Gas station without pumps

2020 October 28

Analog Discovery 2 power-supply noise

Filed under: Circuits course — gasstationwithoutpumps @ 11:38
Tags: , , ,

Last night and this morning I spent some time investigating the noise on the power supplies of the Analog Discovery 2, because some students were having trouble with power-supply noise on their audio amplifiers (an inherent problem with biasing the microphone with just a bias resistor to the power supply).

I looked at the positive supply set to +3.3V using oscilloscope Channel 1, and saw a fluctuation in voltage that was not too surprising for a switching power supply (though the switching frequency seemed ridiculously low).  The power supply is specified to stay within 10mV of the desired voltage, and the voltage seemed to be doing that.

I know that some switching power supplies shut themselves off under low-load conditions, to retain efficiency at the cost of adding low-frequency ripple to the output, so I tried running the power supply with different load resistors.  I did the sampling at 400kHz and took FFTs of the signal (exponential averaging of RMS with weight 100, Blackman-Harris window).

Here are the signals:

The signals show quite a bit of oscillation without a load, but decreasing with increasing load.

Here are the spectra from the Fourier transform (removing the DC spike):

The spike around 57.2kHz is present with all loads and remains at the same frequency even if I change the sampling rate, so is probably the underlying frequency of the switching power supply.

The rather large fluctuations in the audio range are probably the result of the power supply shutting itself off when there is low current draw.  Drawing 10 mA is not quite enough to prevent this shutdown, but 27.5mA seems to be enough.

So there seem to be at least three solutions for students having problems with power-supply noise:

  • Taking enough current from the power supply that the power supply doesn’t shut itself down.  This is a rather fragile technique, as other sources of power-supply noise (such as noise injected by the power-amplifier stage in a later lab) will not be eliminated.
  • Using a transimpedance amplifier instead of a bias resistor to bias the mic.  The bias-voltage input to the transimpedance amplifier can have a low-pass filter to keep it clean.
  • Putting a low-pass filter (with a small resistor and large capacitor) between the power supply and the bias resistor.  The resistance of the filter adds to the resistance for the DC bias calculation, but not to the resistance for the i-to-v conversion.

Leave a Comment »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: