LM386, MIC Help required

Hi All,

I am trying to build a clap recognizer. As a first step i am trying to amplify the output of electrut microphone. My circut is attached, which i found in this forum. I am getting amplified output as 0.65 volts which does not vary with “whatever sound i make!” and fades with time i.e after sometime the output becomes zero. I am checking the ac output in multimeter. What can be the possible problem?

I tried to break the circuit and just check whether stand alone microphone is giving desired output!. Circuit attached. Samething with this also. Output (1.3v) doesnt vary with sound and fades away.

I have checked the connections, capacitor polarity, LM386 connection, all are fine.

FIrst time when i connected circuit i connected all electrolyte capacitors in reverse directions. Capacitors might have gone bad, is that why this behaviour? Also while soldering microphone it got heated up (2 mics i have soldered) will that damage mic?

Im not electronics guy. Please help me move ahead. I will appreacite all your suggestions.

Thanks,
Guru

Amplifier Circuit.bmp (4.74 KB)

Basic Mic Circuit.bmp (5.28 KB)

Are you sure that you have connected the electret microphone the right way round?

Yes, i have connected the mic the right way. I found negative terminal is the one which is connected to the metal shell.

The LM386 is really designed for a low impedance load, like a speaker. 4 ohms to 8 ohms. So its voltage output change is not going to be much. Maybe a volt peak-to-peak at full volume. And the LM386 is A.C. coupled. The "silent" output of the LM386 is normally around Vs/2.

If you are using the A.C. coupled output as in the schematic, try adding a D.C. bias on the ADC input to the Arduino. Try a 10K resistor from the ADC pin to V+ and another 10K resistor from the ADC pin to GND. That will center the input to the ADC at 2.5v. Otherwise, it could wander all over the place as far as center bias. The 10K value is just a guess, but it should get you going in the right direction.

You have a relatively high impedance input on the ADC, so you should be able to use a standard op-amp. I just recommended an LM324 to another user. That would probably work here. That could generate an almost rail-to-rail voltage on the output.

EDIT: I have previously used a gain of 100 with an LM1458 using both amps, each amp set to a gain of 10.
Vin - 4.7K - inverting input - 47K - Vout
I tried it with the gain of 100 all in one amp, but 470K on the feedback resistor did not work well. Too much noise.

gkulkarnii:
I am getting amplified output as 0.65 volts which does not vary with "whatever sound i make!" and fades with time i.e after sometime the output becomes zero. I am checking the ac output in multimeter.
...
Output (1.3v) doesnt vary with sound and fades away.

At which point(s) are you measuring these voltages, exactly? The V_OUT voltage is supposed to stay around 0v and a multimeter may not be able to detect quickly-changing voltages at V_OUT, as it takes a series of readings and displays the average value. Rougly speaking, a soundwave causes the voltage to swing up to, say, +1v, then down to -1v, which averages to zero.

The same holds true for the voltage measured after C7.

You can hook up an Arduino analog input to V_OUT_UC, do analogRead()'s and check in code whether the readings stay close to the mid-point (value of 512) or go up and down (like: below 400, above 600). Or connect a small speaker or a headphone to V_OUT and just listen to the output.

To exclude the microphone from the equation, a portable music player can be used as an input. One channel is connected before C7 and the ground is tied to the LM386 ground.

P.S. It may be best not to use those electrolytic capacitors that got wired in reverse, as they may be damaged.

Hi Dimko,

You are right. Multimeter was toggling between positive and negative readings. I will connect a speaker and check. Same at C7.

Thanks for your replies freinds. Will post back the updates.

Regards,
Guru