Sensitive microphone

I am working on a mini project for a school who want to detect sound levels in the classroom and been going around in circles for a few weeks. They want to measure sound levels but ranging from very low sound i.e. chair moving, door opening to levels of talking within the classroom.

I appreciate to get an accuracy levels a number of microphones located in different parts of the classroom would need to be used. But, just to be able to monitor sounds levels I would like to build a simple sound meter.

I have been trying to use the MEMS Microphone Breakout & Electret Microphone Breakout and not getting any type of response, either the voltage is too low to measure on the arudino and the pre-amplifiers do not work and just seeing loads of noise on the signal line. I have ordered the lm386 OpAmp Breakout which I hope will amplifier the signal from the microphone boards.

I have used many of the Arduino type microphones and built pre-amps using lm358, 386 and BC548 circuits but none of them provide the response I am looking for. Many of the circuits seem to be very noisy and not seeing the response which I am looking for.

Audio analogue electronics is not my area of expertise and looking for some ideas on suitable circuit and design.

Sounds like you just need a more sensitive microphone.

http://www.analog.com/library/analogdia … ivity.html

http://www.sengpielaudio.com/calculator … factor.htm

Or identify where the noise might be comming from. Is it normal white noise, or could there be a switching powersupply messing up the supply to the amplifier. Or some other kind of external noise. Just thinking. Maybe more gain isn’t the solution, just lower noise level.

An oscilloscope can point you where the problem is: in the amp or in the ADC.

The arduino adc is 10 bits IIRC, and probably a bit noisy as well.

You could try to find a higher resolution/low noise ADC (>= 16 bit/40kHz or so)

Alternately you could try to build a linear-> logarithmic converter. Then you can use the 10 bit adc at low sampling speeds.

A suggestion is shown here: figure 1 on http://www.butleraudio.com/tubesvstrans1.html and probably any TL072 or LM358 opamp will do (500pf/1.5k not needed for these opamps).

btw what’s the problem with using cheap ass sound level meters (http://www.dx.com/p/portable-1-5-lcd-di … f22-120765) ?

as marcovansteen said, you need to look at the signal first to figure out what the real problem is… use a scope and then act on else as you mentioned you will be going around in circles.

marcovansteen:
An oscilloscope can point you where the problem is: in the amp or in the ADC.

The arduino adc is 10 bits IIRC, and probably a bit noisy as well.

You could try to find a higher resolution/low noise ADC (>= 16 bit/40kHz or so)

Alternately you could try to build a linear-> logarithmic converter. Then you can use the 10 bit adc at low sampling speeds.

A suggestion is shown here: figure 1 on http://www.butleraudio.com/tubesvstrans1.html and probably any TL072 or LM358 opamp will do (500pf/1.5k not needed for these opamps).

btw what’s the problem with using cheap ass sound level meters (http://www.dx.com/p/portable-1-5-lcd-di … f22-120765) ?

I think so that this [TL072 article is bit helpful for you, and you can easily understand the datasheet to work further](Introduction to TL072 - The Engineering Projects)