----- 1100rpm --- 9v --- >19 dB spcr
U.L.N.A. --- 900rpm --- 7v --- 1.2 mmH2O
-----Noctua -- 13 dB ---- 37 CFM --- (900rpm)
-----xbitlabs -- 33 db ---- 21 CFM --- (900rpm) (no voltage given)
840rpm
-----spcr -----16 dB ----------------- (840rpm) = .6 mmH2O


http://www.xbitlabs.com/articles/cooler ... up_28.html xbitlabs
http://www.silentpcreview.com/article850-page5.html SPCR
http://www.noctua.at/main.php?show=prod ... g=en&set=1 Noctua
http://www.overclock3d.net/reviews/case ... formance/3 home-measured the P12 (12v) at between 1-2 mmH2O


Nexus 120mm Real Silent (D12SL-12) - 36 CFM - 1.2 mmH2O --- (1080 rpm)(12v)
-----11.5v -- xbitlabs --33 dB --- 16.4 CFM - 1 mmH2O --- (910rpm in tube)
-----9v ----- spcr ----- 19 dB --- 23 CFM --- .7 mm H2O -- (850rpm)
-----9v ----- xbitlabs -- 31 dB ---------------- .6 mmH2O ------- (750rpm in tube, 810 w/o)

xbitlabs in-tube and out of tube measurements give the apearance of discrepencies, but if SPCR is using out of tube, then the dB levels and the volt and rpm levels are in line.

http://www.xbitlabs.com/articles/cooler ... html#sect0 xbitlabs
http://www.silentpcreview.com/article69 ... html#nexus spcr sound samples
http://www.nexustek.nl/NXS-nexus120mmre ... asefan.htm nexus

Pluses of the Noctua P12 are the use of a single simple included resistor to reach the desired dB level. However, taking it lower than 7v would require a fan controller.
It's $25+

Pluses of the Nexus are an apparent superiority of sound quality (from scanning comments all over), and that the desired dB is around 9-11.5v, so there's room to tune down more. Also it's $5.
Downside of the Nexus is that it'd require a fan controller (though the Zalman Fanmate 2 is about $5).
Another possible downside is the sleeve bearing: it couldn't be mounted horizontally, and I wonder how much heat the sleeve oil could handle?

[ame]

I just did a quick swap between the Noctua and Nexus to see how they fair.

Both ran at a fixed ~690 using BIOS to control the speed.

CPU cooler is Termalright Ultra Extreme rev 1 that has fairly tight fin spacing.

notice the near identical performance as reflected in the minimum and maximum temps for both these fans. at this speed at 30 cm open case I could not hear either of them.



Noctua @ 690 RPM




Nexus @ 690 RPM




Both are tied for being better than other fans I tried.

[jonjan]

They definitely looks very close. I kept reading that on SPCR pages too. Mostly saying that the Nexus has much quieter at high speeds, and only just quieter at lower speeds.

Due to the Nexus having sleeve bearings I can't go horizontal or use it with heat... so I might be going with the Noctua P12.
Not sure yet... but this thread has certainly helped me narrow my choice down to one of these two!

thanks for all the help