I built (on perfboard) the circuit shown in the attached schematic. And, yes, I know the op-amp is antiquated and inadequate, and the LM386 sounds like a cheap clock radio . But even taking that into account, I can't get it to work properly.

It functions to a certain extent. I can hear output at the speaker when I make a really loud noise in the mic and have the volume control wide open, but the output is not very loud and is distorted. When the circuit is powered up and there is no input to the mic, I can hear a low volume, high pitched whine, which diminishes when there is input to the mic. I have checked and double checked the connections, but can't find a wiring error. I do wonder why R1 is connected to the R4/R5 voltage divider instead of to Vcc, but that's the way the schematic shows it.

Any suggestions are welcome.

Thanks.

 

There should have been another 4.7uF to 100uF cap from the junction of R1/R4/R5/U1-5 to ground.

Also, you should connect U1-3 to that same juction, and U1-2 to U1-1. This turns U1.1 into a voltage follower so it won't oscillate unpredictably.

You'd have better results if you changed the 1458 to a TL072, TL082 or LF353.

 

There should have been another 4.7uF to 100uF cap from the junction of R1/R4/R5/U1-5 to ground.

Also, you should connect U1-3 to that same juction, and U1-2 to U1-1. This turns U1.1 into a voltage follower so it won't oscillate unpredictably.

You'd have better results if you changed the 1458 to a TL072, TL082 or LF353.

Sgt, it's good to hear from you. I hope you are doing well.

I have a TL072 and the LM1458 is in a socket; would that be a drop-in replacement?

Thanks for the help. I'll make the changes you recommended, and post back with the results.

 

C1 needs to be at least 2uF, negative terminal to the mic. 2uF would make the low frequency cutoff 80Hz for that cap. I would also make C2 larger. At 0.1uF, the low frequency cutoff for that cap (with the 10k pot) is 160Hz.

If you are only listening to birds, then these caps are probably OK as shown.

 

Here is an updated schematic with the changes recommended by SgtWookie. As it turns out, the 4.7μF cap from pin 5 of U1 to ground was already installed; I had mistakenly omitted it from the schematic. In addition, the 4.7μF cap I showed across the power pins of U1 was not installed, and I have deleted it from the schematic.

Unfortunately, the situation is unchanged from my first post. I must have a wiring error that I haven't found.

 

Here is an updated schematic with the changes recommended by SgtWookie. As it turns out, the 4.7μF cap from pin 5 of U1 to ground was already installed; I had mistakenly omitted it from the schematic. In addition, the 4.7μF cap I showed across the power pins of U1 was not installed, and I have deleted it from the schematic.

Unfortunately, the situation is unchanged from my first post. I must have a wiring error that I haven't found.

See my previous post. Also, read the LM386 datasheet. You have some parts missing.
Also, you need the 4.7uF on the power supply. Put it back in.
You can put the TL072 in your socket, but you shouldn't need the op amp. You can get gain=200 from the LM386.

 

See my previous post. Also, read the LM386 datasheet. You have some parts missing.
Also, you need the 4.7uF on the power supply. Put it back in.
You can put the TL072 in your socket, but you shouldn't need the op amp. You can get gain=200 from the LM386.

Thanks Ron. I had not seen your previous post when I made my last post. I did put the 4.7μF cap back in as C6, but it didn't help. In fact, C4 was already in place, and may have been sufficient.

Essentially what I was trying to do was duplicate the circuit in the attached pdf. I believed it worked and still think that it probably does, and that I have made some error in assembly.

I have built circuits before with the LM386 that have worked well, but this one has me stumped.

 

Did you increase the values of C1 and C2, as I suggested?
You're right, C4 should be adequate. I missed it before.

You still need to add the series RC network from pin 5 of the LM386 to GND. Otherwise, it may oscillate.

BTW, that's a crappy design in several ways. Besides C1 and C2:
1. The gain of 200 can be had with the LM386 alone.
2. If you insist on using the op amp, it should be configured as noninverting, which will give you much higher input impedance.
3. As has already been noted, the op amp needs to be replaced with a "modern" one.
4. The RC network on the output of the LM386 is missing.

I'm sure Audioguru could chime in with other shortcomings.

 

Did you increase the values of C1 and C2, as I suggested?
You're right, C4 should be adequate. I missed it before.

You still need to add the series RC network from pin 5 of the LM386 to GND. Otherwise, it may oscillate.

BTW, that's a crappy design in several ways. Besides C1 and C2:
1. The gain of 200 can be had with the LM386 alone.
2. If you insist on using the op amp, it should be configured as noninverting, which will give you much higher input impedance.
3. As has already been noted, the op amp needs to be replaced with a "modern" one.
4. The RC network on the output of the LM386 is missing.

I'm sure Audioguru could chime in with other shortcomings.

I didn't change C1 and C2 since I am not concerned with frequency response at this point.

I put in the RC network on pin 5. No change.

I added a bypass cap on pin 7. No change.

I swapped out the LM386. No change.

I swapped out the 1458. No change.

I put in a TL072. No change except for a slight increase in the frequency of the whine.

I know it's not a good design, but it must work; they have sold thousands of them. What I really want is the Nuts and Volts version from December of 2008, but it's more money than I want to spend.

I think I'll go rake leaves; maybe I can succeed at that.

Thanks for your help.

 

I didn't change C1 and C2 since I am not concerned with frequency response at this point.

I put in the RC network on pin 5. No change.

I added a bypass cap on pin 7. No change.

I swapped out the LM386. No change.

I swapped out the 1458. No change.

I put in a TL072. No change except for a slight increase in the frequency of the whine.

I know it's not a good design, but it must work; they have sold thousands of them. What I really want is the Nuts and Volts version from December of 2008, but it's more money than I want to spend.

I think I'll go rake leaves; maybe I can succeed at that.

Thanks for your help.

OK, but C1=0.1uF will cause all frequencies below 1.5kHz to be attenuated - the lower the frequency, the more it will be attenuated. Keep in mind that the telephone voice bandwidth is ≈300Hz to 3kHz.

Who has sold thousands of them? N&V?
Can you post a link to the schematic you are trying to copy?

 

Who has sold thousands of them? N&V?
Can you post a link to the schematic you are trying to copy?

It's attached to post number 7 in this thread.

 

It's attached to post number 7 in this thread.

I was hoping to see the original. I don't think that's it. Is it?

EDIT: never mind. I found it.
http://www.rainbowkits.com/kits/instructions/AA1.pdf

Edit: Have you tries 0.1uF from pin 8 to pin 4 on the op amp, as close as you can get it? It Wouldn't hurt to also add one from pin 6 to pin 4 on the LM386, again with leads as short as possible. The whine you are hearing is probably the result of oscillation.

 

The LM386 power amp has a gain of only 20. Adding one capacitor boosts its gain to 200. Its input impedance is high so it does not load down and attenuate the mic.

The horrible mic preamp opamp would have a gain of 10 if the mic was a very low impedance but it is not. The gain might be only 6.

The horrible mic preamp opamp is inverting with a low input impedance that loads down and attenuates the mic. It should be non-inverting with a high input impedance.

The 3.3k resistor that powers the Jfet inside the electret mic should be 10k with the 9V supply. Then its output signal level will be higher.

The resistor that powers the mic should be fed from an RC filter.
The voltage divider that biases the horrible mic preamp oipamp should also be fed from an RC filter.

The 3.3k resistor is connected to the wrong place.

If the LM386 is set for a gain of 200 and the horrible mic preamp opamp removed then none of these problems will exist.

 

The LM386 power amp has a gain of only 20. Adding one capacitor boosts its gain to 200. Its input impedance is high so it does not load down and attenuate the mic.

The horrible mic preamp opamp would have a gain of 10 if the mic was a very low impedance but it is not. The gain might be only 6.

The horrible mic preamp opamp is inverting with a low input impedance that loads down and attenuates the mic. It should be non-inverting with a high input impedance.

The 3.3k resistor that powers the Jfet inside the electret mic should be 10k with the 9V supply. Then its output signal level will be higher.

The resistor that powers the mic should be fed from an RC filter.
The voltage divider that biases the horrible mic preamp oipamp should also be fed from an RC filter.

The 3.3k resistor is connected to the wrong place.

If the LM386 is set for a gain of 200 and the horrible mic preamp opamp removed then none of these problems will exist.

Yeah, but he wants to replicate a POS circuit. I think I'm gonna bail from this thread.

 

I was hoping to see the original. I don't think that's it. Is it?

EDIT: never mind. I found it.
http://www.rainbowkits.com/kits/instructions/AA1.pdf

Edit: Have you tries 0.1uF from pin 8 to pin 4 on the op amp, as close as you can get it? It Wouldn't hurt to also add one from pin 6 to pin 4 on the LM386, again with leads as short as possible. The whine you are hearing is probably the result of oscillation.

That made a big reduction in the whine. I can still hear it, but it's much reduced. I am going to add the cap across the LM386 gain pins to see what that does.

 

I am going to add the cap across the LM386 gain pins to see what that does.

The datasheet shows you exactly what it does. It increases the voltage gain of the LMN386 10 times. Then if the mic can hear the speaker (headphones?) there will be acoustical feedback howling.

 

The datasheet shows you exactly what it does. It increases the voltage gain of the LMN386 10 times. Then if the mic can hear the speaker (headphones?) there will be acoustical feedback howling.

I wondered when you would make yourself known.

It does increase the gain, but there is no feedback unless I put the mic against the speaker cloth and turn R6 wide open. As is my experience with electret mics, they seem to require very close proximity to the audio source.

ETA: I missed your post #13.

 

Yeah, but he wants to replicate a POS circuit. I think I'm gonna bail from this thread.

Thanks for your help.

 

If the LM386 is set for a gain of 200 and the horrible mic preamp opamp removed then none of these problems will exist.

OK, I'll do that. I thought I remembered from an earlier post of yours that a preamp was required for a mic used with an LM386.

Oh yeah, the Jfet preamp you were working on was to provide a very high input impedance to a guitar pickup. Its voltage gain is nothing or is about 3dB which is slightly more than nothing.
But an electret mic needs a preamp with a voltage gain of 100 to 400.

 

See? "But an electret mic needs a preamp with a voltage gain of 100 to 400."
Adding one capacitor to an LM386 power amp IC boosts its voltage gain to 200 which is a lot of gain. Then the horrible mic preamp IC you have is not needed.