Using constructive interference of audio frequency waves to lower distortion

[arivel]

Hello everybody .
I would like to ask you for an opinion.
audio waves passing through a horn or waveguide experience a small or relatively large distortion.
do you think it is possible to eliminate distortion by making use of constructive interference?

 

[sophiecentaur]

do you think it is possible to eliminate distortion by making use of constructive interference

I'm not sure what you actually mean by the question.

Bearing in mind that a horn is a matching device that works by producing 'constructive interference' of a range of frequencies, the horn will do what it can within the limits of the space available.

The horn is basically a linear device so the distortion is linear. It could be described as a wave filter because it is a distributed, rather than a lumped component filter. Horn design is based on reducing the reflected energy at the mouth by attempting constructive interference in the forward direction and destructive interference of the reflected wave.

You can't 'under drive' two horn speakers and lower the distortion level as you could with two amplifiers by adding the outputs.

If you have room to spare then a larger (longer) horn could perhaps produce less distortion.

 

[Baluncore]

do you think it is possible to eliminate distortion by making use of constructive interference?

Anything is possible.

If you have constructive interference, then you probably also have destructive interference. How can you know which one will cause more or less distortion.

The radiation pattern will depend on the frequency and the diameter of the horn, (measured in wavelengths). You can expect folded horns to have length determined frequency dependent characteristics.

How do you define distortion? As sophiecentaur says it is most unlikely that a horn will generate harmonics of a fundamental wave. On the other hand, the phase delay of the different harmonics in a complex waveform will differ. That will appear as waveform distortion when viewed with instrumentation, but your ear and brain will not be able to hear or determine the phase differences.

 

[sophiecentaur]

If you have constructive interference, then you probably also have destructive interference. How can you know which one will cause more or less distortion.

There won't be any non-linear distortions in any properly made horn. An ideal horn (basically one that is long enough) functions by the interference of reflected and transmitted waves within the length of the taper. A horn is basically a transmission line with a steadily changing characteristic impedance. Energy is constantly being reflected at every point but the reflected energy is canceled by energy reflected at other points. Ideally, there will be no net reflection over the operating bandwidth and all the energy is transmitted. It's an excellent broad band matching unit, but bulky! There will be a wavelength beyond which this effect starts to fail and a horn loudspeaker can seldom be long enough to deal with very low audio frequencies. That will produce a low frequency droop (a distortion). A folded horn introduces further distortions because the transmission line is not 'ideal'.
Across the mouth of any real horn, there will be some phase irregularities and that can affect the beam pattern at different frequencies. I have no idea whether or not it would be possible to combine (different?) horns in some way to compensate for phase problems inside the horns by suitable positioning of the horns but I wouldn't mind betting that you could do better just by making a single, bigger horn.