The system needs a way to listen to the sound around it and the natural way to do this is with an electret microphone. I will probably include a 3.5mm jack input to allow you to plug music straight in too.

I was looking online for some guidance on designing the circuit and I found a good reference design from Texas Instruments. I followed through the steps that they took in coming up with the design. This design is basically what I need so I am copying it almost exactly (hey, that’s what reference designs are for) except that I am replacing the TI op-amp for an NE5532 (my go-to op-amp), and the 75kΩ resistor with a 200kΩ potentiometer so that I can vary the gain.

I built the circuit up on a breadboard.

IMG_0673.JPG

Power-supply-noise-coupled
The audio signal in red, and the power supply noise in blue.

The red trace shows the audio signal after it has been amplified and it looks pretty clean. However, the power supply has a very repetitive ~45mV pulse on it every ~600µs and this is getting coupled through to the audio signal. You can see that every time there is a pulse on the power supply (blue), there is a little bit of noise on the audio signal (red).

It’s not clear right now how much of an issue this will be. As the pulse is occurring with a frequency of ~1.5kHz, I’m worried that the 1kHz or 2.5kHz filters will amplify the noise and that this will mess with the output. This can be fixed by adding a filter to the power supply or by getting a cleaner power supply. Rather than waste time trying to fix this now, I’m going to wait to see if this is actually an issue.

Advertisements