Electret Microphone circuit design

Thread Starter

soulhealer95

Joined Jan 17, 2017
9
Hey everyone,

I'm trying to work with this electret microphone (http://www.digikey.ca/product-detail/en/pui-audio-inc/POM-3535L-3-LW100-R/668-1496-ND/5414026), but it just wont work. I'm certain it must be something silly I can't see. Anyway, attached is the circuit. Vcc is 12V.

Here's what I have tried:
- Using a voltage divider
- Varying the Cap values
- Varying Vcc (5-10V)
- Using different Resister values
- Using different Microphones
- Amplification to up to 100 gain (using LM741 and LM348n)

I have tried using a scope right at the base of the microphone, when I tap it, there is very, very, subtle pulse but that doesn't happen always and no voice goes through at all (apparently). I have been trying to troubleshoot for four days now.

Please help me see the little bug. Tips for troubleshooting. Things to try. ANYTHING would help. Thanks! :)
 

Attachments

Thread Starter

soulhealer95

Joined Jan 17, 2017
9

Alec_t

Joined Sep 17, 2013
14,280
The mic spec states "2V". Applying 5-10V almost directly to it, if your pot/rheostat happened to be set for a very low resistance, might have fried it.
 

Thread Starter

soulhealer95

Joined Jan 17, 2017
9
The mic spec states "2V". Applying 5-10V almost directly to it, if your pot/rheostat happened to be set for a very low resistance, might have fried it.
I almost always had a voltage and current limiting resistor. But I'll try another piece with 2V.


Thanks for replying! :)
 

hp1729

Joined Nov 23, 2015
2,304
Hey everyone,

I'm trying to work with this electret microphone (http://www.digikey.ca/product-detail/en/pui-audio-inc/POM-3535L-3-LW100-R/668-1496-ND/5414026), but it just wont work. I'm certain it must be something silly I can't see. Anyway, attached is the circuit. Vcc is 12V.

Here's what I have tried:
- Using a voltage divider
- Varying the Cap values
- Varying Vcc (5-10V)
- Using different Resister values
- Using different Microphones
- Amplification to up to 100 gain (using LM741 and LM348n)

I have tried using a scope right at the base of the microphone, when I tap it, there is very, very, subtle pulse but that doesn't happen always and no voice goes through at all (apparently). I have been trying to troubleshoot for four days now.

Please help me see the little bug. Tips for troubleshooting. Things to try. ANYTHING would help. Thanks! :)
Do those op amps work well with an input signal close to ground on a single power rail application?
LM741 does not work well close to the negative rail (either power rail). Use + and - power, or an op amp, like LM358 or LM324, that works well on the low rail.
 

GopherT

Joined Nov 23, 2012
8,009
Do those op amps work well with an input signal close to ground on a single power rail application?
LM741 does not work well close to the negative rail (either power rail). Use + and - power, or an op amp, like LM358 or LM324, that works well on the low rail.
It is capacitively coupled so the bias sets the distance from the rail.
 

ian field

Joined Oct 27, 2012
6,536
Hey everyone,

I'm trying to work with this electret microphone (http://www.digikey.ca/product-detail/en/pui-audio-inc/POM-3535L-3-LW100-R/668-1496-ND/5414026), but it just wont work. I'm certain it must be something silly I can't see. Anyway, attached is the circuit. Vcc is 12V.

Here's what I have tried:
- Using a voltage divider
- Varying the Cap values
- Varying Vcc (5-10V)
- Using different Resister values
- Using different Microphones
- Amplification to up to 100 gain (using LM741 and LM348n)

I have tried using a scope right at the base of the microphone, when I tap it, there is very, very, subtle pulse but that doesn't happen always and no voice goes through at all (apparently). I have been trying to troubleshoot for four days now.

Please help me see the little bug. Tips for troubleshooting. Things to try. ANYTHING would help. Thanks! :)
Why shunt the mic with a 1k resistor? - I doubt that's helping anything.

Electret mics usually like something in the region of 2V - a dead simple way of arranging that is to put it in the emitter circuit of a grounded base stage. Simply provide the base with a decoupled bias of 2.7V - that just happens to be a standard Zener voltage.

Its not ideal because it has low input impedance and you'll have to select the collector load resistor by trial and error. But it will work consistently enough to weed out any flaky mic capsules. Although it has high voltage gain, the current gain is slightly less than 1 - you need to buffer the common base stage with an emitter follower.

A few years back I had a TL431 based electret booster published in Elektor magazine - it seems to have escaped into the WWW, but I don't have the link.
 

Thread Starter

soulhealer95

Joined Jan 17, 2017
9
Thanks a lot guys. turns out my circuit was working all along, except a minor detail that the voltage was too low even after the 200x amplification. I am getting something like around 400mV after the amplification. I send it to a voltage follower before sending it to BPF. Attached is the whole circuit. I was told it might be something to do with the first coupling cap. As in it doesn't have a path to dis-charge except the Op-Amp. So, I thought the shunt would help. Any suggestions what I can do to significantly increase the gain enough to play it on an 8 ohm speaker. I was hoping for an output of atleast 2Vpp and a DC offset so its positive. Thanks again! :) I'll try using the transistor.


Also, since a lot of you were wondering, I'm using LMC660 op-amp now (previously LM348N) which is cool because I can use single ended mode and 5 volts.
 

Attachments

ronv

Joined Nov 12, 2008
3,770
Thanks a lot guys. turns out my circuit was working all along, except a minor detail that the voltage was too low even after the 200x amplification. I am getting something like around 400mV after the amplification. I send it to a voltage follower before sending it to BPF. Attached is the whole circuit. I was told it might be something to do with the first coupling cap. As in it doesn't have a path to dis-charge except the Op-Amp. So, I thought the shunt would help. Any suggestions what I can do to significantly increase the gain enough to play it on an 8 ohm speaker. I was hoping for an output of atleast 2Vpp and a DC offset so its positive. Thanks again! :) I'll try using the transistor.


Also, since a lot of you were wondering, I'm using LMC660 op-amp now (previously LM348N) which is cool because I can use single ended mode and 5 volts.
You still have a few problems.
1- It sounds like you are using a single power supply - 5 volts. So you should generate a virtual ground Vcc/2. This will allow both the positive and negative portions of the signal to be amplified.
2- I don't know the frequency range of your signal, but the gain/bandwidth of the 660 is only 1.4 Mhz. So if you want a gain of 200 your frequency needs to be below 7Khz.
3- The LMC660 can only drive 18ma. So the most you can get into 8 ohms is about .14 volts.
Give this a read about single supplies:
http://www.ti.com/sc/docs/apps/msp/journal/nov2000/nov_08.pdf
Then maybe you can use a low cost power amp like the lm386 to get additional gain and the ability to drive a speaker.
 

Thread Starter

soulhealer95

Joined Jan 17, 2017
9
You still have a few problems.
1- It sounds like you are using a single power supply - 5 volts. So you should generate a virtual ground Vcc/2. This will allow both the positive and negative portions of the signal to be amplified.
2- I don't know the frequency range of your signal, but the gain/bandwidth of the 660 is only 1.4 Mhz. So if you want a gain of 200 your frequency needs to be below 7Khz.
3- The LMC660 can only drive 18ma. So the most you can get into 8 ohms is about .14 volts.
Give this a read about single supplies:
http://www.ti.com/sc/docs/apps/msp/journal/nov2000/nov_08.pdf
Then maybe you can use a low cost power amp like the lm386 to get additional gain and the ability to drive a speaker.
Hello!

Thanks again! :)

1. I have been suggested that before too. but when you say virtual ground, is that for the components going to the inverting input's ground?

2. frequency range is 20 Hz - 20 kHz. :/

3. That confused me too but when I tried to play a sine wave at 2 V, it played it well. :O So I figured it should be fine if I can amplify the signal somehow.
 

ronv

Joined Nov 12, 2008
3,770
Hello!

Thanks again! :)

1. I have been suggested that before too. but when you say virtual ground, is that for the components going to the inverting input's ground?

2. frequency range is 20 Hz - 20 kHz. :/

3. That confused me too but when I tried to play a sine wave at 2 V, it played it well. :O So I figured it should be fine if I can amplify the signal somehow.
I'll draw something up tomorrow.
 

GopherT

Joined Nov 23, 2012
8,009
Here. The capacitors are "optional" if you don't have any handy. Almost any size from 0.1 to 100uF are fine. Same for the resistors on the voltage divider (1k to 100k are fine).

Connect the (+) input of the left two op amps to the virtual ground with a 100k to 1M resistor. There is no resistor on the left-most currently and the second-from-left filter op amp has one going to true ground (move it to virtual ground).

Also, the two connections to the real ground (one capacitor and one resistor) should be connected to virtual ground.

image.png
 

ronv

Joined Nov 12, 2008
3,770
Hello!

Thanks again! :)

1. I have been suggested that before too. but when you say virtual ground, is that for the components going to the inverting input's ground?

2. frequency range is 20 Hz - 20 kHz. :/

3. That confused me too but when I tried to play a sine wave at 2 V, it played it well. :O So I figured it should be fine if I can amplify the signal somehow.
upload_2017-1-30_12-25-30.png
If you need to run more opamps than just this one from the virtual ground you should add the circuit Gopher posted
 

Thread Starter

soulhealer95

Joined Jan 17, 2017
9
Here. The capacitors are "optional" if you don't have any handy. Almost any size from 0.1 to 100uF are fine. Same for the resistors on the voltage divider (1k to 100k are fine).

Connect the (+) input of the left two op amps to the virtual ground with a 100k to 1M resistor. There is no resistor on the left-most currently and the second-from-left filter op amp has one going to true ground (move it to virtual ground).

Also, the two connections to the real ground (one capacitor and one resistor) should be connected to virtual ground.

View attachment 119642
Thank you guys again! I tried the Virtual Ground circuit and attached is the waveform for when I tap the microphone continuously after a two stage gain. first one 7x (turned the first voltage follower to provide a gain) and another one with 200x at end of filter. I see an output of 1.44Vpp but doesn't seem to pick up voice :/ Any Advice?
 

Attachments

Top