There are commercial noise cancelling headphones available which work very well. But as I like to experiment, I thought I would try to design a pair which fit in the ear. I finished up using a pair of Sony folding 'in ear' headphones that I had owned for a while and didn't mind modifying. The following design is not perfect but does work at certain frequencies. The idea of including this page is that hopefully someone will modify the design and make it even better and then tell me!
|
The picture left shows one side of the modified headphones. A small electret condenser microphone has been glued to the side of the headband with araldite near the headphone. A small round notch was filed in the headband to accommodate the microphone and keep it in place whilst the glue hardened. The cable (removed from another 'cheap' pair of 'in ear' phones) is glued along the length of the original cable with superglue. This is of course done on both sides of the head phones so that there is one microphone for each headphone. |
Design Approach
The basic idea of noise cancellation is to feed a signal into the headphones which is 180 degrees out of phase (antiphase) with the outside noise which you want to cancel. On commercial systems, this is achieved by placing a small microphone between the headphone and the ear, which picks up the noise reaching the ear and feeds an antiphase version back to the headphone at the same volume, causing cancellation. Problems occur with feedback at certain high frequencies as the system self oscillates if you are not careful. It is not easy for the amateur to make such a system using todays headphones which tend to be quite flat when placed on the ears, leaving no room to fit a microphone. This sytem is very good at cancelling low frequencies if the headphones are large enough and have a good bass response. My design uses a different approach. The signal arriving at the headphones is picked up by the small microphone in the picture above, amplified, inverted to provide the antiphase version and fed to the headphone. As the microphone and headphone are so close together, the time difference and phase lag of the signal is small (but nevertheless is present). The drawbacks of the system are as follows and this is where I would like some help.
Circuit Description
 |
The circuit shows one channel of the system. The whole circuit requires just one quad op-amp, a TL084 type. It is very simple to understand. The signal from the microphone is amplified by the first op-amp and this signal is passed to the second op-amp via a preset labeled 'null'. The final signal is fed to the headphone via a passive mixer (2x100 ohm resistors) which allows the headphones to be connected to a 'walkman' stereo cassette at the same time, or the headphone socket in an aircraft entertainment system. |
The 1uF capacitor in parallel with one of the 100 ohm resistors compensates for a small loss of high frequency response due to the inclusion of the 100 ohm resistor in the first place!
The headphones are worn whilst listening to a noise source. A computer fan or boiling kettle are ideal. The 'null' preset is then adjusted to give the best cancelling effect. There is a problem with earphone position within the ear as altering the depth can cause the cancelling effect to alter. But it is easy to find the most useful position and calibrate for this and return to this position by noting the cancelling effect at the same time. If no cancelling effect can be found, then it is possible that the headphones are wired in the opposite phase and will need the wires swapping around. Also make sure that the mic on the left goes to the headphone on the left and the same on the right.
The circuit is powered from a PP3 type 9v battery which will last many hours. I used one battery from England to Australia (24 hours continuously) and back again and it was still OK. What's more is that the cancelling effect did not alter during the whole time.