Piezo transducer as both microphone and speaker

Question

I'm trying to find the best piezo transducer for my project, which is a Morse code sender and receiver.

Since a piezo transducer can both act as a speaker and a microphone, and I only send/receive at a very specific frequency (700 Hz), and I only send or receive (not both at the same time), I think I can pull this off with just a single transducer connected to both the ADC and DAC of my microcontroller.

When looking through component websites (Digikey, etc.), I can't find any piezo transducers that have data for usage as microphones. Even though they are called transducers, the intention seem to be to only use them as speakers. To find the ideal component, I need to see microphone sensitivity, microphone noise level, etc.

Has anyone done something similar, and can give me some pointers?

Answer 1

Many small piezo transducers come mounted in an enclosure which acoustically resonates at a specific frequency - usually much higher than the pleasing 700 Hz tone desired...a TDK piezo described here...

This one's acoustic resonance is near 4 kHz. But just how seriously resonant is it? Acting as a raw microphone, it was monitored with an oscilloscope. White(ish) noise was blown across it momentarily from a few centimeters away:

The enclosure surrounding the embedded piezo greatly magnify sound intensity - but only at resonance. At 700 Hz, acoustic efficiency is very poor...there is very little amplitude at 700 Hz, but large amplitude at 4000 Hz.

Usually, much larger transducers have lower resonant frequency. You might find one whose resonance is nearer to 700 Hz than ubiquitous tiny ones.
A highly-resonant enclosure like this makes it a poor microphone for voice use. For Morse code, a 700 Hz. resonant transducer would be more useful, since bandwidth of Morse code is so narrow.

Based on the amplitude generated by the 4kHz transducer (above), a single-transistor amplifier stage might be useful to drive an ADC with a larger voltage than the ~50mV shown here.

Answer 2

Not with a piezo but with a standard loudspeaker, yes. Many years ago I designed a comms system with many distributed loudspeakers which were used to sense ambient sound (and thus adjust the talk signal volume). It worked pretty well and AFAIK is still in use in a London theatre.

I don't know why you chose piezo, but if you are not locked into that you could experiment with inexpensive speakers for the same purpose (which probably sound less annoying and give you more flexibility anyway). You won't find any manufacturers data, it is a pretty non-standard use. Just try it. As the voice coil is low impedance it is rather easy to amplify. (And if you do need piezo for some reason, I'd say the same - just try it and see how it works.)

Some recording engineers use a speaker of about 6" as a mic on kick drum as well. Apparently it works pretty well.

Answer 3

I agree that a piezo element is the wrong part for the job. A small, relatively high impedance speaker, such as 32 or 60 ohms, works just fine as a microphone, a common technique in many intercom systems. However, it needs a preamp.

Because of that, consider separating the tasks and using an actual microphone. For your application, a preamp stage for an electret microphone cartridge is not a complex circuit - 1 transistor and 2 resistors (plus coupling caps). This is arguably less effort than whatever you were going to do to switch the piezo element between the DAC output amplifier and the ADC input. If you were thinking of driving the piezo directly out of the DAC, so the DAC output and ADC input were tied together, that will not work for several reasons.