Noctua = Overhyped, noisy little buggers

[Bemaitea2]

Just received 2 Noctua's to put in my P180

Unfortunately to get the same CFM of my current Tri Cools I had to plug in a Fanmate becuase the UNLA feature only halves the RPM (5v equiv)

But to my suprise they weren't whisper quiet!

As a matter of fact, to push the same amount of air as my Tri Cool, it made even MORE noise.

Needless to say, they are going back and I'm sticking to the Tri Cool on Low.

I'll tell you, at lower RPM's these fans are awesome with virtually no noise. But the reason they make so little noise is because there is no air being moved, just kinda swirled around.

If you've got a P180 and a Zalman9500, do not use these as reliable case fans, your temps will go up, not to mention they aren't as quiet as many people claim.

Noctua
RPM........................................................... CPU
600RPM=whisper quiet..................................40C
800RPM=smooth whoosh, but not annoying.....39C
1000RPM=loud for SPCR standards.................37C
1200RPM=not even considerable....................36C

Normal CPU temp with Tri Cool on low: 36C

So as you can see the NOctua doesn't do anything for CPU heat, which in turn increase the thermal adjusting CPU fan, effectively increasing overall noise.

Not to mention since it pushes such little air overall case heat increases.

Conclusion: If you can live with little air movement in your case then this is your only choice as it made no noise at all (ie lightly equipped rigs).

If you at least play 1 3d game that will utilize your CPU and GPU, then look elsewhere as this is just an expensively loud fan.

Sorely dissapointed since SPCR gave it such awesome reviews

[qviri]

Is that temperature on idle or load? What is the CPU? What is the heatsink?

[Bemaitea2]

Temps are idle, I didn't want to push my system to load with the noctua blowing.

Its a AMD4400+ @ 2.5GHz, Zalman CNPS9500.

I'm not really sure whats going on, but at the same RPM and more importantly noise level as the Tri Cool, it blew LESS air.

I think if you took it out of its frame it might be a better contender, as a fanswap candidate, but other than that its money that is better spent elsewhere.

[josephclemente]

I've been very happy with my Noctua fans. They are all 800 RPM editions. I did try a 1200 RPM version which came with my Noctua heatsink, and I thought it was pretty noisy at 12 volts. Instead of using any voltage adapter, I just put an 800 RPM Noctua fan on the heatsink instead.

[Bemaitea2]

I will agree with you, that yes at 800RPM the fan is not as noisy as at 1200RPM.

But the noise/CFM ratio from my perspective seems to be high.

Plus, the price tag of $20 I'd expect something more appealing.

I guess this is just a case of high expectations.

[Devonavar]

You're not the first person to make a report like this. This issue has been on my mind for a while; so much so that I am currently re-evaluating the test procedures we use for airflow.

However...

At this point, I don't think it's fair to say that the Noctua provides lower airflow ... all it's safe to say yet is that it behaves differently from conventional fans.

Thus far, we have tinkered with our test procedure improve repeatability and consistency across a wide range of fan designs, and we are currently trying to ensure that the procedure will correspond to thermal results.

Our thermal bench so far is a Ninja on our heatsink test bed, where the Noctua beat the Tricool by ~2°C, both running at a constant 800 RPM. For the record, that's about 70 RPM slower than the "L" setting for the Tricool.

At this point, all I can glean from these reports is that the Noctua behaves oddly; sometimes it seems better, sometimes not.

[Bemaitea2]

Devonavar, I have a suggestion for testing realistic airflow.

What I did to test the Noctua vs. the TriCool was a bit different.

What I did was tape a piece of paper to the front of my P180 with about 5 inches or so hanging over the edge like this diagram:




Then under the paper, on the floor, I layed a ruler down with the 0" part at approximately the edge of the paper.

Next I placed a small stack of books in front of the part of paper hanign over the edge. I placed, first, the TriCool on the books so that the bottom of the fan was at the bottom of the hanging paper at approximately the middle of the hang over (measured with ruler). I made 2 marks on the book where the fan was.

Next I turned on the computer and let the fan reach its speed, then from a top view, looking down on the floor, I measured how many inches the paper bent back with the air flow, using the ruler I had placed on the floor already.

I got about ~1 inch of bending.

I then went to medium, and then to high and noted the bending (at high the bending is basically perpindicular to the face plate of the p180 becuase of so much air.


Next I did the same using a Noctua. But since it doesn't have the similar L,M,H as the Tricool, I used a FAnmate to control the speed. So then I just simply adjusted teh speed of the fan until I got about ~1 inch of paper bend, just as the Tri-Cool had on low.

So now that the Noctua was calibrated to move as much air as the Tri Cool, I plugged both in at the same time and had a listen. I didnt have a dB meter, but from my hearing alone, I could tell that my 1 year old TriCool was making less noise than my Noctua while both were blowing the same amount of air.

That was my method at least and from that, and a few more tests with different setups inside a case, I came to the conclusion that even though at low RPM (low airflow) the Noctua is silent, it is no where near perfect.

[[F]bernZ]

But I do believe the Noctua has a good sonic signature. I've had my fair share of Tricools and compared to the Noctua (I have the 1200RPM version at 800rpm), the Tricools clatter a hell lot more compared to the Noctua.

[Bemaitea2]

There is no question that there is less noise on the Noctua, but as a respectable fan it needs to push as much as air as it is quiet.

I simply found the noise/CFM ratio very dissapointing. IT seems as though they've sacrificed way too much airpressure and air flow for a mediocore silencing method.

[Felger Carbon]

I simply found the noise/CFM ratio very dissapointing.

Some folks love them, some hate them. My guess is that they work well in cases where there's a low-resistance air path, and stumble badly if there's a high-resistance path. Noctua seems to have deliberately optimized for high airflow when there's a low resistance, by their high blade pitch.

[Bemaitea2]

That seems understandable, but it doesn't explain why when I had the fans outside the case it made more noise than the Tri Cool.

The only reason I can think of is since the Noctua reduces air noise by reducing air movement and RPM, in order to get to a reasonable level of air movement it needs to spin much faster than the Tri Cool, which makes sense.

But at this higher RPM I found the noise character of the Tri Cool more tolerable, especially inside the case.

[Calgarian]

I'm one that loves them.

Fan, CPU cooler with fan and chipset cooler.

Very pleased.

[spookmineer]

As stated a few times in these forums (hard to keep track, I agree), Noctua fans aren't meant as heatsink fans because when they face any backpressure, the performance drops like a brick.

Their best use is as a rear case fan, sucking air out instead of pushing air in.

What makes it more hard to compare these fans to others is: Noctua fans have a different blade design. Like you, most say it swirls air around outward, rather then in a straight path. So, because of the pitch of the blades, a considerable amount of air isn't pushed out directly along the axis of the fan, but on a tangent.
If this is the case, measuring how much this fan pushes a paper backwards isn't doing it justice, it pushes more air, just not in the "right" direction (the direction of measurement).
This would explain why these fans do well as rear case fans (just getting air out) and badly when used in combination with a heatsink (where you need a directed airflow).

[I don't have a Noctua fan, condensed the comments of other people in this forum]

[Stefx]

There is no question that there is less noise on the Noctua, but as a respectable fan it needs to push as much as air as it is quiet.

I simply found the noise/CFM ratio very dissapointing. IT seems as though they've sacrificed way too much airpressure and air flow for a mediocore silencing method.

I have 4 x Noctua NF-S12-1200 in my P180 (2 intake, 2 exhaust), and I have a feeling (not scientific) that those fans are silent but push little airflow when faced to a resistance.

The test conditions that measure airflow should reflect the typical resistance faced when installed in a typical PC case

I'm ok for now (not overclocking) but when I overclock I may have to change those fans for quiet fans but better air-movers

[Bemaitea2]

I may have not been clear, I was using the Noctua's as Case Fans, not as CPU fans.

I tested the Noctua out first without mounting to see how much noise/CFM the thing had. After inital testing I was thrown back about how little air is moved and didn't want to risk using it as a HSF in fear of overheating.

But even as case fans there are inadequete. Temps go up all around, especially around the CPU in a P180, and to get good cooling they make a distinct high frequency noise.

As far as my method for testing air flow, it may not be perfect and requires further tweaking, but if my calculations are right, the Noctua should have had the advantage, according to you're theory of it pushing air in a non cylindrical shape.



As you can see the Noctua, according to you're theory, should have actually pushed MORE air than the Tri-cool at the same RPM.

Not perfect math, but it should give you an idea.

There should be no perpindicular air movment on the Noctua becuase it whould be infeasible in such a fan structure.

[Devonavar]

It looks to me like the bent-paper method of measurement is actually measuring pressure, not airflow. I think the point that the Noctua does poorly in situations that require high pressure is well taken, and I think that's what your test seems to demonstrate.

On the other hand, the results I got with the Ninja seem to suggest that, at least with the Ninja, the extra airflow can have an effect. I'm not quite sure how to determine which situations these are though...

[spookmineer]

As you can see the Noctua, according to you're theory, should have actually pushed MORE air than the Tri-cool at the same RPM.

Not perfect math, but it should give you an idea.

There should be no perpindicular air movment on the Noctua becuase it whould be infeasible in such a fan structure.

A = pi*r^2*h
This is the amount of airflow (assumed per time interval) if the air leaves the fan in a unidirectional way. It is not an area (A) but a volume (V).

The area the force is put upon is relevant, but so is the force itself (with increasing area, the force put upon it will decrease).

Any airflow which has a component flush with the paper surface, will have loss in the direction it is measured. The speed vector can be disected to a component directed perpendicular to it, and flush with it.



The speed flush with the paper will not be measured by movement of the paper but it still gets rid of the air.
Ofcourse, this happens with all fans (turbulence) but because of the pitch of the blades, more so with Noctua fans.

The only way to measure the fan performance is to just measure the temperature results when using this fan. Measuring the amount of airflow is very hard (and in the end, not as important as the temperature results).

[Bemaitea2]

I absoluetly agree with you, temps and silence speak louder than airflow and pressure.

Sorry I had meant Volume, not area in my above equations.

But yes I can see where the spread of air over a greater area yields less air pressure, but same amount of air. My testing "technique" still needs to get worked out.

But I can safely say at the same Air Pressure as my Tri cool on Low, the Noctua makes more noise while not cooling the system down at all.

As a low noise case fan at >800RPM, it is silent and should be considered by a more passive user who simply wants a silent rig, but by no means would it be sufficient for a heftier rig.

I had 1 Noctua replacing my exhaust fan in my P180, the top port hsa been covered to reduce noise.

From that position the Tri Cool was slightly more noisy (mostly due to airflow) on Low than the Noctua at half speed (UNLA and 1/2 turn on FanMate), but that made up by the obviously greater airflow.

My adjusted conclusion: From my experience, it seems that the noise/air pressure ratio of the Noctua to the is greater than my Tri Cool.

Of course this is assuming that spookmineer is right in his conical theory. I will pull out the Noctua again and do more testing and let you guys know what else I learn.

As of now I would not reccomend anyone needing a silent, efficient case fan (don't even consider it as an HSF) to purchase the Noctua at the $20 price tag. There are cheaper alternatives that are more quiet.

[Fungi]

Which is more important (performance/silence) depends on your use, and that's why the double graphs of Madshrimps are so useful.
You can compare with noise OR airflow depending on needs and find the one that suits your needs the best.

[J. Sparrow]

As a matter of fact, to push the same amount of air as my Tri Cool, it made even MORE noise.

Have you considered the possibility that your Noctua have been damaged during transport?

I don't have any Noctua, but I do have a TriCool and my feeling is that it moves little air compared to the RPM. The Low speed setting is mostly useless to me as I always use the PC about full load.

Regarding the point Noctuas not being fitting fans for heatsink use, just two things I'd like to point out:

1) The guys at Noctua sell their own S1200 with the U12F heatsink, so they're pretty confident about the fan performance.

2) In a recently posted test, the Noctua showed to be a pretty decent fan for heatsink use. The performance only dropped decidedly when used on a radiator (much more restrictive than a common heatsink)

However, this whole thread is a proof that I was right some time ago when I wrote about the existence of a "Noctua dilemma".

[Bemaitea2]

I had considered a fan problem, but both fans at teh same time?

Not to mention it was sent via UPS.

I can't think of a viable excuse why tehy would act like that.

As far as a HSF, the 1200RPM model will psuh enough air for acceptable cooling, but at a very noisy cost.

My current Zalman9500 with a Nexus 92mm mod pushes more air than the Noctua and sounds nicer too.

I really would like to belive that my experience with both of my Noctua's were flukes, but I'm not hopeful.

[Felger Carbon]

That seems understandable, but it doesn't explain why when I had the fans outside the case it made more noise than the Tri Cool.

Especially if you confirmed the two were spinning at the same RPM.

It's your money and if you prefer redheads to brunettes, er, Tricools to Noctuas you are absolutely entitled to your choice. But please don't try to broadcast, repeatedly, that brunettes are a bunch of crap. Others are entitled to their opinions too, you know!

[Bemaitea2]

Yes Felger, its true that everyone is entitled to their respective opinions and you may have mistaken me, I am not denouncing anyone elses opinion.

I was ready for the Noctua to be the perfect fan for me. It wasn't. Others may find that they love it.

For the record I do not prefer Tri Cools, they are alright but nowhere near perfect. I had such a high expectation for the Noctua's that I was pretty let down.

I'm realy not sure what I should be looking for now as far as a silent case fan goes. I guess Nexus or Scythe.

[Felger Carbon]

I was ready for the Noctua to be the perfect fan for me. It wasn't. Others may find that they love it.

Others have found that they love it, and have so reported here in the SPCR forums. Still others agree with you. I don't think any of these guys are idiots - they hang out with you and me, so they must be really smart as well so there must be an explanation for their different reactions to the same device.

Full disclosure: I do not own either Noctua or Tricool fans, and have never heard either of them.

There has to be a reason for the variance that the SPCR crowd is collectively reporting! My guess, as I've said, is that the Noctuas work well in low-resistance air paths but have poor noise performance in high-resistance air paths.

[J. Sparrow]

I guess Nexus or Scythe.

I was about to buy Noctua but then went for Scythe. Why? 1) I listened to the sound recordings on SPCR and I didn't like Noctua at all. 2) Scythe costs 25-50% less 3) Noctua marketing material is IMO pretty mean.

I've ordered two more Scythe, too bad I didn't see that Sharkoon review earlier, otherwise I would have been waiting for two golf-balls (they are cheaper, too!)

I won't insist on your fans being defective, but IMO it's a possibility since you received the two together.

[Bemaitea2]

There is a chance of defectiveness, but I don't want to go through the hassle of finding out if in fact both of the fans I received were defective.

Also Felger, you could possibly be right about your theory. Noctua's do in fact use a funny blade design which may react differently than conventional fans.

[s_xero]

Since I did RMA with my first whining fan, I got one which is effectively inaudible. No weird noises.
It is possible that there is some samples variance in the Noctua's.
My second fan had a paler color beige then my defective one.

Also - The Tri-Cool maybe good for a Stock-fan, but for the sake of good fans, it just never belongs there.

I guess the best bet is now the Scythe FSB-series. I've got an 120mm DFS-series from my Ninja as a top-casefan. It's the roughest of the system, by far. On the other hand - the system in total is very quiet by any means.
I guess, except for the low rattle, the DFS1225L12 is in many prospects similar to the Nexus. It's better than the Papst 4412FGL in any way, which is better then the Tri-Cool.

Here it comes: My CPU-probe (in an 4400+(120W) w/ Ducted Ninja) reported 3 Degrees Celsius higher temps with the DFS compared with the Noctua. I guess the ninja has close to no resistance

[jaganath]

I had such a high expectation for the Noctua's that I was pretty let down.

At their price, I would have high expectations too.

My guess, as I've said, is that the Noctuas work well in low-resistance air paths but have poor noise performance in high-resistance air paths.

It would be very interesting to get a hold of the P-Q graphs for the Noctuas. If your guess is correct it should intersect the y-axis very low down.

[thejamppa]

I had few noisy Noctua's. I contacted Noctua directly and they sent me replicament fans. No problems and noise was gone. I had high pitch tone which probably was cause some malfunction or damage in fan. Generally Noctua's have been good and silent fan, even their statistic pressure is bit low.

If you think your Noctua's are noisy, just e-mail Noctua, descripe problem and fan noise. Noctua in my experience takes very seriously their claim of having silent fans and have prooved in my occasion that their customers service is top notch.

It cost me nothing to get those two noisy Noctua's exchanged for new ones. I gained free promo pack and working fans ^^

[Willy Higinbotham]

I was planning to get Noctua fans both for HSF and as case fans in my upcoming build, now after reading this thread, I'm not sure anymore.

It's not easy to decide when there's too many different products to choose.