The board in question uses 25 ADS1278s in five chains of five chips. Please send me an email and I can share schematics.
Also, I should clarify that we are actually measuring intracardiac electrograms, not ECGs from the skin surface. The signal characteristics are basically the same as for ECGs, but about an order of magnitude larger (say around 10mV instead of 1mV).
With regards to the scope captures, this problem has so far only occurred when the device has been in use at an external test facility where we are unable to hook a scope up. I do however have screenshots of the clean and the “blocky” signals as displayed in our software program after some additional signal processing operations. See images below. The full scale for each trace in each image is 20mV and the pulses are occurring at a rate of approximately 2.5Hz. The screenshots are taken about 15 minutes apart (blocky occurs first) and there was no change in the ADC configuration or signal processing during that period.
I’ve been doing some reading on sigma delta modulations and have come across the concepts of limit cycles and idle tones. I’m still working on properly understanding them, but I was curious if you’ve ever come across them before and if either or both of them could result in the symptoms seen in the block example image below?
Normal: Relatively smooth transitions.
Blocky: The signals seem to have much more jagged transitions and appear more “square.”
