Low Frequency Electret Mic

Go large on the coupling cap and likely most will work.

WTF makes you think otherwise??

.... Phil

For simplicity you will need an omni-directional capsule, rather than a cardioid one - and the larger the diaphragm, the better.

The L.F. sensitivity will be greater with a larger diaphragm because it is more flexible and can enclose a larger air space. The larger capacitance gives a longer time constant with any resistance to earth at the gate of the FET, so the electrical roll-off frequency is lower. You won't need to worry about diffraction effects from the larger diaphragm if you are only interested in low frequencies. Beware of built-in transformers, cheap ones often distort at low frequencies and even good ones would have problems handling full power at 15c/s - this applies all the way through the signal chain.

If you want to use a cardioid capsule, these can be compensated electrically for the 6dB/octave roll-off below about 270c/s . I have taken the Panasonic WM 55A103 down to 30c/s by this method and, no doubt, could have extended it another octave lower if I had needed to. However, cardioids tend to lose directionality at these frequencies and the extra L.F. boost makes them very susceptible to flicker noise on the power supply; a simple voltage regulator is far too noisy for quality work, batteries can't be relied upon either, so special supply circuits must be used.

If you think you might want to extend your response down a lot lower in future, it is best to consider R.F. bridge capacitor transducers, or even optical systems, rather than high impedance electrets

Use the one with the built-in follower and use a LARGE cap, also baffle the pickup area to enhance low frequency response.

From memory, cheap electrets went down below 30Hz, but you can kind of test and add a bass-boost circuitry, too.

A while back I had an electret booster published in Elektor - it was DC coupled.

The electret likes close to 2V, stabilising the operating point of a TL431 by rolling it back with loads of DC nfb will put a steady 2.5V on the control input - that's close enough for an electret.

Just connect the control input to GND via the electret and set the operating point with a 47k pot from cathode to control, once you've determined the value for a fixed resistor you can split that resistance for a tapping point, then you can shunt the AC nfb to GND for awesome gain.

Most of the omni capsules are sealed, but might have an air pressure relief port. This gives the ability for LF response. If you baffle, it's likely to enhance upper range response, not LF. I've mounted a capsule through a hole in a pressure zone plate. You can seal the back if you want to, but I have doubts it will help. You can always test.

Greg

Maybe you should look into a MEMS accelerometer.