Main homepage
mail me !
- piezo-electric transducers
- electret microphones
- electrostatic transducers
Piezo transducers
Piezo transducers are usually specified for a single frequency to work on, for example 40 kHz. They have low sensitivity to other frequencies. Fortunately, a lot of bats (most dutch bats) emit sounds at 40 kHz, so this is a lucky coincidence.The frequency response of a piezo transducer can be altered by adding components like coils and resistor accross them. I now have a separate page about piezo transducer detuning.
Tony Messina wrote to me about this subject:
'Some piezo transducers can have their frequency response modified (detuned) by adding inductance across them. What this does is effectively smash the peak of the frequency curve so the overall frequency response is flatter. This allows you to increase the gain of your 'flat frequency' preamp to detect sounds further from the center frequency of the transducer - effectively widening the bandpass. Another surprise feature is that, sometimes the inductance adds nodes of sensitivity to the sensor. When I added the coil to the MuRata transducer, I picked up a sharp response at 20 kHz and 33 kHz as well as getting extended response from 40 kHz to 60 kHz. This detuning trick does not seem to work with all transducers - only 1 of the 4 transducers I tried responded to detuning. A bad side effect of adding the coil across the transducer is the coil makes an excellent magnetic pickup for electrical noise - like light dimmers, electric motors, and oscillators in TV's, computers, and the like. It's OK if you are going out into the field away from electrical devices - but it renders the bat detector almost useless around urban areas.'
Take also a look at this picture at the site of a company called Massa. By adding coils and resistors parallel to the transducer, the response can be significantly broadened. I have also tried to raise the resonance frequency by gently removing some of the metal connected to the piezo-electric crystal. This resulted in greatly reduced overall sensitivity, probably due to damage.
Electret condenser microphones
Electret microphones usually are specified for up to 20 kHz, about the upper limit of the human hearing range. Above 20 kHz they have lower sensitivity, but I have no knowledge about how fast the sensitivity actually drops with increasing frequency. I expect something like a second order roll-off with a -12dB/octave slope. This lower sensitivity can possibly be corrected by using a microphone amplifier that has increasing gain with increasing frequency. It can be expected that the signal-to-noise ratio deteriorates at higher frequencies.
To use an electret microphone, you need to supply a bias voltage across
its terminals, feeding the built-in amplifier. In most cases I have seen,
there are only two terminals on the microphone and the bias voltage is
supplied through a resistor to the positive terminal. The microphone signal
is taken from the positive terminal through a capacitor.
The required circuit is shown on the right.
It is also a good idea to decouple the supply of the electret to prevent
crosstalk through the power supply lines.The resistor should have a value of 1 to 10 kOhm. The capacitor should be something like a few nF.
When I pried open an electret microphone, I saw a three-legged black component, probably a JFET. The outer leads were connected to the external terminals, while the middle lead was connected to a small metal disc with some holes in it. On top of this disc was a small plastic, metallised membrane separated from the metal disc by a white plastic ring. The membrane is electrically connected to the case of the microphone. I suppose the disc and the membrane together form a capacitor and the sound induced capacitance change causes a voltage change which is amplified by the transistor.
One particular property of an electret microphone is that it is usually omnidirectional, meaning that it sensitive to sound from any direction.
Electrostatic microphones
I have some experience with these microphones.These are said to have a reasonably response of even up to 200 kHz, not terribly flat but better than piezo transducers.
To use these, you need to supply a high DC voltage (up to 200 V) to the microphone.
I've experimented a bit with various types of circuit to produce such a high voltage.
Polaroid makes these microphones for distance finding in some of their autofocus cameras. They are also used in so-called electronic tape measures. Polaroid has have some information regarding directivity and frequency response here.
My first experiments with a Polaroid transducer are encouraging: good sensitivity and good frequency range. It is also very directional (owing to its large physical size), so problems with feedback are not very likely. You can find these microphones at e-bay for example.
Recommended microphones
Some microphones that have been recommended in the articles that I have, and by the people that I corresponded with, are:-
Knowles model
EK-3132
and the model CA-2833, electret microphones used in hearing aids.
This microphone is very small and very sensitive up to about 70 kHz, but very hard to find. - Monacor MCE-2500, MCE-2000 (electret). They are not as sensitive as the EK-3132 and do not have not so great frequency range, but are still quite good. The MCE-2000 is easy to find in electronic shops. The MCE-2500 might give slightly better performance but is also slightly harder to find.
- Toko, SE-08F-40R (40 kHz) or EDS-A250 (33 kHz) (piezo transducers)
I think for piezo transducers the choice is not so critical as it is for electrets. - Piezo transducers that I tried myself, are a transducer with the markings 'CTD Taiwan 40R' and the MuRata MA40A5R. The MuRata MA40A5R actually showed two resonance peaks, one at 40 kHz and another at 48 kHz.