Hi everyone,

I am looking for a good way to make a voice amplifier circuit using LM386 that ensures low noise at the amplified voice (output).

Regards,

 

Hello there

using LM386 that ensures low noise at the amplified voice (output).

Ha.Ha.Ha.Ha. pull the datasheet under application notes you'll find many circuits adapt one or copy it straightforward and then put it on a breadboard you're going to love it.

 

What are you going to use for the input? There are lots of electret microphone amplifier circuits using LM386 on the internet.
What are you concerned about; background noise or electrical noise?

 

What are you going to use for the input? There are lots of electret microphone amplifier circuits using LM386 on the internet.
What are you concerned about; background noise or electrical noise?

both

 

Is the noise you are concerned with caused by high frequency hiss from the amplifier? Or from acoustical feedback howling? Or from background sounds?

The LM386 is old and cheap so it produces some hiss noise when its gain is high.
If you filter out the high frequency hiss then you also filter out the important consonant sounds of speech then words sound like vowels, grunts and groans. Listeners will say, "What? What did you say?" over and over.

 

Another kind of noise is severe distortion produced by the amplifier clipping or an improperly powered electret mic that is clipping.

 

I bought the LM386 kit and the LM1875 kit I hope that LM1875 will make a good second stage.
These projects are not the highest priority. This is where I am studying Audio amplifiers in more depth.
In years past I did'nt approach RF amplifiers in a traditional way but was sucessful with MMIC boards.

I am not sure which software to test the audio, I reviewed Virtins Muti-Instrument and Analog Discovery.
I expect that I might need to turn the volume down on the LM386 40 to 50 %
I have built about 10 different audio pre-amps in 2020 which gives me something to compare.
If the amplifier reproduces flawlessly I will mount it in an enclosure with BNC port for channel A
Channel B for experimental board for op amps such as NE5532A. The feeltech 6900 function generator
connects channels in order to compare opamp effects on one channel along with the straight LM386 on other scope channel.
It uses a relay to switch back and forth from scope to the LM1875 attenuator or listen via speaker output.
At least that is what I think might save time and allow quick disconnection of the function generator and scope.

 

If you are serious about the fidelity of the audio that you are amplifying, use the LM4558 op-amp for your pre-amplifier stages. It has the same pin-out and similar spec to the much more common LM358 but It has much lower noise and distortion figures..

 

An LM386 is a power amplifier, not a preamp. Use an ordinary low noise audio opamp for a preamp.

 

I guess I will have to rethink LM386 as a vocal pretone preamplifier. I am sorry AG, I appreciate your feedback.
I have several op amp pcbs that came with NE5532 I want to gut all the parts on those boards actually,

I have the LM4558 I wanted to change out other parts also. The original purpose was for low noise was ELF.
The intended purpose for vocal using LM386 was simplicity 0.08 % THD but that would not eliminate the noise.
For vocals I had ordered LM833 the gain I estimated to be slightly more than unity gain for parts of the 3-4kHz range
and slightly lower in other parts of the frequency range. Possibly more effective filtering at preamp level
using low noise op amps will improve quality instead of introducing power. The high fidelity is how well the singer feels about the performance.
I don't understand profesional sound. I have a few sound files and need to enhance the vocal for just one song.
I think possibly getting the singer to set an equalizer will give the response that is needed.