Hello Bob,
Thanks for the response.
I'm running fclk at 2MHz, and SCLK at 200kHz.
As far as we can tell, the part still operates fine, it just runs at a new rate. We can configure the A/D without a reset. The way our software operates, though, our customers tend to issue a reboot command to clear it up.
I agree that this is transient related. We have several customers reporting this that are using our device to measure parameters on jet engines. That's a very noisy environment that I can't duplicate in my lab.
One customer reported that putting ferrite chokes on the inputs reduced the events significantly. His test cell was shut down for a time and the building power was updated. New transformers, etc. After that, he had no more problems.
This is one of the strangest problems I have ever encountered. The FF is not a power on default value. We have more than one user reporting the same change to the same register. It took quite a bit of sleuthing to figure out what was happening inside the A/D, because the events are so rare.
Thanks,
Chris