5

I am working on a project using a microphone/microcontroller combination to detect if a clap has occurred.

Thus far, I am using the Electret Microphone Amplifier - MAX4466 with Adjustable Gain from Adafruit, which uses a CMA-4544PF-W Elecret Condenser Microphone as well as the MAX4466 LNA.

For this project, I need to be able to detect claps from up to 30 ft away. After some testing with the microphone on maximum gain, I found that the microphone can certainly pick up sound from that distance but only strong, loud sounds. Thus, I am inclined to believe that I need a bit more gain from my microphone.

I don't know too much about audio amplification, and am curious what direction to proceed in. I would estimate that I need a gain in the ballpark of ~6dB-20dB ON TOP of the already existing microphone amplifier that I linked above.

Thus far, my thought process has led me to 3 options:

  1. Cascade another LNA with the current amplifier.

  2. Use a single-ended op-amp with a low gain. My logic here is that an LNA in the adafruit mic/amp has already amplified the signal ignoring as much noise as possible, I could possibly just use a standard amplifier at this point.

  3. Use a completely different LNA with higher gain with the microphone. This was my first idea, but I was finding that the MAX4466 already has a gain larger than most of the other LNA's I was looking at.

Any input or other ideas would be appreciated!

Zearia
  • 353
  • 1
  • 5
  • 15
  • Don't guess how much gain you need, do some measurements and calculate – PlasmaHH Jan 10 '18 at 21:00
  • My ballpark estimate is based on measuring the voltage at the microphone output from a person making noises at 30' away. It's very hard to pin an exact number on it, it's a compromise either way. At a higher gain it will work better from a distance but rail easier up close, and vice-versa the other way around. Especially since different people make different volumes of sound if I ask them to make a noise. – Zearia Jan 10 '18 at 21:11
  • You should be able to get a good estimate of the gain required by taking the loudest sound you want to record without clipping – PlasmaHH Jan 10 '18 at 21:16
  • my experience with those microphones weren't that great. SNR is not the greatest – Makoto Jan 10 '18 at 21:21

1 Answers1

3

At least for now, just add a small amplifier after the one you already have. A gain of 20, with maybe a volume control in front of it should give you the range you need with some margin.

You can rig up a little amplifier like that with something like a TL072 opamp and ±12 V supply.

Once you have your system working, and therefore know what you really need, then you can look at streamlining the whole amplifier chain.

Keep in mind that is not mostly about gain, but about signal to noise ratio. Amplifying a noisy signal does nothing useful.

Olin Lathrop
  • 310,974
  • 36
  • 428
  • 915
  • So in general, when choosing a simple op-amp like you mention should I just choose the amp with the best SNR within reason that works with my supply etc.. ? – Zearia Jan 10 '18 at 21:21
  • Not for this application, each amplifier stage amplifies the noise from the previous stages, so the noise floor of the microphone is the most critical, followed by the total noise from the first stage amplifier. The noise from your second stage is likely to be non-critical in system performance. The MAX4466 does not have particularly good noise characteristics and the microphone is just ok noise-wise. – Dean Franks Jan 10 '18 at 21:55