Hi guys, my question is how to listen to ultrasound or at least visualize it with say a VU meter?
I have this ultrasonic transducer and I would like to build a "ultrasonic leak detector" that would help me find minute leaks in an oxygen producing concentrator machines.
I was testing my ultrasonic transducer with my scope and I get 20-40mV voltage in the 50khz range when I produce some ultrasonic sounds by rattling my keys on my keychain.
I did not test it yet but do you think that just amplifying this voltage would be enough ? Or maybe you have some other ideas?
Maybe it is possible to do some DSP with the Arduino and have PWM leds light up depending on the frequency range the ultrasound is located and the brightness would indicate the relative "loudness".
The right sensor for the job depends on the job 
Detecting a leak in an oxygen tank with sound doesn't seem like the best option to me. Are you trying to detect the sound that the leak makes when leaving the tank/tube/hose/enclosure? I guess that would be possible but considering the sound would change depending on the volume of leaking material, the location of the leak, and many, many other environmental factors, I'm not sure it's worth attempting like that.
There are fairly inexpensive sensors for detecting various gases. I'm not sure if those are sensitive enough for your purposes but you might look into those.
is how to listen to ultrasound
The simplest way is to get a bat detector. I once got a low cost kit from these people:-
http://www.magenta2000.co.uk/
Basically the ultrasonic is hetrodyned (mixed) with an oscillator to bring the frequency down.
You need a microphone rather than the normal ultrasonic transducer because these have a very limited band width.
Actually using digital ic's like 4024 in heterodyne circuits is like digital processing to me but thanks for directing me into bat detector circuits. I found one very interesting approach on; PicoBat - www.micro-examples.com This guy is using the ultrasonic piezo to act as a clock source for the PIC and you can't get a simpler code, unfortunetly I don't know how to translate it into Arduino code.
Actually I wanted to detect pressure losses in a pressurized oxygen producing machine. The oxygen concentrator as it is called uses a very noisy compressure in order to pressurize the air and push it through nitrogen absorbing micro granules. When it's working I can't listen to leaks cause the noise from the comressor is too loud, and I thought that by using the ultrasonic listening device I would locate leaks that cause degradation in oxygen purity comming out of the machine.
When it's working I can't listen to leaks cause the noise from the comressor is too loud
Maybe if you introduced a small amount of hydrogen and a spark in the vicinity of the suspected leak.
You'd hear that ;D
Have you tried getting an audio spectrum analyser to at least get you into the right ballpark frequency?
cause the noise from the comressor is too loud
So what makes you think that the compressor is not emitting sound in the ultrasonic region as well?
That pic circuit is just a divide by N circuit that toggles the speaker for every 16 or so input pulses. Easy enough to do on an arduino. Just count the number of transitions on the input line and toggle an output line every 16 times.
Maybe if you introduced a small amount of hydrogen and a spark in the vicinity of the suspected leak. You'd hear that
or at least get some water out of it
Well today I tried my modified VU meter that lights up when it hears ultrasound. I went around the oxygen concentrator "looking" for ultrasonic sources, I found too many. I have to somehow distinguish one from the others. I'll try the spectrum analyser tomorrow.
The hydrogen and sparks are a no no in this industry [smiley=shocked.gif]
Just to let you guys know that there was no ultrasonic sounds comming from my oxygen machine. I found out that my ultrasonic transducer is sensitive to low frequencies between 20- 60 hertz The sensor was picking up low frequency noise. I give up.