How to go about reducing noise in a power supply

Hardware
1

I’m using a DC-DC Step Down Buck Converter to try and power a photon. But I’ve found the converter suppiled power is introducing noise in one of the sensors being used in my project. Specifically it’s occurring with a sound sensor. If I power the Photon with the converter the background sound/noise value is 150 but if it’s powered by a battey (i.e. a clean power supply) the background sound/noise value significantly reduced to 50.

I understand that capacitors can be used in some way to help here. But am having trouble understanding the details of how to go about doing this. I’ve been reading capacitor examples trying to get some clarification but am kind of lost.

Would I be looking to create a decoupling capacitor? If so how would I know what value capacitors to use? The input power to the buck converter is from an AC to DC adapter and is 27 volts at 55ma.

Is this somehow helpful? LM2596 pdf specs. On page 11 is shows what looks like a table of capacitors for various voltages in and out.

Thank you.

2

Try adding a 0.1uF and 10uF to filter power supply noise from low to high frequencies.

3

I guess they would go across the load? And the voltage does not really matter as long as it’s above 5 since that is what the Photon VIN is?

Buck Converters: http://www.learnabout-electronics.org/PSU/psu31.php

4

Yup correct. Don’t exceed the Vin limit though :slight_smile:

5

@JoeT, switching regulators have ripple by design. As such, power supply sensitive sensors may be affected, even with decoupling caps. Which sound sensor are you using?

6

SparkFun Sound Detector

‘The envelope output allows you to easily read amplitude of sound by simply measuring the analog voltage’

This envelope pin has a higher base value, with no sound present, when the power is coming from the buck converter. I’ve been assuming that a higher base reading means a loss in the ability to detect lower volume sounds. But maybe I’m wrong about that.

Because of the way I’ve soldered my project together, and attached it to my garage ceiling, it’s a pain to take down, and start doing some testing. So I’m just going to buy another converter and do the capacitor experimentation with that one. So in a couple of days I should hopefully have some results to pass on.

7

@JoeT, the microphone circuit uses a high gain (100) input stage with a resistor divider providing a voltage reference for the op-amp used. This is standard design. However, because Vcc is used as the supply, any ripple on Vcc will translate into the input of the op-amp and get amplified. Furthermore, the amplifier stages don’t provide any filtering to reduce high frequency noise effects. The ideal scenario would be to use a linear regulator to avoid switching noise. However stepping down from 27V to 5V is a pain. Let me know how the filter caps work (or not). :wink:

1 Like