building a "sound beam" device

Hi,
I am working on a project called "parametric array". It's a device that makes sound to be very directional. If you want to learn more about it you can visit the site: **broken link removed**
Actually to have the directionality effect, you have to mix an ultrasound signal to the audible sound you want to transmit. I realized an Amplitude Modulation circuit using as a carrier a 38kHz sine wave and as a signal 1kHz sine wave.

My college had the ultrasonic emiter given by ATC for a demo. So I just had to match the emiter to an amplifier. I have been told by ATC that my emiter had a 160nF capacitance and a 2 Ohm resistance. I built a 8W amplifier and plugged an inductor of 150µH in series with my transmitter to compensate the capacitance of the emiter.

That went good, I had quite a low volume but I could hear that the device produced very directionnal sound.


I am now trying to plug it on an audio 50W amplifier. I firstly checked that my amplifier (Rotel RA931MkII) worked good at frequencies as high as 60kHz (pluggin it on a 4Ohm resistor and studying voltage vs. frequency). That was ok.

Then I tryied it on my emiter.To be more cautious, I plugged a 4Ohm resistance in series with the inductor and the emiter. But, when I tried to make it work, the voltage at the output of the amplifier didn't go above 1V (with 1V input) and the amplifier began to spread roasted components smell :roll: so I turned it off...


Do you know something that could help me in understanding my problem?

Thanx

fadY

Audio amp works upto 60kHz ?

I am amazed :shock: that your audio amp (the Rotel RA931MkII) can cope with ultrasonics ??

I wonder if the complex load of the emitter and series resistor could be causing something like an SWR - reflecting power back into your amp and causing the output transistors to dump all this reflected power as heat.
{>> This is a GUESS <<} but you must admit you are way beyond the amp's design parameters :!:
The only other thing I can think of is that the Rotel amp's output stage is actually destroying itself as one transistor starts to conduct before the other stops (as the signal passes through quiescent) - maybe some internal flywheel effect that isn't a problem at audio frequencies {but I suppose this should be evident with your test load resistor as well?}

Given that you are still in development here I would suggest building your own simple power amp - no audio bandwidth compensation, no equalisation, no bells, no whistles - just BASIC.
A simple op-amp input stage and very simple power transistor output.
This will at least let you disprove my guess of SWR problems ?

Thanx a lot for your reply, mechie.

I am now building this power amp I will give you some news when it will be done