As I think I have told the group back in my original introduction,
I work at Bose Corporation in Framingham, MA.  As it so happens, I have
worked as part of the team that designed the worlds first active noise
canceling headset that we announced in 1989 and the third generation
headset (the Bose Aviation Headset X) this past summer.

As was mentioned in MMD 981111, these headsets have a small microphone
that sits just outside the users ear, and a driver (speaker) that
produces both the radio signals that you want to hear and the noise
cancellation (out of phase) signal that significantly reduces the
unwanted noise that comes in through the earcup.

Sound is a series of positive and negative pressures that the eardrum
moves in response to.  In a noise canceling headset, the microphone
and electronics work together to attempt to keep a constant pressure
in the earcup.  When a positive pressure enters the cup, the driver
pulls back, thereby increasing the volume and decreasing the pressure.
Conversely, when a negative pressure enters, the driver moves forward.
The wanted signal is added on top of this motion.

In an ideal situation, the unwanted noise source and the canceling
source come from the same point in space.  In practice this is not
possible, but with the geometry of our earcups, it is not that far
from correct since most of the unwanted noise passes through the plane
of the driver.

If the two noise sources are not coincident like in a larger enclosure,
then there will be places that the noise is canceled and places where
it is doubled.  If you happen to be standing where the noise is doubled
due to the interfering sound waves reinforcing each other rather than
canceling, this is truly the wrong place to be!

There is another effect that is important.  The frequency that can be
canceled is a function of the size of the cavity.  The earcups, being
very small, allow us to cancel up to about 1 kHz.  As the cavity gets
larger, the cancellation frequency rapidly goes down.  We had many
debates over a few cubic centimeters of volume in the product design
(1 teaspoon = 5 cc!).  By the time you get to a blower enclosure, it's
probably not worth the effort.

Don't let me discourage anyone, but before you go experimenting,
remember that what these sets do is to put the microphone right in
front of the driver.  Anybody who has ever heard microphone squeal
knows that this feedback can cause some rather loud, unpleasant
effects.  It takes very careful design of the feedback electronics
to maintain margin and stability.  It is a fine line between positive
feedback like you get when you put a microphone in the path of the
speakers at an auditorium and the correct feedback you need to make
noise cancellation work.

I hope this helps.  Please do not ask for specific details, as these
are company secret and therefore cannot be divulged.

Jack Breen
Southboro, MA