I am looking at the MPC5777 micro, specifically the built in sigma-delta ADC (SDADC). The datasheet has offset and gain error values on page 26. As I interpret these specs, it seems to imply that even with the built in calibration completed I could still expect many millivolts of offset and gain error depending on how much temperature, Vdda, Vrh_sd, etc vary or drift from the moment calibration was done last. Is this interpretation correct?
For my application I need a high degree of DC accuracy. The signal I am trying to measure is a small amplitude and slow varying signal but requires accurate DC measurement where the many millivolts of offset and gain error in the spec sheet are not acceptable. I don't need to do anything fancy with the ADC data, I just need to measure a signal to see if it is above or below a threshold but I need to know this with relatively high DC accuracy.
Can I run the SDADC calibration sequence in real time, say every second or two, such that I can constantly calibrate out the offset and gain error drift due to temp, voltage, etc? Or is the calibration something that shouldn't be done in a real world running environment with actual analog signals coming into the SDADC?
Assuming I can run calibration every few seconds to constantly eliminate any DC error in the SDADC, how accurate is the SDADC after I complete a calibration? Can I assume there is effectively no offset or gain error at the moment calibration is complete?
Does NXP have a spec for how much offset and gain can drift per degree C of temp change (i.e. delta in mV/C)?
Any other thoughts or suggestions for high DC accuracy applications?