I need VERY slow (Noctua?) fan for heatsink

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

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
axee
Posts: 170
Joined: Sat Nov 15, 2008 1:01 pm
Location: Slovenia

I need VERY slow (Noctua?) fan for heatsink

Post by axee » Sat Nov 09, 2013 11:15 am

So I bought S12A PWM and I'm impressed. I can run it at higher speeds than any other fan (500RPM) without hearing it. My only concern are hot days (it can get pretty hot in my studio).
So I'm planning on adding another fan on CPU cooler (HR-02 Macho, passive i7), just to be able to run 2 fans at very low RPM instead of one fan spinning at higher RPM (when on high load).

My predictions:
Adding 1 very slow (300-350RPM) fan on HR-02 won't make any audible difference on idle, but it will lower CPU temperatures on full load. That will prevent exhaust fan spinning too fast - anything over 500 RPM I can hear.


NF-S12A seems like good option, since I have non restrictive heatsink and I have great experience with
NF-F12 seems like good static pressure option, but is it effective or/and necessary at so low speeds on HR-02?
NF-A15 seems like good option, just because it will probably cool other components around CPU socket (my mobo must be running quite hot with so little airflow)

What would you choose, feel free to share any thoughts.

Thank you!
Last edited by axee on Tue Nov 26, 2013 2:24 pm, edited 1 time in total.

Abula
Posts: 3662
Joined: Tue Nov 02, 2004 12:22 pm
Location: Guatemala

Re: [call me a FANboy] I need VERY slow Noctua fan for heats

Post by Abula » Sat Nov 09, 2013 2:40 pm

I would go with Noctua NF-A15 PWM, out of
1) Its design for heatsinks, similar model than what comes with the NH-U14S. Per Noctua website, "The NF-A15 is tailor-made for use on today’s high-end CPU coolers"
2) Per SPCR First 140 mm Fan Roundup: Noctua NF-A15 PWM, "The A15 is strikingly similar to Thermalright's TR-TY140", so it should match very well, the HR02 included fan clips work, check below for pics.
3) It can be drop extremely low, even lower than your NF-A12A PWM, check the below pictures, in both cases is ran on CPU_FAN header (true pwm header).

Image

Image

Worth mentioning is that Asus motherboards only have 1 true PWM fan header, if you want to control them independently then you need 2 pwm fan header (CPU_OPT doesnt count as its just a mirror, it will get the same PWM signal as the CPU_FAN), so assuming that you want the fan to be off under idle conditions, i would chose the one that start lower on the CPU_FAN header, in this case the NF-S12A PWM = 15% minimum and on the CPU_OPT fan header the one that starts higher, NF-A15 PWM = 17% minimum, either way its very very close so it might still run, depending on how exact is the PWM design of the mobo and the fan, but i can tell you its inaudible at those rpms, it will barely move any air at all.

As a warning if you use a Noctua PWM fan on a 3pin voltage controlled pin (aka CHA_FAN 1,2,3), it might be unstable, first it might not drop as low and second it might drop to voltages into where its not design to operate. I tested both NF-A14 PWM and NF-S12A on 3pin headers on my Asus, in both cases it didnt drop as low and they stoped in time. If you plan on going with CHA_FAN headers then go with 3pin fans, and if you go with CPU_FAN then go with 4pin pwm fans.

Image

Image

As final warning, Noctua doesn't recommend the NF-S12A on heatsinks or heavily restricted spots, where the design of the blades will struggle moving air, given that HR02 has its fins with more space, i dont think its enough to warrant using it over the NF-A15 PWM thinking that both drop so low and that A15 is design for heatsinks, and will match the HR02 so well into its size matching the recommended fans on the HR02 macho (TY141/147). In case you want to look into how the HR02 with A15 looks ill leave you a pic of my build (its no longer there not out of performance but wanted to try the TY150).

Image

Image

Image

axee
Posts: 170
Joined: Sat Nov 15, 2008 1:01 pm
Location: Slovenia

Re: [call me a FANboy] I need VERY slow Noctua fan for heats

Post by axee » Sat Nov 09, 2013 7:29 pm

Wow, thank you for extensive report!

A15 seems like obvious choice, the only problem are Asus pwm headers. I will probably want to run both from cpu fan header, others simply cannot go that low..

It's really a shame asus does this..

I'm just wondering how the can I run scythe kama pwm at 280 rpm on chassis fan header (40% duty cycle). I guess kama pwm has a very strange fan speed curve..

Abula
Posts: 3662
Joined: Tue Nov 02, 2004 12:22 pm
Location: Guatemala

Re: [call me a FANboy] I need VERY slow Noctua fan for heats

Post by Abula » Sat Nov 09, 2013 8:11 pm

axee wrote:Wow, thank you for extensive report!
Np, glad to help.
axee wrote:A15 seems like obvious choice, the only problem are Asus pwm headers. I will probably want to run both from cpu fan header, others simply cannot go that low.. It's really a shame asus does this..
Asus fault is only in their marketing saying it has all PWM Fan headers, when in reality is only one. But these CHA_FANs (the fake 4pin that are voltage controlled) can go as low as the CPU_FAN (PWM control), its a matter of choosing fans that can drop as low on voltage control, again is all about understanding how the motherboard was design and its limitations to end up with what you want.

By design the NF-A15 PWM and NF-S12A can drop below 200rpm on PWM signal, but around 400 or so on voltage control, but this a limitation of the fan design by noctua, other fans are different. For example i were to be still on Asus with FanXpert2, i would probably chose different case fans, that can drop super low on 3pin voltage control (aka Asus Fake 4pin headers). I'm going to explain a little further, check the following

Page 1 from SPCR Second 140 mm Fan Roundup: Antec, bequiet!, Corsair, Scythe,

Image

Another, page 5 from SPCR Fan Roundup #6: Scythe, Noiseblocker, Antec, Nexus, Thermalright

Image

Now what do i want to show on those, that Lawrance is using FanXpert2 and on a CHA_FAN header, the same that you assume don't go that low, but these fans are reaching 200rpm on pure voltage control, its all about how the fan is design and what header are you using it.

When using Asus motherboards and wishing to get a super low rpms, its important to chose the proper fans for each header, and for this is important to understand the limitations and design that asus implemented with their motherboards and fanxpert.
axee wrote:I'm just wondering how the can I run scythe kama pwm at 280 rpm on chassis fan header (40% duty cycle). I guess kama pwm has a very strange fan speed curve..
Read the reply i posted on the other thread you asked for an explanation.

axee
Posts: 170
Joined: Sat Nov 15, 2008 1:01 pm
Location: Slovenia

Re: [call me a FANboy] I need VERY slow Noctua fan for heats

Post by axee » Sun Nov 10, 2013 6:24 am

Well, I'm using CPU fan header for S12A, minimum speed I can set is 21%, which corresponds with 360RPM. If I'm adding heatsink fan, I will use PWM splitter and run all fans from CPU header. So basicly I need fan with similar duty cycle/RPM response curve.

I'm not sure who can we blame CHA fan headers limitation, did you actually try them with voltmeter? I think I'll try that, I'm still not convinced that CHA fans are just voltage controlled 4pin headers.. are you really sure that they are really not just PWM headers.

NF-S12A test
Image

Scythe Kama PWM
Image

I guess you're right, CHA 4pin headers are in fact NOT PWM headers.

Abula
Posts: 3662
Joined: Tue Nov 02, 2004 12:22 pm
Location: Guatemala

Re: [call me a FANboy] I need VERY slow Noctua fan for heats

Post by Abula » Sun Nov 10, 2013 7:40 am

axee wrote:Well, I'm using CPU fan header for S12A, minimum speed I can set is 21%, which corresponds with 360RPM. If I'm adding heatsink fan, I will use PWM splitter and run all fans from CPU header. So basically I need fan with similar duty cycle/RPM response curve.
If your FanXpert2 fan control is similar to mine (it should, mine is AI SUITE III but FanXpert2), then i think NF-A15 PWM is a good match, the control range is almost the same 15% to 17% minimums respectively. Still i have my doubts being totally the same since you are on 360rpms while im at 250-275 rpms on idle, regardless, the fan curve is very similar on both fans so even if there is a bump in minimum speeds should be both the same in terms of PWM SIGNAL.

Image

Personally i would run the NF-A15 PWM as the rpm signal connector on the splitter, as it has a minimum of 17%, this would bump the NF-S12A from 15% to 17%, but the difference should be negligible, but this will ensure that both fans spin always. But feel free to test both fans on your hardware/software as not all samples of fans are 100% identical, so its best to test once you have it and you decide which of the two runs on the RPM signal connector on the splitter (one has 4 cables the other has 3).

Now on a lot of ASUS motherboards, you have a CPU_FAN and CPU_OPT, the last is just a mirror from the first, its not controllable, but it has an advantage over running it with a splitter, that CPU_OPT does read the RPMS, so you will have display on you AI SUITE of both fans RPMS, i cant say for sure as i don't know which motherboards you own, but check if you have the CPU_OPT header, and if you do, use that instead of the splitter.
axee wrote:I'm not sure who can we blame CHA fan headers limitation, did you actually try them with voltmeter? I think I'll try that, I'm still not convinced that CHA fans are just voltage controlled 4pin headers.. are you really sure that they are really not just PWM headers.
All i can tell you is that i have done a lot of testing on my Asus Maximus VI Gene, and read a lot of other people tests, but i havent used other asus motherboards with fanxpert2 to know for sure if this is how it is on all, its just my personal assumption. I also didnt bother testing all CHA_FANs, maybe one isnt voltage controlled but a true PWM header. But if you do run voltmeter test let me know, im interested on this, as ASUS market their motherboards as full PWM headers, just not my experience thats true, again i can be wrong, so please let me know if you do find it. Another thing you can do if you feel like it, is enter the bios and go into fan control and change the Q-FAN from AUTO to ADVANCED, this will make the CPU_FAN switch from being a true pwm header to voltage control header, run fanXpert2 on it like that see if the curve is the same as what you are getting on the CHA_FAN headers. If you find anything post it, im interested on this, as this was one of the reasons im not buying ASUS motherboards.
axee wrote:Scythe Kama PWM
Image
While i love scythe fans, specially slipstreams and kamaflex pwms, these were the first pwm fans that were done right, even before noctua, for someone looking for a very low dropping quiet fan. But in terms of PWM they are a little (not much) wierd, in some cases their PWM fans drop lower on voltage control then on PWM signal (this is not the case on Noctuas). Lawrance from SPCR has recently started dual testing the fans on the reviews, so you can see some of this cases there also.

From Scythe Ashura CPU Cooler

Image Image

From Scythe GlideStream, Slip Stream XT, and Grand Flex Fans page 3 n 4

Image Image

Image Image

As you can see from the three examples (there are more), some scythe "PWM" fans drop lower on voltage control fan than on true PWM control, but this is how these fans are. There are other Scythe PWM fans that are not like that, they are like noctua that drop much lower on PWM than on voltage control, the only way to know is to check reviews like SPCR or do your own testing. This is the main reason i decided to keep my asus motherboard for pure fan testing, as not all fans are the same.
axee wrote:I guess you're right, CHA 4pin headers are in fact NOT PWM headers.
Don't accept what im saying is right, test it and draw your own conclusion. But i do recommend you to read ASUS Z87-Deluxe fake 4-pin headers & other fan control info, its a good discussion on the subject and the experience of some users.

axee
Posts: 170
Joined: Sat Nov 15, 2008 1:01 pm
Location: Slovenia

Re: [call me a FANboy] I need VERY slow Noctua fan for heats

Post by axee » Mon Nov 11, 2013 5:29 am

Advice from Noctua support:
Since you've already bought a NF-S12A-PWM fan, I would suggest that you try it on the CPU cooler as it is likely that it will provide very good results in this application as well. I wouldn't recommend to fit a NF-F12-pwm though, because this fan is optimized for heatsinks with tight fin spacing ans usually it is rather noisy when used in low-impedance applications. The NF-A15-pwm is somewhere in between these two fans, but you would have to check whether the additional air flow is required after all.
It's nice to see honest opinion, that both F12 and A15 will probably be too loud for my situation. Enough spending money on pricey fans, Slipstream will do for now. It has some kind of ticking noise, like many PWM fans, but If it starts rattling like my previous Slipstream, I will replace it with another S12A PWM.

Post Reply