Need help to improve my circuit to amplify an electret microphone signal with an AD6220 In-Amp

Thread Starter

programmer786

Joined Nov 13, 2018
6
Hello,

I 've tried to connect an electret microphone to an AD620 In-Amp. I have made a circuit that works with a single 5V supply but its sound quality is quite bad:


Basically most frequences seem not to work at all, even tapping directly on the mic doesn't show any output, only if I blow air into the mic, I will get some very distorted sound.
I have tried this circuit on a breadboard first and thought the issues would come from that. I have soldered them on a PCB with components very close to each other for minimal interference but that still doesn't remove those issues.

Does anybody know how I could improve this circuit please ?
 

bertus

Joined Apr 5, 2008
20,189
Hello,

What is the application?
Why would you want to use an instrumentation amplifier?

There is no connection to the ref input.
You are using a single supply, but the IA is ment to be used fro dual power supply.

Bertus
 

Audioguru

Joined Dec 20, 2007
11,251
An ordinary low voltage audio opamp like an OPA134 works fine powered from +5V as an electret mic preamp if it is non-inverting with a input resistance of about 20k or more and has its (+) input biased from an RC filtered about half the supply voltage.

Your electret mic has its output attenuated by the extremely low value of R3 (1k) which should be 4.7k when the supply is almost 5V.
The extremely low value of 0.1uF for C2 is useless with R4 (1k) as a filter and should be 22uF. R4 and C2 can also power the voltage divider that biases the (+) input of the opamp.

EVERY electronic circuit needs a supply bypass capacitor (use 47uF) located close to the IC.
Use this circuit:
 

Attachments

Thread Starter

programmer786

Joined Nov 13, 2018
6
An ordinary low voltage audio opamp like an OPA134 works fine powered from +5V as an electret mic preamp if it is non-inverting with a input resistance of about 20k or more and has its (+) input biased from an RC filtered about half the supply voltage.

Your electret mic has its output attenuated by the extremely low value of R3 (1k) which should be 4.7k when the supply is almost 5V.
The extremely low value of 0.1uF for C2 is useless with R4 (1k) as a filter and should be 22uF. R4 and C2 can also power the voltage divider that biases the (+) input of the opamp.

EVERY electronic circuit needs a supply bypass capacitor (use 47uF) located close to the IC.
Use this circuit:
Thank you very much for your reply. I have already tried using an normal op-amp as a pre-amp (the LM393) and, while it could could output the sound, its quality wasn't high enough.
I am trying to record a very low-volume sound. If I connect the electret microphone directly to the mic input of my laptop, I get a very good recording but if I use an LM393 as a pre-amp and connect headphones to it, the sound quality isn't good enough: there is some audible background noise and the sound is slightly distorted. In my experiment, the sound needs to be very clear.
I am not sure if the OPA134 is enough in this case since I have never used it and don't know if its output is better than the LM393 or not.
I just bought 2 OPA134 online and I will test out your circuit and will try to change the passive components on the AD6220 too.
Thank you very much for your help again
 

Audioguru

Joined Dec 20, 2007
11,251
You need to learn about opamps, comparators and instrumentation amplifiers. Look in Google.
The datasheet of every opamp shows many linear amplifier circuits. The datasheet of the LM393 comparator shows no linear amplifier circuit, instead its shows logic switching circuits.

1) An LM393 is not an opamp, it is a comparator that switches its output low with a low current. It cannot be used as a linear amplifier.
2) The opamp preamp I showed is designed to drive the fairly high input resistance of a power amplifier, not low resistance headphones.
The minimum load for most opamps is 2k ohms but the OPA134 can drive 600 ohms. Your headphones are probably 32 ohms for each one.
The OPA134 can drive your headphones perfectly as detailed at https://tangentsoft.net/audio/cmoy/ .

Do not bother trying to use the AD620 in-amp because it needs an additional negative battery and its output current is too low to drive headphones.
 

Thread Starter

programmer786

Joined Nov 13, 2018
6
You need to learn about opamps, comparators and instrumentation amplifiers. Look in Google.
The datasheet of every opamp shows many linear amplifier circuits. The datasheet of the LM393 comparator shows no linear amplifier circuit, instead its shows logic switching circuits.

1) An LM393 is not an opamp, it is a comparator that switches its output low with a low current. It cannot be used as a linear amplifier.
2) The opamp preamp I showed is designed to drive the fairly high input resistance of a power amplifier, not low resistance headphones.
The minimum load for most opamps is 2k ohms but the OPA134 can drive 600 ohms. Your headphones are probably 32 ohms for each one.
The OPA134 can drive your headphones perfectly as detailed at https://tangentsoft.net/audio/cmoy/ .

Do not bother trying to use the AD620 in-amp because it needs an additional negative battery and its output current is too low to drive headphones.
Hello,
I have gotten 2 OPA134 ICs yesterday from different sellers. I have tried 2 circuits to make a working pre-amp : your circuit and another one I found online :
electret mic preamp.png
This circuit unfortunately didn't work at all. Before trying that one, I've used your circuit and, while I could hear my voice for a few minutes on my headphones with it, the low-dB sounds I wanted to hear weren't audible and the circuit destroyed the power supply after a short while. This might have to do with the problem you raised about the circuit needing higher resistance.
I have looked at the link you sent about the cmoy. This is the diagram they've given :
cmoy.PNG
I have tried to modify it for a single 5V input, is this correct please ?
cmoy5V.png

Until now, the only circuit that worked for me was the commercial one that came with the mics, which used the LM393 comparator. I found the manufacturer's diagram for it now :
-881631505.jpg
The module looks like this :
-2117952631.jpg
As said, this module does work but I need better audio quality.
Sorry for my big number of questions and my long message.
 

bertus

Joined Apr 5, 2008
20,189
Hello,

The sound quality of the last shown circuitboard will be very low.
It uses a comparator and not an opamp.

As for the single supply circuits, have a look at chapter 4 of the attached PDF.

Bertus
 

Attachments

Thread Starter

programmer786

Joined Nov 13, 2018
6
Hello,

The sound quality of the last shown circuitboard will be very low.
It uses a comparator and not an opamp.

As for the single supply circuits, have a look at chapter 4 of the attached PDF.

Bertus
Yes, you are absolutely right. The quality is very low compared to that of my laptop's soundcard for example but for now it is the only working circuit. I've looked at the 4 cases in the PDF and the first 2 seem to be the best. I will try to use the first one if possible.
Does someone know the gain range of the comparator circuit please ? I would like to use the same gain range for the OPA134 and the case 1 circuit.
 

Thread Starter

programmer786

Joined Nov 13, 2018
6
I have now measured the signal of the electret mic, it usually goes from -0.2V to +0.2V with sometimes some small surges of +/- 0.3V or in very rare cases up to 0.5V (while moving the mic around, but that signal isn't important and ,if possible, it should be filtered out).
 

Audioguru

Joined Dec 20, 2007
11,251
My preamp circuit is properly biased with a voltage gain of 215 times when the trimpot is set to 470k Then the 10mV when speaking at conversation level 10cm away from the mic will be 2.15V RMS which is 6V peak-to-peak but the 5V supply voltage is much too low for such a high output level then the output will have severe clipping distortion. It is impossible for it to destroy a power supply unless the opamp is a fake Chinese one.

The preamp circuit you found online is biased properly but its opamp bias and mic bias are missing the filter mine has, so the preamp will amplify any noise on the +5V power supply. Its maximum voltage gain is low at 101 times when the trimpot is set to 0 ohms.

The cmoy circuit you posted has a voltage gain of only 11 times for a line level input, not a from mic.

Your modification of the cmoy circuit is completely wrong with no bias causing it to rectify the audio with severe distortion and no voltage gain.

The Chinese module seems to be a circuit that blinks the LED when it hears a hands clap sound. It has an audio filter that cuts frequencies above 1600Hz. Its transistor is biased wrong and its comparator is not an audio amplifier, it has a digital output squarewave with a uncontrolled voltage gain of from 50 thousand to over 250 thousand as shown on its datasheet. If you are lucky and find a transistor that works with the bias given to it then it will multiply the gain 100 times which is from 5 million to over 250 million times.
 

Audioguru

Joined Dec 20, 2007
11,251
I notice that you are using only 5V to power the OPA134 which is its minimum supply when it barely works. You have its inputs biased at only 2.14V but the datasheet says the bias must be at least 2.5V higher than the negative supply and at most 2.5V less than the positive supply.
Use equal resistor values to set the bias at 4.5V with a 9V supply.

The bias and the power for the resistor feeding the mic should have an RC filter (1k and 47uF) like my circuit has so that the preamp does not amplify noise from the supply.
 
Top