Fans - How To Reduce Noise

Giasas, George. The sound of the air flow shouldn't be too loud. Most likely the fans produce the noise. I have tried several times to lubricate the bearings of the fans. That implies to remove them, detach the propeller, clean, lubricate, and put all back. For a (short) while the noise is much reduced. It comes back in time. I think that you should replace them with new ones with ball bearings. Have fun (and an ouzo for me).

Hi Indel.

Yeap, I'd prefer fun instead of... fan.

. . . . . . . . . .

Cheers.

If you read the specs, you'll see the noise level of one fan is 58.5 dB(A). Four fans would thus be 64.5 dB(A), the approximate sound level of two people talking at once.

Have you measured the sound level? Do you know the frequency spectrum of the noise?

The simple answer is to buy quieter (this usually means larger, slower) fans. Can you do this?

Can you duct the exhaust of the warm air to a muffled area?

Hi TVP45. You said: "The simple answer is to buy quieter (this usually means larger, slower) fans. Can you do this?"... I'm afraid not. The height of the box is standard 1U. The fans that I use are slightly shorter than this. So I can't use larger and slower fans (in order to achieve the same air flow).

In that case, spec a lower noise fan (usually more $), eliminate as much flow resistance as you can (use flow directors, even flow straighteners in a pinch) to cut down on turbulence, and duct the exhaust. Insulate the fan box if you can. Examine the noise spectrum and look for transmission routes appropriate to that frequency.

Can you duct to some sort of exhaust plenum? This can be as primative as a couple of folders held in place by paper clips

Maybe a temerature control could help reduce the noise. Usually fans are overdimentionned, and could keep your device cool even when running on a lower speed. Lower speed = lower noise !

I used a very simple method. Only two components can do the job. An NTC resistor's resistance gets lower when the temperature rises. If you put an NTC in series with the fan, the fan will run faster if the temperture rises, and of course the fan will run slower if the temperature drops. To be sure that the fan starts running at startup, I put a capacitor in parrallel over the NTC.

I can't give any correct values for these components, all depends on the situation, specs of the fan etc. You'll have to experiment a little ... but usually the NTC has a low value ( between 33 and 100 Ω ) and the capacitor is somewhere between 10 and 100 µF / 16V. (be sure to put the (+) of the capacitor at the + side of the supply) Mount these components so that they are in the airflow of the fan.

I would make this circuit 4 times (one for every fan) because 4 fans together probably are going to pull to much current and warm up the NTC.

I'll post a diagram if I have a lil more "spare time" ...

Or this gives better temperature control and you can use a low power NTC.

But, again you may need to tweak the design to make sure the fan starts.

Hi 44mEurope. The system has already a (more sophisticated) control circuit for the fans (temp. sensor → processor → p.s. → fans). Usually the fans operate at a voltage lower than 12V (i.e. 8V or 10V) and the sound is not so loud. But sometimes (because of the temp. rise of the system) the control circuit forces them to operate at 12V. Then the sound is loud.

Your idea with the NTC resistor (and the capacitor) is really good though (if a simple and rough control is needed). I'll have it in mind for other similar applications.

Use eight of the low power (LN: low noise?) fans instead of 4 of the high power ones (VHN very high noise?)

4 off 58.5 dB-A gives 64.5 dB-A as TVP45 says

8 off 39 dB-A gives 48 dB-A

Air flow would be decreased by more than the reduced spec. suggests because the max voltage of the top fan is lower, but it would be better distributed.

You'd also be using less power (4/6.5).

Hi Randal. Unfortunately I can't use eight low power-low airflow fans due to the lack of space. Moreover, I can't use only four such low power fans because (probably) they will be insufficient in a high-temp (heavy load) situation. The thermal measurements show that these fans do nice job, so I can't take the risk to replace them with low power fans.

Thanks anyway... ...

Probably gonna be difficult to find a fan that small (38 mm) that gives such a big airflow ... I searches a few catalogs, but I couldn't find any better...

The sollution with the 8 fans is probably the best.

Or if you could get a 2U unit, you could use bigger fans who give the same airflow with less dB's ...

Anyhow ... GOOD LUCK !