I have a MEMS microphone which outputs PDM and is connected to an STM32U5. It's connected via the integrated Audio Digital Filter (ADF) peripheral which implements conversion from PDM including decimation and anti-aliasing filters.
I'd like to input some test stimulus to the ADF to validate performance as I have it configured. For some of my testing I'd like to exclude the microphone part of the chain and test just the DSP. In other words, I don't want to be limited to acoustic stimulus.
Is there a common practice way to do this?
One idea I had was to replace the microphone with a 2nd MCU programmed to be a PDM generator. I could bit bang output the PDM data values on clock line edges. The PDM encoding algorithm looks simple enough to implement and I wouldn't need to store full test waveforms in memory since I expect all my test signals could be generated algorithmically (impulses, white noise, sinusoids, etc)
Is there an easier option? I looked for an off-the-shelf device for this application but found nothing.