Audio Signal Processing

Designer_Big

Audio Signal Processing [Copy link]

 

Hello everyone

I plan to use this circuit to amplify the audio signal collected by the microphone. The frequency is 100~16KHz, the highest two-stage amplification gain G=500, and the gain-bandwidth product of AD8542 is 960KHz

Power supply is 5V, providing 2.5V DC bias

My problem is that when I build a circuit on a breadboard for testing, the waveform captured on the oscilloscope is not an audio signal, or in other words, the captured signal does not match the following (Internet picture)

I generate audio signals by playing music on my phone

ps: After testing, it can normally amplify 20mVpp~1Vpp sinusoidal signals

maychang

[My problem is that when I build a circuit on a breadboard for testing, the waveform captured on the oscilloscope is not an audio signal]

If your audio signal is generated by playing music on a mobile phone, then the waveform on the oscilloscope should be roughly the same and should look like this.

maychang

[ps: After testing, the sine signal of 20mVpp~1Vpp can be amplified normally]

This is unlikely. The supply voltage of your op amp is 5V, so the output amplitude of the op amp cannot be greater than 5V. If the input signal is 20mVpp and the amplification factor is 500 times, then the output should be 10Vpp. If the input signal is 1Vpp, then the output should be 500Vpp.

Designer_Big

maychang posted on 2023-2-26 12:19 [ps: After testing, it can normally amplify 20mVpp~1Vpp sine signals] This is unlikely. Your op amp supply voltage is 5V, so the op amp...

The gain is adjustable, and since I couldn't find a variable resistor in the LT, I chose the parameter for maximum gain.

In other words, my circuit is fine in theory, but the actual result is wrong. It is caused by the introduction of unknown noise when I built it.

Designer_Big

Is it inappropriate to use a breadboard to build this type of audio processing circuit? Will it introduce noise?

maychang

Designer_Big posted on 2023-2-26 13:11 Is it not suitable to use a breadboard to build this kind of audio processing circuit? Will it introduce noise?

The bandwidth is only 16kHz, no problem with the breadboard.

If noise is introduced, it is probably caused by your input leads (the connection line from C4 to the signal source and the ground line). In addition, the decoupling capacitor of the op amp power supply (this is not shown in the electrical schematic) is far away from the op amp, or even not there at all, which will also introduce noise if the electromagnetic environment is not good.

maychang

Designer_Big posted on 2023-2-26 12:36 The gain is adjustable, because I couldn't find a variable resistor in LT, so I chose the parameter with the maximum gain. In other words, my circuit is theoretically OK...

Variable resistors are often large in size and easily introduce noise. If the electromagnetic environment is not very good, especially if the connecting wires from the variable resistor to various components are long, it is even easier to introduce noise.

maychang

Designer_Big posted on 2023-2-26 12:36 The gain is adjustable, because I couldn't find a variable resistor in LT, so I chose the parameter with the maximum gain. In other words, my circuit is theoretically OK...

Whether noise from the environment has been introduced requires careful analysis and cannot be judged solely by the output waveform.

Short-circuit the input. If the oscilloscope shows a straight horizontal line, it is most likely not noise.

In addition, when testing your amplifier circuit, the signal source should be a signal generator, not music played from a mobile phone.

maychang

Designer_Big posted on 2023-2-26 12:36 The gain is adjustable, because I couldn't find a variable resistor in LT, so I chose the parameter with the maximum gain. In other words, my circuit is theoretically OK...

Your waveform does not represent environmental noise. If it is a music signal, it is normal to see this waveform on the oscilloscope.

Gen_X

maychang posted on 2023-2-26 16:03 Your waveform does not indicate environmental noise. If it is a music signal, it is normal to see this waveform on the oscilloscope.

It should be normal indeed! Music is like this too, and you can only see the noise clearly by zooming in on it.

wangerxian

The waveform seems to be the same when the oscilloscope is directly connected to the positive and negative poles of the speaker.

xutong

Designer_Big posted on 2023-2-26 12:36 The gain is adjustable, because I couldn't find a variable resistor in LT, so I chose the parameter with the maximum gain. In other words, my circuit is theoretically OK...

Maybe introducing the .param directive will solve the problem?