BeagleBone Black ADC signal drift FYI

I am posting this solely in case it helps someone else. There’s no question to be answered, but comments are fine.

Our problem was that, when reading a 0-1.6V DC signal with the ADC on AIN0, the signal drifted up quickly after the ADC was started using PRU0.

The issue was that our signal source was too high impedance. This also produced some noise as well.

We solved this by running the signal through a unity-gain buffer amplifier already available on our custom board, which lowers the signal source impedance. Now we don’t have the drift and our noise is much, much lower.

Here is the transcript from the TI.com session which has some links and more details.

The errata mentioned in TI’s response explains the best practice for unused ADC channels. We could not easily tie our unused ADC inputs to AGND as indicated so we had to cut the runs and leave them open.

AM3358_ Signal drift when reading with ADC - Arm-based microcontrollers forum - Arm-based microcontrollers - TI E2E support forums.pdf (320.3 KB)

1 Like

That’s good news!

We really need more people taking time to feed information like this back to the community.
Thank you for taking said time!

You are welcome. I am the founder of a startup and do a lot of development alone so this group has been very helpful over the years. I want to give back when I can.