How size affect fan noise?

[Zucca]

Quick one, not 100% sure that's why I decided to ask here.

I would like to understand better how fan noise is influenced by the fan size.
Imagine you swap a fan with one double as deep as the original (theoretical example: 35mmx35mmx10mm with 35mmx35mmx20mm), RPM stays the same (again in theory).
What is the expected noise level compared with the original fan?

And what if the fan input voltage stays the same not the RPM?

Many thanks in advance.

[mendip_discovery]

This might help,

https://www.processingmagazine.com/home/article/15587087/understanding-fan-acoustics

[thm_w]

Just changing the depth of the fan on its own, assuming everything else would stay the same, I can't see that having a significant effect on noise.

But everything else would not stay the same. You'd lose a lot of airflow and static pressure. The lost airflow would mean lower noise.
If you held airflow constant then expect more noise (since the fan has to spin faster).

https://www.corsair.com/ca/en/explorer/gamer/accessories/fan-thickness-30mm-vs-25mm-vs-15mm/

I don't know what you mean by "fan input voltage stays the same".

[Postal2]

How size affect fan noise?

You just need to reduce the fan size to zero and get the result. Where it is necessary, for example in the bathroom, the fan should be on a rubber suspension. Although usually called "silent" fans, in fact, these are ordinary fans for the house, unless that house is a metal workshop.

[wraper]

Blade profile means a lot. With thicker fan you can easily get more airflow/static pressure at the same noise level. On other hand with blade count/profile and RPM being about the same, thicker fan will be more noisy but also have higher performance. In general, larger fan means more performance per acceptable noise level. But of course you can make them more noisy than a small fan, something like 120x38mm fan at 10000 RPM is like a jet engine.

[Postal2]

... acceptable noise level. ..

I've already been told I'm not blind enough. Now they're going to tell me I'm not deaf enough.

[Zucca]

Good links:

https://blog.orientalmotor.com/fan-basics-air-flow-static-pressure-impedance

https://www.cuidevices.com/blog/understanding-airflow-fundamentals-for-proper-dc-fan-selection

Considering only the fan (and not the enclosure where it is moving air) it seems the noise level depends only on the fan RPM speed...

To answer my questions:

1) Doubling the depth and keeping the rpm constant the noise does not change (again not considering the enclosure and other variables like blade profile...)

2) By keeping the voltage constant, the rpm in free air does not change (yeah I know, but a pretty good estimation). Doubling the depth double the air flow.
Now putting a double air flow fan in the same enclosure, the rpm/static pressure and air flow will change according to the enclosure air impedance.
So the rpm will be lower than the one in free air.. so the fan will be producing less noise than in free air.
But the air flowing will be much more (in theory double more)... so the air hissing noise through the enclosure will be more (log(double)?).

This is my best educated guess.

[Zucca]

You just need to reduce the fan size to zero and get the result.

Of course if you turn off the light switch the room will be dark.

[Postal2]

Of course if you turn off the light switch the room will be dark.

You are a master of metaphors. You can also close your eyes to this. And if you close the valve, the gas burner will go out. Or the flow of water will stop.

I actually hate fans in computers and will never buy a computer with a fan again.

[IanB]

What kind of fan are you talking about?

If you say 35 mm x 35 mm, that would be a case fan for a computer, for example? Not a room fan, nor a kitchen/bathroom extractor fan?

Broadly speaking, fans with a larger diameter can move the same amount of air more quietly than smaller fans. So a 70 mm x 70 mm fan could be quieter than the 35 mm fan, because it can turn more slowly to achieve the same air flow. (But this implies that the speed should be regulated accordingly.)

Most of the noise from fans comes from the speed of the blades, and the acceleration of the air moving over the blades.

Secondary ways to reduce noise can include blade profile, bearings, mechanical vibration (balance) and mountings. All of these things are amenable to good design, so well-designed fans can be quieter than poorly designed fans.

To some degree, you get what you pay for, which is why "quiet fans" tend to have a price premium.

[Postal2]

A professional artist records an audio book in his home, and in the background you can hear - guess what. Recordings from the 2000s.

[wraper]

Considering only the fan (and not the enclosure where it is moving air) it seems the noise level depends only on the fan RPM speed...

Wrong, and can be easily figured out from fans of the same size and RPM having vastly different noise level.

[Zucca]

Considering only the fan (and not the enclosure where it is moving air) it seems the noise level depends only on the fan RPM speed...

Wrong, and can be easily figured out from fans of the same size and RPM having vastly different noise level.

You are correct wraper, my statement above was of course neglecting the blade profile, bearing type and so on.

How about same size, RPM and air flow. Do I expect again a "vastly different noise level"?
I would tend to disagree.

That said this is also truth:

The acoustic noise produced by a fan will increase by 15 dB when the speed of the fan is doubled.
A 10 dB increase in acoustic noise is typically perceived by human hearing as a doubling of the noise level

Now if we double the depth, the blade profile change as well by definition so it is impossible to make a comparison.

Now consider this:

FAN A: 35mmx35mmx20mm 5000rpm
FAN B: 35mmx35mmx10mm 10000rpm

The airflow should be almost the same in the two fans above, which one is expected to make more noise?

Moreover:

FAN A: 35mmx35mmx20mm 1rpm
FAN B: 35mmx35mmx10mm 2rpm

Same as above? Would a human hear notice the difference? I'm not so sure.

[Zucca]

What kind of fan are you talking about?

Let's say DC brushless fan, since they should be the ones used the most in electronics.

[wraper]

How about same size, RPM and air flow. Do I expect again a "vastly different noise level"?
I would tend to disagree.

Still nope. Noise optimization is a thing. Otherwise everyone would be making cheap silent fans. Also beside noise from moving air, many brushless fans have annoying motor noise.

[Assafl]

Disregarding bearing squeals and other noises what causes fan noise is the same as other "whoosh" noise makers. Which is mainly the speed of the outermost tip of the blade. The other sides of the blade also make noise but much lower in intensity (ignored).

That is why larger fans can move larger amounts of air at lower noise.

The number of blades modulate the "whoosh" causing a low frequency beat. That is how SONAR operators know how many blades a submarine has.

They make huge ceiling fans with many blades (I've seen 6 blades) that run at very low RPM. You can sleep under one, hardly hear it - yet catch a cold.

So at a given CFM the larger fan has lower noise - ALWAYS (again - I know some bearing squeal and some have shot bearing that rattle - that is OT).
And lower RPM is lower noise. But the same RPM for a larger fan means higher CFM AND higher noise (not because of the CFM - but because at the same RPM the velocity of the tip at greater radius is higher).

With fans bigger is always better. Except when it doesn't fit. Then it isn't better.

[Zucca]

How about same size, RPM and air flow. Do I expect again a "vastly different noise level"?
I would tend to disagree.

Still nope. Noise optimization is a thing. Otherwise everyone would be making cheap silent fans. Also beside noise from moving air, many brushless fans have annoying motor noise.

https://www.digikey.com/en/products/filter/dc-brushless-fans-bldc/217?s=N4IgjCBcoCwGxVAYygMwIYBsDOBTANCAPZQDaIMADHAJwCslIAuoQA4AuUIAyuwE4BLAHYBzEAF9CYGhGggUkDDgLEyIAEwBmdTADs65m06Qe-YWMngamxPLRY8hEpHJw9ADkYsQHLr0GiEpYGLiAAJkgABBhCzOJAA

35mmx35mmx10mm 5.5CFM

max-min noise level 3db(A) over 27 different fans.

[wraper]

^AFAIK Delta never really cared about noise, also they're notorious for motor noise. I don't recall a single delta fan I've seen to not have it. Most of Sunon fans I've seen were pretty good though in this regard. Although expensive noise optimized PC fans I've used were more silent with other specs roughly equal.

[Zucca]

So at a given CFM the larger fan has lower noise - ALWAYS (again - I know some bearing squeal and some have shot bearing that rattle - that is OT).
And lower RPM is lower noise. But the same RPM for a larger fan means higher CFM AND higher noise (not because of the CFM - but because at the same RPM the velocity of the tip at greater radius is higher).

With fans bigger is always better. Except when it doesn't fit. Then it isn't better.

This was the quick answer I had already, apparently to quantify and get numbers it is not so easy.

[nctnico]

How about same size, RPM and air flow. Do I expect again a "vastly different noise level"?
I would tend to disagree.

Still nope. Noise optimization is a thing. Otherwise everyone would be making cheap silent fans. Also beside noise from moving air, many brushless fans have annoying motor noise.

https://www.digikey.com/en/products/filter/dc-brushless-fans-bldc/217?s=N4IgjCBcoCwGxVAYygMwIYBsDOBTANCAPZQDaIMADHAJwCslIAuoQA4AuUIAyuwE4BLAHYBzEAF9CYGhGggUkDDgLEyIAEwBmdTADs65m06Qe-YWMngamxPLRY8hEpHJw9ADkYsQHLr0GiEpYGLiAAJkgABBhCzOJAA

35mmx35mmx10mm 5.5CFM

max-min noise level 3db(A) over 27 different fans.

In my experience the noise specification doesn't say much about how a fan sounds when put into an enclosure. The noise depends highly on the pressure a fan needs to build up. See the attached datasheet for an ultra-low noise fan and what parts have been optimised to keep noise low. It is clear that there isn't a single aspect that dictates fan noise; it is the entire design.

[wraper]

^BTW noise level by itself is not the whole story. Frequency distribution is important. Some frequencies sound/are noticeable way more than others.

[Zucca]

^AFAIK Delta never really cared about noise, also they're notorious for motor noise. I don't recall a single delta fan I've seen to not have it.

Two manufactures involved Delta and Wakefield-Wette.

Let's see another example with ebm papst

https://www.digikey.com/en/products/filter/dc-brushless-fans-bldc/217?s=N4IgjCBcoCwGxVAYygMwIYBsDOBTANCAPZQDaIMADHABwDMNIAuoQA4AuUIAyuwE4BLAHYBzEAF9CYAJwRoIFJAw4CxMiDgB2TWABMIQlu2bmbTpB79hYyeGl1ECtFjyESkcmBpw4dByxAOLl5BUQlbfQ8QABMkAAIMIVMNRwForjBKXQQzYKswwnYAT1ZcLnRsFHFxIA

119mmx119mmx32mm 100CFM

27fans at 45db(A), one fan at 30db(A) @ 72$.

so half the noise by paying almost the double.

So wraper is right if we consider those unicorns, generally speaking with the same budget, same dimension, same speed => same noise.

[Zucca]

^BTW noise level by itself is not the whole story. Frequency distribution is important. Some frequencies sound/are noticeable way more than others.

True, also old people does not hear high frequency as well anymore. 

[hans]

nctnico is right, the noise properties of a fan are so dependent on the environment. If the "impedance" changes, that will have a significant effect on the noise. If you ever try to "choke" a fan from air, you'll hear the noise pitch and volume change a lot.

There are also other aspects. RPM, blade count, is one thing. Blade design of course as well. Blades are spinning airfoils with a certain angle of attack and rotation speed. As any airfoil will have, it will have a certain stall speed, turbulence on edges, etc. Consider that the rotational speed near the hub is much lower than at the edge tip of a blade, and there is a lot of art that goes into a fan design. Simply doubling the fan size and halving the RPM for a "lower noise fan" must mean that a blade redesign also must be reconsidered. Perhaps the airfoil near the hub are stalling and creating a bunch of turbulence.

Then "noise" is often expressed in dBA, which is already a slightly subjective weighting of the pressure levels (Pa) that is produced. Nevertheless, if you take 2 fans and spin them together: the noise can completely fine or annoying as hell. I was watching a review/insight on a new Noctua CPU cooler by GamersNexus the other day that went into this. Basically their new CPU tower cooler will have a "faster" and "slower" fan because their modern PWM controller will regulate the fan speed to such a fine degree, that the mixing of frequency components of the blade frequency from both fans (which due to tight regulation may only differ by a few Hz) will make a very annoying "woop woop" sound.

[wraper]

^AFAIK Delta never really cared about noise, also they're notorious for motor noise. I don't recall a single delta fan I've seen to not have it.

Two manufactures involved Delta and Wakefield-Wette.

Let's see another example with ebm papst

https://www.digikey.com/en/products/filter/dc-brushless-fans-bldc/217?s=N4IgjCBcoCwGxVAYygMwIYBsDOBTANCAPZQDaIMADHABwDMNIAuoQA4AuUIAyuwE4BLAHYBzEAF9CYAJwRoIFJAw4CxMiDgB2TWABMIQlu2bmbTpB79hYyeGl1ECtFjyESkcmBpw4dByxAOLl5BUQlbfQ8QABMkAAIMIVMNRwForjBKXQQzYKswwnYAT1ZcLnRsFHFxIA

119mmx119mmx32mm 100CFM

27fans at 45db(A), one fan at 30db(A) @ 72$.

so half the noise by paying almost the double.

So wraper is right if we consider those unicorns, generally speaking with the same budget, same dimension, same speed => same noise.

BTW I don't get why people are so excited about overpriced ebm-papst. IME of seeing them in laboratory equipment working in the same equipment for many years, their reliability was much worse than of much cheaper Sunon.