Best sound from a DIY microphone

Thread Starter

Lectraplayer

Joined Jan 2, 2015
123
Although I know I probably could buy a microphone that does some of what I'm wanting to do much better, I am wanting to experiment with building some audio capture hardware from spare parts gained from tearing down broken devices. I may mostly be using dynamic mics from small (1.5 inch or so) speakers as my primary pickup. My goals are simply to get the best sound I can, whatever that may be. What are some ways of improving the response of most mics? I'm guessing that using speaker tuning methods, such as a resonator chamber, packing, ports, etc will make a significant difference. I have also been looking at shotgun mics lately as well as cartioid and omnis.
 

Thread Starter

Lectraplayer

Joined Jan 2, 2015
123
I actually have a range of purposes in mind, ranging from talking to friends on Discord (on a computer) and toying around with vintage recording equipment (those would typically be an omni or cartioid, picking up 1KHz or so, and "flavoring" the sound is acceptable) to trying to catch something far off (shotgun mic.) I am mostly experimenting though. I plan to address the amplification and equalization once I have the mic constructed and under test, so I'm mostly wanting to concentrate on the sound capture aspect of mic design here.
 

KeithWalker

Joined Jul 10, 2017
2,602
Electret capacitive microphones have very flat characteristics and are quite sensitive. For distant sounds, have you considered a parabolic reflector? They are very effective. You can do most of the other stuff with an inexpensive equalizer board. I got one from Aliexpress, put it in a case with a power supply and added it to my sound system.
Regards,
Keith
 

Audioguru again

Joined Oct 21, 2019
5,404
A small speaker is probably the worst sounding microphone since it resonates strongly at about 200Hz like a bongo drum. They are used in gas station outdoors cheap intercoms.

Nowadays, reporters in the TV news are in virus lockdown so they use Skype. Every few seconds they sound like chipmunks. Why is the sound quality so bad??
 

Thread Starter

Lectraplayer

Joined Jan 2, 2015
123
...so an anti-200 filter may help. :D Depending on other characteristics of the mic's performance, this may be something I can handle. (Again, these won't be the best mics in the world. I'm just experimenting here.)

Gotta love lag. Bane to both gamers and videochat.
 

KeithWalker

Joined Jul 10, 2017
2,602
...so an anti-200 filter may help. :D Depending on other characteristics of the mic's performance, this may be something I can handle. (Again, these won't be the best mics in the world. I'm just experimenting here.)

Gotta love lag. Bane to both gamers and videochat.
You can characterize a speaker if you have an audio oscillator and a scope or ac voltmeter that is fairly flat up to 15 KHz. An ideal speaker should give a flat amplitude response across it's spectrum but if resonance occurs, it will be very noticeable on your monitoring device. You can use a low level to drive the speaker because, as a microphone, it will only be handling small signals. The characteristics will change if you use higher level signals, as a loud speaker. Once you have done that, you know at what frequencies resonances occur and how flat it is over the rest of the spectrum. It will also show you how the enclosure affects the response of the speaker. It will give you a good starting point for further experiments.
Regards,
Keith
SpeakerTest.jpg
 
Last edited:

MisterBill2

Joined Jan 23, 2018
13,686
Although I know I probably could buy a microphone that does some of what I'm wanting to do much better, I am wanting to experiment with building some audio capture hardware from spare parts gained from tearing down broken devices. I may mostly be using dynamic mics from small (1.5 inch or so) speakers as my primary pickup. My goals are simply to get the best sound I can, whatever that may be. What are some ways of improving the response of most mics? I'm guessing that using speaker tuning methods, such as a resonator chamber, packing, ports, etc will make a significant difference. I have also been looking at shotgun mics lately as well as cartioid and omnis.
The very first step is to decide what you mean by the best sound. That could mean the loudest, the clearest, the broadest flat frequency range, the most sensitivity, or the least distortion.
Speakers can function as microphones but you do not get microphones from speakers.
The classic "shotgun" microphone uses an array of resonant tubes to provide gain in a single direction, and that leades to needing a microphone with a larger sensitive area. And the frequency response is determined by the number and length of the tubes, so they can be a bit big.
 

Thread Starter

Lectraplayer

Joined Jan 2, 2015
123
The very first step is to decide what you mean by the best sound. That could mean the loudest, the clearest, the broadest flat frequency range, the most sensitivity, or the least distortion.
Speakers can function as microphones but you do not get microphones from speakers.
The classic "shotgun" microphone uses an array of resonant tubes to provide gain in a single direction, and that leades to needing a microphone with a larger sensitive area. And the frequency response is determined by the number and length of the tubes, so they can be a bit big.
My purpose isn't to build the "world's best microphone" from spare parts, as much as to investigate actually turning sounds into electrical signals based on what I have handy, creating a large number of working microphones in the process for varying purposes. That said, I'm not really expecting to match the performance of anything that I can buy. ...though yes, I also do know from research that a well made shotgun microphone can be two or three feet long, and quite a chunk. So far, my logic works as follows:

Loudness: Can be amplified or attenuated as needed, but louder is always better.
Sensitivity: Probably limited in my choice of elements, but I intend to evaluate as much as practicable.
Clarity: Probably limited in my choice of elements, though I will be looking to maximize performance. Most gains my be in my environment instead of the mic itself.
Response: To be determined as much as possible, though most mics will be focused on voice (I'd guess 500 to 5KHz.) Some mics may be tuned to other frequencies. If I can start turning out some decent full range mics, I will.
Distortion: To investigate, though less is always better. Likely not easily adjustable given my choice of elements, aside from amplification.

I have also seen some people who build shotgun mics use a funnel to concentrate the sound on a smaller element, though how well it works, I don't know yet. (Probably about as well as a minivan at Talledega.)
 

Audioguru again

Joined Oct 21, 2019
5,404
Response: To be determined as much as possible, though most mics will be focused on voice (I'd guess 500 to 5KHz.)
144 years ago AG Bell invented the telephone. He claims it had a frequency response of from 300Hz to 3kHz but it was probably narrower. The narrow response was the best he could do in those days. Later telephones copied the narrow frequency response because to make it wider would cost a lot more.
You should get a audio equalizer program to hear what your 500Hz highpass and 5kHz lowpass sound like. Listen to your own voice, it probably goes down to 60Hz and produces the important consonants of F, TH, S SH and many other up to 14kHz or higher. Your cutoff at 500Hz will make your voice sound like a chipmunk and your cutoff at 5kHz will have listeners saying, "What did you say?" often. A lousy old AM radio also has a cutoff at about 3kHz, but they boost frequencies from 1kHz to 3kHz to make voices sound better.
I have an audio equalizer program called Peace and I use it to try making narrow band voices and music sound more normal.
 

MisterBill2

Joined Jan 23, 2018
13,686
My purpose isn't to build the "world's best microphone" from spare parts, as much as to investigate actually turning sounds into electrical signals based on what I have handy, creating a large number of working microphones in the process for varying purposes. That said, I'm not really expecting to match the performance of anything that I can buy. ...though yes, I also do know from research that a well made shotgun microphone can be two or three feet long, and quite a chunk. So far, my logic works as follows:

Loudness: Can be amplified or attenuated as needed, but louder is always better.
Sensitivity: Probably limited in my choice of elements, but I intend to evaluate as much as practicable.
Clarity: Probably limited in my choice of elements, though I will be looking to maximize performance. Most gains my be in my environment instead of the mic itself.
Response: To be determined as much as possible, though most mics will be focused on voice (I'd guess 500 to 5KHz.) Some mics may be tuned to other frequencies. If I can start turning out some decent full range mics, I will.
Distortion: To investigate, though less is always better. Likely not easily adjustable given my choice of elements, aside from amplification.

I have also seen some people who build shotgun mics use a funnel to concentrate the sound on a smaller element, though how well it works, I don't know yet. (Probably about as well as a minivan at Talledega.)
Microphones often have trade-offs, some are intended for strictly speech communication, others are aimed at a wide frequency range with uniform response over the range. That is a reall accomplishment, and often they are not as sensitive as others. I have read a few articles about folks who actually have built their own microphones, and it seems that there are a large number of variables involved. One challenge will be the piezo-electric element because the output is at a very high impedance and low voltage, so they always get an FET transistor amplifier right in the microphone element. But they are easy to make and they often sound good.
 

MisterBill2

Joined Jan 23, 2018
13,686
Here is the frequency response of a dynamic (coil and magnet) Shure vocals mic. They boost the mid-highs to give it some shrill "presence".
Here is the frequency response of an ordinary electret mic:
The spec for the SM-58 looks a lot like the spec for the ASTATIC D-104 Crystal Mic. And the claimed curve for that electret mic is a whole lot flatter than the specs for the electret elements in the DigiKey catalog. And the desired frequency response for any microphone depends a whole lot on what the use of the signal is going to be.
 

MisterBill2

Joined Jan 23, 2018
13,686
I think I used an Astatic crystal mic about 63 years ago.
Here is how it compares to the Shure vocals mic:
OK, the same sort of shape of the curve but different frequency peaks. The D-104 was intended for use with tube-type AM transmitters. It also works well with newer transistor radios with the impedance matching transistor amplifier added.
 

MrChips

Joined Oct 2, 2009
27,651
Loudspeakers make lousy microphones.

The best allround microphone is the electret microphone. You can salvage them from so many devices: telephones, mobile phones, headsets, cameras, audio recorders, etc.
 
Top