20khz sound
A brief explanation. It's my bedtime and I'd need to draw some diagrams to explain this properly.
This system relies on frequency mixing in your ears. There are small bones in your ear that transmitt
sound to the frequency selective nerves. I suspect that this system is making the ear bones grind against each other or move as far as they can so the transfer becomes non-linear.
Watts of ultrasonic sound in a room give me a headache and make my eyes hurt. I'v read of cases
of ultrasonic movement detectors that were on all the time and people who worked in the area
took much more sick time off work and were less happy.
When audio engineers talk about mixing they mean adding the amplitudes of two sounds together, for
example mixing the sound from a microphone and a guitar so you hear both from one loudspeaker.
When radio engineers talk about mixing them mean frequency mixing. If you have a 20Khz tone and
and 21KHz tone then a mixer will produce, 1KHz, 20khz, 21khz and 41khz.
This system shifts a signal from ultrasonic down to audio by mixing with a carrier.
Shifting a carrier that is frequency modulated with an audio signal down to audio frequncys will be unintelliagable.
If you whistle into the selective shouter at 1KHz then you get a 21KHz signal out. IF you whistle
2KHz you get 22KHz out.
If you put a 1KHz audio signal into a AM modulator with a 20KHz carrier you get a signal out that
contains energy at 19KHZ, 20KHz and 21KHz.
You add another 20KHz carrier from a different ultrasonic transducer.
Your ear picks up both signals. Depending on where you are standing in the room the
20KHz carriers will arrive with different phases. In some places they cancel out, in other
places they add. Your left ear and right ear will get different levels.
Lets say one of your ears gets a loud 20KHz sound. You get a mixer product of 19KHz and 20KHz,
the mixer product of 20KHZ and 19KHZ, and the mixer product of 19KHz and 21KHz.
You now have 2KHz as well as 1KHz. When the audio signal is speech or music instead of a tone
that 2nd harmonic distortion is going to sound really awful.
Also the two 1KHz mixer products could add or cancel depending on the relative phase which depends
on where your ear is in the room, the response of your ear to the ultrasound and the behaviour of
your abused ear bones. The frequency response will probably vary considerably across the audio
range.
You will get some sort of result with AM but I doubt it will work well.
Thats why you need to use an SSB modulator.
Let us know how you get on.
Added after 37 minutes:
Here's the selective shouter article. It's probably not worth building this. I think you could generate suitable 20KHz ultrasonic signals with some software and a good soundcard.