fan noise vs. voltage

[kallenin]

Hey all,

I was wondering if anyone knew in practice how the noise of fans scales with the voltage (like if you used one of those things to control fan speed). I know that the rotational speed should be directly proportional to the voltage, but I'm not sure how that translates into noise (the whole 20 log |H(w)| thing makes my head go funny). Is it mostly linear, or can there possibly be some funky behavior?

-Kallenin

 

[Madwand]

I'd say that the relationship is more complicated than that, and also that in practice, you need to consider more than just the overall dB figure.

Why?

1. Because some of the noise, particularly at high speeds, changes in frequency as you change voltage / speed. Naturally, but this can be significant because some frequencies will resonate / be more bothersome / travel better than others.

2. Because sometimes there is some specific noise that arises / resides that does not significantly affect the single dB figure, but can be the problem for some users. I practically gave up on the usage of SPL measuring on this finding -- comparing two different fans, A measures better than the B, yet A has a significant whine that bothers me while B's noise is at a low frequency and not objectional. Here the hard data is telling me something very misleading, and the ears do a better job.

3. Because some fans generate different noises at different voltages (some buzz/etc. when down-volted; the apparent level of these noises can change, and at least the relative level changes. I.e. at high speeds, you wouldn't notice these, at lower speeds until near cut-off, these noises are the problem, and don't change much.

4. Because the general goal is not max speed / noise, but rather minimum noise at sufficient speed. Alternatively, it's maximum speed at noise below threshold / noticeable.

So yes, there is some funky behaviour, and the controllers work well in practice, because they allow you to tune (4) for your needs and circumstances.

 

[kallenin]

I suppose you'd have to do some sort of audio waveguide analysis of your case to find out what kind of resonant modes are going on in there. I wonder if they think about it at that level of technicality when they design a case, or if they just test it to see if it works or not... (they're probably more concerned with modeling air flow...)

-Kallenin

 

[Madwand]

I suppose you'd have to do some sort of audio waveguide analysis of your case to find out what kind of resonant modes are going on in there. I wonder if they think about it at that level of technicality when they design a case, or if they just test it to see if it works or not... (they're probably more concerned with modeling air flow...)

-Kallenin

I don't think that the typical $50-$100 case of the month gets very much in the way of design thought, and for evidence I have the day to day experience of how badly some of these things are designed in terms of usability. At most, I think the typical manufacturer might pay some attention to Intel's thermal guidelines and try to pass their tests, and mostly pay attention to price and saleable features of the month, and on occassion copy each others' successes.

My observations above were practical based on personal experience. If you want to go deeper into the theory part, I think you might like following site, together with its links. I find it interesting, but in the end, we often just end up turning the pot until we get a good temp / noise balance.

http://www.cpemma.co.uk/index.html

 

[kallenin]

I think it would be an interesting project to put a few mics around your case (the really tiny micro kind that you can get with some electronics kits), feed it to the line input of your sound card. Combine this input with the thermal monitoring of your motherboard, and you might be able to write some sort of program that would automatically balance noise and temperature.

-Kallenin

 

[nobly]

Also remember that all fans are not made equal. Different manufacturers use different bearings for fans. Some will just use ball bearings, some will use different technologies. I think these parts contribute a significant part of the noise in addition to the fan motion.