Hi. This is @Audioguru again's electret mic circuit (more/less). It runs on 9v and powers an electret mic putting the audio signal at Vcc/2 into the single sided op amp (TL072), and has a potential gain of 11.



If I wanted to use the same TL072 in a differential configuration at +/- 15v, what would I change? I shouldn't need to bias the input do I? I probably don't want 15v on an electret, something like 4-6v maybe?

Thanks!

 

You would be running the TL072 at its maximum rated supply voltage. I wouldn't recommend that if you want the IC to have a long and happy life.

 

Data sheet says +/- 20

 

hi rp,
As the inputs and outputs are AC coupled, have you considered using just a single +15V supply?
E

Please post a diagram of your proposed differential circuit.

 

I shouldn't need to bias the input do I?

To make the circuit the same, put a 100K resistor from + input to ground.

 

Hi. This is @Audioguru again's electret mic circuit (more/less). It runs on 9v and powers an electret mic putting the audio signal at Vcc/2 into the single sided op amp (TL072), and has a potential gain of 11.

View attachment 352982

If I wanted to use the same TL072 in a differential configuration at +/- 15v, what would I change? I shouldn't need to bias the input do I? I probably don't want 15v on an electret, something like 4-6v maybe?

Thanks!

Just leave out R2.

 

Actually, to maintain the same frequency response, R3 should be changed to 50K.

But the low frequency cutoff is ridiculously low at 3Hz in the original circuit.

 

Actually, to maintain the same frequency response, R3 should be changed to 50K.

But the low frequency cutoff is ridiculously low at 3Hz in the original circuit.

50k/100nF gives 32Hz, but C5/R5 will be setting the low-frequency -3dB point at 72Hz (admittedly, C5/R5 only rolls the gain off as far as unity)

 

50k/100nF gives 32Hz, but C5/R5 will be setting the low-frequency -3dB point at 72Hz (admittedly, C5/R5 only rolls the gain off as far as unity)

For some reason. I thought the coupling cap was 1uF. 100nF is better.

 

I was taught that where there are two time constants determining a roll-of characteristic, it was not a great idea to make them around the same value as it gives a 2nd order roll-off with a very low Q. It is better to have one of them determine the frequency response, and move the others a decade way where they have very little influence, and also that it was always better to have a film or NPO capacitor determine the frequency response, not an electrolytic.

 

Like this?

 

Like this?

View attachment 353022

Yes.
You might not need quite as many coupling capacitors with a ± supply.

 

Does this work?

 

"differential configuration" usually refers to the input circuit, but in your case I think you are referring to the power supplies. To modify the circuit for operation on dual supplies (+15 V, -15 V, and GND):

1. Delete R2.
2. Connect the bottom end of C8 to GND.
3. Connect the bottom end of R1 to GND.
4. Connect the bottom end of the mic capsule to GND.
5. Connect the bottom ends of C5, R100, and the LED to GND.
6. Add decoupling capacitors from both X1 power pins to GND.
7. Add another diode and 100 uF capacitor for the negative power input.

ak

 

You might not need quite as many coupling capacitors with a ± supply.

There is only one (C4), and it is best to leave it in there because you never know what downstream borscht might present to the output.

ak

 

Does this work?

View attachment 353023

No. You will have a DC offset on the output voltage equal to the microphone bias voltage. Remember that the op-amp has unity gain at DC. It is C4 that you might not need.