Active Noise Cancellation 2

I have installed a couple of active noise systems on cars. One was internal, for engine noise, and one was external, for exhaust noise.

Both worked.

I also helped with a couple of road-noise cancellation projects.

I am 90% certain that you will get a better more cost effective solution if you use normal acoustic techniques (shielding, isolation, absorption and silencers) rather than ANC.

However, if you are determined to persist there are a few comapnies out there who will sell you a kit. A google search will find them quickly, Lord in the USA and some mob in Australia come to mind.

ANC can work well if you can predict what the sound is going to sound like by the time it reaches the listener. There's two ways of doing that - get advance notice of what the sound is , by putting a mic close to the source (in which case the maximum frequency you will be able to cancel is controlled by the speed of sound and the distance), or else using teh previous cycle to predict the next cycle, which only work with deterministic signals.

Incidentally free field cancellation of random noise is not practical, and I think that may be your aim.



Cheers

Greg Locock

It seems to me that a conventional reactive or passive silencer would be sufficient for the pneumatic exhaust. An active system is much more costly and has maintenance/durability issues.

Unless you have the restraints of size, temperature, pressure drop and or dirty gas flow etc, use a conventional silencer. C. Hugh (

When you talk about reducing noise from a pneumatic conveyor exhaust, do you mean simply reducing the noise from air exhausting from a pneumatic motor or some such? If so, you'd perhaps be better off using a conventional pneumatic muffler.
The field of active noise cancellation was vastly over-hyped a few years ago, with people applying active noise to many problems for which it's not at all suited. It is best for applications that simply cannot be solved by other means, due to the high cost of active noise solutions.
For your conveyor problem, if you can't isolate a discrete source of the noise, active noise would be a poor solution. Cancellation depends on being able to have the "antinoise" signal 180 degrees out of phase with the noise; this in turn requires one of two things: 1) knowing exactly where your receiver or listener is located, since differences in spatial position are equivalent to differences in phase in the time domain; or 2) providing cancellation at the source.