Let's say we have a device placed in a 5m x 7m room. The device dimensions are 50cm x 50cm x 10cm. The device has an internal fan which is producing acoustic noise. The device also has an internal pump which makes noise when activated through a button press on a digital display.

Is it possible to somehow cancel or suppress the noise that the device emits?

For example by placing a speaker and microphone inside the device where the microphone measures the emitted noise and the speaker outputs a sound wave which suppresses/cancels the emitted noise (like active noise cancellation)?

Or can this only be accomplished by placing one or more external microphones in the room and have those microphones measure the emitted sound so that a cancelling/suppressing sound wave can be emitted from a speaker.

The idea/goal is that people who are in the room while the device is emitting noise will perceive less noise from the device because a cancelling sound wave is being emitted at the same time. Sort of like the same ANC technology you find in cars.

  • $\begingroup$ Are you looking for a natural sounding solution ? Is it a fixed point in the room? Or a moving point cancelation ? Is this a musical application, where acoustical fidelity is upmost important? or just an industrial application, where just a little less motor noise is the only important concern? $\endgroup$
    – Fat32
    Oct 5, 2021 at 22:56
  • $\begingroup$ I don't know what you mean by natural sounding solution. The noise emitting device will be fixed to one location in the room and will not move. The device emits stationary fan noise as well as a pump noise that lasts for some seconds. $\endgroup$
    – james3849
    Oct 6, 2021 at 8:12
  • $\begingroup$ Maybe @richardlyons can chime in... $\endgroup$
    – james3849
    Oct 6, 2021 at 12:16

1 Answer 1


In the domain of noise control, the noise can generally be controlled in three differenct places: the noise source, during propagation, and the reciever.

The third one is relatively cheap and easy to implement, such as the well-known ANC earphone. If you don't want to wear earphones, that's the target of local active noise control. This is very hard and the practical frequency is limited to relatively low frequencies compared to the size of quiet zones.

You want to place a loudspeaker and a microphone inside the device which is the first case, noise control at the sound source.

Is it possible to somehow cancel or suppress the noise that the device emits?

Theoretically, yes. Assume that the noise source is a point source which has omni-directivity and infinity small size, then place another point source at the same location as the noise source but vibrating out of phase and all noise is cancelled by the second point source. However, you can't place two real sources in the same location, you can only place them very close together. This pair of sources is called acoustic dipole and has a directivity of figure 8.

The average radiation sound power of a dipole is given as $$ W \approx \frac{2}{3}\pi \rho_0 c_0 k^4r_0^4l^2 u^2 \tag{1} $$ when $kl\ll 1$, where $\rho_0$ is air density and $c_0$ is sound speed, $k=2\pi f/c$ is the wave number, $r_0$ is the radius of spherical source, $l$ is the distance between two sources, and $u$ is the surface vibration velocity of the spherical sound source.

The radiation sound power of a monopole is given as $$ W_0 \approx 2\pi \rho_0c_0 k^2r_0^4 u^2 \tag{2} $$ when $kr_0\ll 1$.

We can derive that the ratio between the radiation sound power of a dipole and a mono-pole with identical vibration velocity is $$ \frac{W}{W_0} = \frac{k^2l^2}{3} \ll 1 \tag{3} $$ Keep in mind that Eqs. (1) and (2) are derived under low-frequency assumption that both $kr_0$ and $kl$ are greatly less than $1$. Eq. (3) tells us that at low frequencies, dipole radiates much less energy than a monopole. This is because the two sources are out of phase and the sound generated by them are cancelled each other. A loudspeaker unit without sealing or baffling can be assumed as a dipole at low frequencies -- when the paper cone is vibrating, the medium (air) at one side of the membrane is in compression and the other side is in rarefaction, and this phenomenon is called acoustic short circuit.

Back to your question, in practice, it is difficult to realize noise cancelling at the source end. In terms of hardware, you need one or two loudspeakers which have the similar directivity pattern with the noise source, the fan and the pump, so that the sound wave they create can suppress the noise at all direction; of course the loudspeakers should have small sizes to put them close to the noise source, and ability to generate sound loud enough without distortion (should avoid acoustic short circuit); you also need one or more microphones and a capable digital signal processor. In terms of software, a narrow band ANC algorithm is required -- fan noise should be composed of fundamental frequency (related to rpm) and its harmonics but I'm not familiar with the pump noise.

Well, a fan and a pump, is it a PC case? Maybe you should try passive noise insulation and assorption first.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.