Heatsink

And copper is very expensive...

yes, i think that this is why copper is not so used, but in extreme cases you can use the copper from a PCB as a heatsink...well not so extreme

May I add that when using a heatsink it may be wise to use silicone grease for a more efficient heat transfer from the device to the heatsink material .

Ongoing Subject ...

Bogdanfirst
Take a look at ...
https://www.electro-tech-online.com/threads/interfacing-ht9200-to-telephone-line.963/

This explains the soundbeam theory.

Heatsinks -- I have used copper (bits of old hot wter cylinder) - perfectly OK, steel works (not as good but still fine).

bogdanfirst said:

how about shareing your directional amp, by adding some schematics of it and so on....
but how small is the "tube" of the sound...if you understand what i mean, wow much can you focus it?

Click to expand...

here is the modulator circuit:

C1=100n
C2=100n
C3=100n
R1=56R
R2=1K
R3=1K
R4=6.8K
R5=470R
R6=470R
R7=470R
R8=680R
R9=potentiometer 50K
R10=56R
R11=680R
R12=56R
R13=3K
R14=3.9K
R15=3.9K
ic1, ic2=CA3046

I send a 34kHz sine wave as a carrier and the audio signal I want to hear as signal. I, then, send the output to the amp.

Then the output of the amp is connected in series with a 150µH inductor, a 4.7R power resistance and the ultrasonic (25µ PVDF film) speaker which has a resistance of 2Ohm and a capacity of 160nF.

The most wide the speaker the most directional sound you have because you have less diffraction. The speaker is hard to find; my school had it given by ATC which has developed this application ( https://www.electro-tech-online.com/custompdfs/2003/04/01-WPHSS-0001-RevC.pdf ).

Concerning the narrowness of the beam, I haven't measured it yet but the effect is quite amazing!

I tried my circuit on a class T amp and I have problems with the output filter. Actually class T amps modulate the analogic signal before amplifying it. After JFET transistors amplification, one have to demodulate by low pass filtering.

The filter I use is a second order LC low pass circuit which resonantes at 80kHz. But the filter has to be damped by an RC filter to prevent resonant curents so the RC filter in parallel with the capacity of the LC previous filter has to decrease the quality factor Q of the LC filter.


My problem is that, at ultrasonic frequencies, a lot of current go through this filter which make it heat (I put a high power resistor) a lot and make me loose a lot of the power given by the amp.
Do you know how I could to improve that?