Experimental true condenser microphone.

Thread Starter

Arvorsteve

Joined Jan 15, 2022
9
Hi this is my first post on this forum, so thanks for inviting me in.

I am experimenting with making a large diaphragm (50mm) true capacitor (condenser) microphone using just a diaphragm and back plate mounted into an insulated housing that I will place in a mesh Faraday cage for noise reduction along with a pre amp that I need some help with please.
My aim isn't to construct something better or even as good as something than I could buy, but more to see if I can actually machine ( I have a lathe) something accurately enough that would work of sorts, bearing in mind I will need to keep the air gap separation between diaphragm and back plate as small as possible and work with some pretty thin splattered mylar materials for the diaphragm.

If you like it's a mix of an electrical and a mechanical engineering challenge.

Most of my research leads me to electret microphone circuits, which this is not as I am just using a simple two plate construction.

I know I will need to generate a bias voltage for the capacitor plates and plan to produce this using small 3v button cells in series to generate my 48V, which seems to be a standard voltage used on condenser microphones and phantom supplies. Using batteries and having them located right at the capacitor plates should keep noise to a minimum. Current drain should also be minimal as I am just using it to polarise the plates of the capacitor.

I don't really know what size of usable signal I will be able to generate but I suspect it will be small and of a high impedance.

I don't want to go and buy a pre amp ready made as that would be cheating and I won't learn much, so I was thinking of using a three op amp instrumentation amplifier to give me the high to low impedance conversion but not sure if is this is a good idea or not. Maybe just a FET input stage would be sufficient. Any recommendations on a practical design from someone that has trod a similar path as opposed to a theoretical one and how I could couple the two plate mic to the pre amp would be greatly appreciated.

Many Thanks
 

Audioguru again

Joined Oct 21, 2019
6,672
The mic diaphragm must be extremely lightweight with a very small gap to the plate. It must be rigid enough operate all in one piece like a piston.
A condenser or electret mic has no resonance or a small resonance at 18kHz but yours will have a large low frequency resonance peak caused by the weight of the diaphragm and an extremely low output level caused by the gap.

An electret mic has the 48V "built-in" to its electret material and has a Jfet and very high value resistor inside it.
 

Thread Starter

Arvorsteve

Joined Jan 15, 2022
9
Why so large? It is likely to have very poor high frequency response.

Bob
Thanks for getting back to me Bob. High frequency response drop off is not a concern, in fact I was going to go bigger once I get an amplifier that works as more interested in low frequency sound. Not looking to reinvent the perfect mic. I reckon larger means that plate separation could also be greater but would the bias voltage have to be increased as I also go larger. I suppose it would. As long as I can get something working in the speech frequency range that's all I am trying to do at this stage. Part of the fun will be to see how different sizes work out. and what I can use for the capacitive transducer.
 

Thread Starter

Arvorsteve

Joined Jan 15, 2022
9
The mic diaphragm must be extremely lightweight with a very small gap to the plate. It must be rigid enough operate all in one piece like a piston.
A condenser or electret mic has no resonance or a small resonance at 18kHz but yours will have a large low frequency resonance peak caused by the weight of the diaphragm and an extremely low output level caused by the gap.

An electret mic has the 48V "built-in" to its electret material and has a Jfet and very high value resistor inside it.
Hi and thanks for getting back to me. I can see that there are challenges ahead but I like a challenge. What size gap are we talking about and can you offer any suggestions regarding pre amp design and biasing, which is where I need some assistance?
 

Audioguru again

Joined Oct 21, 2019
6,672
The usual phrase said on an old telephone that restricts high frequencies is, "What? What did you say?" because all the important high frequency consonant sounds of speech are missing. We do not speak in only vowels.
 

MisterBill2

Joined Jan 23, 2018
18,167
There is another scheme for a capacitor microphone, which is to have it tune an RF oscillator, and then send that frequency to an FM receiver. No high voltage, but the oscillator might be in the FM band or even higher. And if the frequency is chosen correctly there would a quiet spot on an FM radio to hear it. Not much capacitance change to cause a real frequency change at 100 megahertz.
 

Ian0

Joined Aug 7, 2020
9,667
There is another scheme for a capacitor microphone, which is to have it tune an RF oscillator, and then send that frequency to an FM receiver. No high voltage, but the oscillator might be in the FM band or even higher. And if the frequency is chosen correctly there would a quiet spot on an FM radio to hear it. Not much capacitance change to cause a real frequency change at 100 megahertz.
https://en.wikipedia.org/wiki/The_Thing_(listening_device)
Full details in Spycatcher by Peter Wright
 

MisterBill2

Joined Jan 23, 2018
18,167
The condenser microphones that I am familiar with use a higher voltage, often 300 volts. Of course they also have a very low noise vacuum tube amplifier in the same enclosure, In fact I think that at least one of them has a two-stage amplifier..
That 48 volts, or less, from Phantom power, and those circuits, seem a lot more like electret microphone circuits.
And at least some of the microphones I have worked with have relatively stiffer diaphrams. The trick is to form some rounded groves in the foil around the perimeter so that the foil can move like a piston instead of moving like a flexing element.
With a metal lathe cutting the die to for forming the grove and ridges will be a rather simple exercise. Just keep the edges of the grooves smooth.
 

Thread Starter

Arvorsteve

Joined Jan 15, 2022
9
The condenser microphones that I am familiar with use a higher voltage, often 300 volts. Of course they also have a very low noise vacuum tube amplifier in the same enclosure, In fact I think that at least one of them has a two-stage amplifier..
That 48 volts, or less, from Phantom power, and those circuits, seem a lot more like electret microphone circuits.
And at least some of the microphones I have worked with have relatively stiffer diaphrams. The trick is to form some rounded groves in the foil around the perimeter so that the foil can move like a piston instead of moving like a flexing element.
With a metal lathe cutting the die to for forming the grove and ridges will be a rather simple exercise. Just keep the edges of the grooves smooth.
Thanks Bill. Pretty sure I read somewhere that 48V can work on true condenser Mics but I have also seen higher voltages. I think this circuit here that uses 200V will be a good starting point for my experiments.

https://www.researchgate.net/figure...ement-condenser-microphones-56_fig7_265036087

Also interesting to hear that the diaphragms can be stiff and still work and yes I can make a die to ridge it as I am starting with some .002" copper sheeting. I can't believe the price of high value resistors. Might make my own or use a 1n4148 reverse bias for the low voltage resistor but not sure if that would produce too much noise. What kind of gap do you think will be required between the two plates? Obviously the smaller the gap the better but I am starting with 0.5mm and hoping I will get something I can go smaller but with 200V the risk of flash over becomes a real probability as the gap decreases!
 

Ian0

Joined Aug 7, 2020
9,667
Cheers' The full article calls for 1T ohm resistors but 10G ohm is a good starting point.
Do you really hope to achieve 1TΩ? Can you actually keep it clean enough? 10GΩ seems to be good enough for AKG!
Is the low frequency cutoff equal to 2πRC where C is the microphone self-capacitance? In which case 10G would give a 20Hz cutoff for a 1pF microphone.
 
Top