Public Address System

[p.gholap93]

I am doing a small public address system in which whatever I talk on microphone I want to hear that on a speaker.Now I am hearing the sound but the problem is that intensity of the sound is too low.I am using LM386 as power amplifier.Its output is DC shifted sine wave with DC shift of 4V.The peak amplitude is 0.4V.So now I want to amplify only this AC component without any DC so that I will hear sufficient sound.So what could be the solution for that?I also attach the circuit diagram.

 

[Nigel Goodwin]

The LM386 is a only a tiny amplifier, you need a bigger amplifer, and also a preamplifier.

 

[moffy]

You might want to use headphones. If you get too much signal (speaker volume) near the mic you can get feedback.

 

[ronv]

I don't know how much louder you need but you could get quite a bit more by using the Gain = 200 circuit found here. https://www.electro-tech-online.com/custompdfs/2011/03/LM386-1.pdf You may also want to use 9 or 10 volts for the power supply when you do that.

 

[audioguru]

Since you have the LM386 amplifier set with a voltage gain of only 20 and you measure an output peak level of 0.4V then its input level has a peak of only 20mV which is the level from a microphone.

Look on the datasheet of the LM386 amplifier where it shows how to connect a 10uF capacitor to it (between pin 1 and pin 8) to increase its voltage gain to 200. Then its output will try to have a peak level of 4.0V but it will be clipped because it does not have enough supply voltage.

The 0.01uF input capacitor feeding the 10k volume control cuts all low frequencies. It reduces 1928Hz to half power and cuts lower frequencies much more. Use a 0.22uF input capacitor and a 50k volume control so that 29Hz will be at half power.

 

[p.gholap93]

Since you have the LM386 amplifier set with a voltage gain of only 20 and you measure an output peak level of 0.4V then its input level has a peak of only 20mV which is the level from a microphone.

Look on the datasheet of the LM386 amplifier where it shows how to connect a 10uF capacitor to it (between pin 1 and pin 8) to increase its voltage gain to 200. Then its output will try to have a peak level of 4.0V but it will be clipped because it does not have enough supply voltage.

The 0.01uF input capacitor feeding the 10k volume control cuts all low frequencies. It reduces 1928Hz to half power and cuts lower frequencies much more. Use a 0.22uF input capacitor and a 50k volume control so that 29Hz will be at half power.

Ok I will try your second trick.I connected a pot in series with capacitor between pin no 1 and 8 to vary the gain.The gain is varying.I also connected only 10uF between pin 1 and 8 without resistor but still not sufficient whatever the maximum comes out to be. I was thinking of using a preamplifier (op amp noninverting using LM324) but still the problem of DC shift remains.

 

[p.gholap93]

The LM386 is a only a tiny amplifier, you need a bigger amplifer, and also a preamplifier.

Ya but what about DC shifting.Output of microphone is DC shifted with very small amplitude.So if I use a preamplifier naturally this DC shift will be raised at the output thereby causing distortion.

 

[p.gholap93]

I don't know how much louder you need but you could get quite a bit more by using the Gain = 200 circuit found here. https://www.electro-tech-online.com/custompdfs/2011/03/LM386-2.pdf You may also want to use 9 or 10 volts for the power supply when you do that.

The sound should be audible at least.Now I have to take my ears close to the speaker to listen whatever I speak in front of mic.

 

[Nigel Goodwin]

Ya but what about DC shifting.Output of microphone is DC shifted with very small amplitude.So if I use a preamplifier naturally this DC shift will be raised at the output thereby causing distortion.

DC shift isn't relevent, it's an audio amplifer so you only amplify the AC component. The input capacitor blocks DC going in, and the output capacitor blocks DC going out.

 

[audioguru]

You have something wrong with your circuit. Post your schematic.

I have an LM386 with its gain set to 200 with a 10uf capacitor between pin 1 and pin 8. My electret mic is biased with a 10k resistor from the positive supply and is coupled to the LM386 input with a 0.22uF coupling capacitor feeding a 100k volume control. It is so loud that I turn down the volume control.

 

[carbonzit]

You have something wrong with your circuit. Post your schematic.

I have an LM386 with its gain set to 200 with a 10uf capacitor between pin 1 and pin 8. My electret mic is biased with a 10k resistor from the positive supply and is coupled to the LM386 input with a 0.22uF coupling capacitor feeding a 100k volume control. It is so loud that I turn down the volume control.

Audioguru, if you don't mind indulging me a moment, I've got a few questions/comments that might take us on a bit of a tangent in this conversation.

1. Regarding that schematic: you didn't draw that, right? It looks a lot like the National Semi style of schematics one sees in books, etc. (I think it's National Semicircular, anyhow.) I happen to like that style of schematic drawing (planning on starting a thread about schematics here soon).

2. In that circuit, a couple of things bug me. First of all, what's the purpose of that 1K resistor and 100uF cap there on the left? Power-supply decoupling? Why not just leave them out and run the 10k resistor right from V+?

Also, what's the purpose of the small (,047uF) cap and 10-ohm resistor on the output? Shouldn't the 8-ohm speaker load be sufficient (through the coupling cap, of course)?

3. Strictly speaking I wouldn't say that the electret is "biased" by the 10K resistor: it's actually powered by it, right? AFAIK, electrets (usually) contain a FET amplifier, no?

4. Interesting that you say it's loud enough to have to turn down the volume. Those op amps are pretty amazing devices. But again strictly speaking, shouldn't we be talking about the output in terms of power (watts, or at least milliwatts), and not in terms of voltage, since we're using this as a power amp?

5. Finally, isn't it a bit ridiculous to be using a puny op amp for such an application? When the OP posed a question about a PA system, I ass-u-med that they were asking about a proper PA, one that could actually be used to, well, address the public. This clearly won't do it; how come nobody has pointed that out?

Not meaning to knock ya, but I like to point out these things when I encounter them.

 

[Nigel Goodwin]

2. In that circuit, a couple of things bug me. First of all, what's the purpose of that 1K resistor and 100uF cap there on the left? Power-supply decoupling? Why not just leave them out and run the 10k resistor right from V+?

You already answered that, it's power supply decoupling - running the 10K directly to the supply is likely to encourage instability and oscillation.

Also, what's the purpose of the small (,047uF) cap and 10-ohm resistor on the output? Shouldn't the 8-ohm speaker load be sufficient (through the coupling cap, of course)?

No, it's called a Zobel network, look it up - again it's for stability, and VERY important.

3. Strictly speaking I wouldn't say that the electret is "biased" by the 10K resistor: it's actually powered by it, right? AFAIK, electrets (usually) contain a FET amplifier, no?

4. Interesting that you say it's loud enough to have to turn down the volume. Those op amps are pretty amazing devices. But again strictly speaking, shouldn't we be talking about the output in terms of power (watts, or at least milliwatts), and not in terms of voltage, since we're using this as a power amp?

It's not an opamp, it's a power amp - and loudness is relative.

5. Finally, isn't it a bit ridiculous to be using a puny op amp for such an application? When the OP posed a question about a PA system, I ass-u-med that they were asking about a proper PA, one that could actually be used to, well, address the public. This clearly won't do it; how come nobody has pointed that out?

As I mentioned way back at the beginning of this thread.

 

[carbonzit]

It's not an opamp, it's a power amp - and loudness is relative.

Whoops, my bad: I saw the op amp symbol and ass-U-med that's what it was. A '386 is indeed an audio power amp.

(Still not what I would use for anything I'd call a "PA" ...)

 

[audioguru]

1. Regarding that schematic: you didn't draw that, right? It looks a lot like the National Semi style of schematics one sees in books, etc. (I think it's National Semicircular, anyhow.) I happen to like that style of schematic drawing (planning on starting a thread about schematics here soon).

I make and modify schematics with Microsoft Paint program.
I copied and pasted part of the LM386 from its datasheet and added my micophone biasing.

2. In that circuit, a couple of things bug me. First of all, what's the purpose of that 1K resistor and 100uF cap there on the left? Power-supply decoupling? Why not just leave them out and run the 10k resistor right from V+?

You might be able to leave the filter out if the 9V is regulated but with a little 9V battery and its current jumping up to 150mA peak (when the amplifier outputs 0.45W and makes 0.5W of heat) then the battery voltage will be jumping with the music and the LM386 will amplify the jumping which will cause motorboating (putt, putt, putt).

Also, what's the purpose of the small (,047uF) cap and 10-ohm resistor on the output? Shouldn't the 8-ohm speaker load be sufficient (through the coupling cap, of course)?

The RC is called a Boucherot cell (also called a Zobel network) and stops the output from oscillating when the inductance of a speaker causes it to be a high impedance at high frequencies. The datasheet for the similar LM380 describes ringing on half the waveform when the load current is high and the RC stops the ringing.

3. Strictly speaking I wouldn't say that the electret is "biased" by the 10K resistor: it's actually powered by it, right? AFAIK, electrets (usually) contain a FET amplifier, no?

Yes, the 10k resistor powers the Jfet impedance converter (not much if any voltage gain) inside an electret mic).

4. Interesting that you say it's loud enough to have to turn down the volume. Those op amps are pretty amazing devices. But again strictly speaking, shouldn't we be talking about the output in terms of power (watts, or at least milliwatts), and not in terms of voltage, since we're using this as a power amp?

The LM386 is not an opamp, it is a little audio amplifier with a voltage gain of 20 to 200 and even more with some extra resistors and capacitors.
Its open-loop gain is very high so it uses a lot of negative feedback to reduce the gain to a useable amount, reduce distortion and reduce its output impedance to about 0.04 ohms or less. Because its output impedance is so low then its output is a voltage source that does not change when the load impedance changes. The very low output impedance also damps the resonances of a speaker very well.

5. Finally, isn't it a bit ridiculous to be using a puny op amp for such an application? When the OP posed a question about a PA system, I ass-u-med that they were asking about a proper PA, one that could actually be used to, well, address the public. This clearly won't do it; how come nobody has pointed that out?

An LM386 makes too much sound when it is a PA system and the apartment complex (flats) that has rice-paper walls. My home is very different ands it might not be heard in the next room if the doors are closed.

 

[p.gholap93]

You have something wrong with your circuit. Post your schematic.

I have an LM386 with its gain set to 200 with a 10uf capacitor between pin 1 and pin 8. My electret mic is biased with a 10k resistor from the positive supply and is coupled to the LM386 input with a 0.22uF coupling capacitor feeding a 100k volume control. It is so loud that I turn down the volume control.

So this is my final circuit diagram I am currently working on,it is picked as it is from datasheet.I am using a condenser mic and I also attach its picture.I have biased mic with 1k resistor and have connected 1nF capacitor in shunt with it.
What is the role of bypass capacitor at pin no.7?Also can you please explain why have you connected the capacitors and resistive network at pin no.2?Is it meant for filter purpose?

 

[p.gholap93]

**broken link removed****broken link removed**

So this is my final circuit diagram I am currently working on,it is picked as it is from datasheet.I am using a condenser mic and I also attach its picture.I have biased mic with 1k resistor and have connected 1nF capacitor in shunt with it.
What is the role of bypass capacitor at pin no.7?Also can you please explain why have you connected the capacitors and resistive network at pin no.2?Is it meant for filter purpose?

 

[audioguru]

So this is my final circuit diagram I am currently working on,it is picked as it is from datasheet.I am using a condenser mic and I also attach its picture.

You do not have a condenser mic. You have an electret mic. It should have only 2 wires with its ground wire connected to its metal case. Its output will be very weak if you connect its polarity backwwards.
A 3-wires electret mic needs to be connected differently.

I have biased mic with 1k resistor and have connected 1nF capacitor in shunt with it.

Your circuit is missing a power supply voltage, a supply bypass capacitor (470uF in my circuit), a filter for the bias current to the mic (1k resistor and 100uF capacitor in my circuit) and a very important input coupling capacitor (220nF capacitor in my circuit).
Since your volume control is only 10k ohms then your input coupling capacitor should be about 2.2uF and then it will be polarized so its (+) wire should connect to the mic.

The resistor to the mic should be 10k ohms (not 1k like you have) if the supply is 9V and should be about 4.7k ohms if the supply is too low at 5V.

What is the role of bypass capacitor at pin no.7?

It reduces hum if a cheap AC-DC power supply causes hum.

Also can you please explain why have you connected the capacitors and resistive network at pin no.2?Is it meant for filter purpose?

I explained it in my last post.