Hello, sorry for late answer,
we have some mechanisms that might help there: post processing filtering (FIR or IIR) might work for you. When the filtering finishes, the new result should be compared against the boundary. Filtering is limited to 2nd order max. If you want to do more, you would need to use SW, but then boundaries are not directly linked. Also possible to accumulated via HW (up to 4 times accumulation) and then shift for averaging.
On HW side, you could use a small C or RC next to ADC pin to clean up the signal. Another idea: you can play with sampling frequency and sampling time. Depending on your noise characteristic, sampling faster or slower might help you getting better results: for example if you have spurious spikes eventually, sampling time tuning might help filtering out those.
Regards,
P Yelamos
we have some mechanisms that might help there: post processing filtering (FIR or IIR) might work for you. When the filtering finishes, the new result should be compared against the boundary. Filtering is limited to 2nd order max. If you want to do more, you would need to use SW, but then boundaries are not directly linked. Also possible to accumulated via HW (up to 4 times accumulation) and then shift for averaging.
On HW side, you could use a small C or RC next to ADC pin to clean up the signal. Another idea: you can play with sampling frequency and sampling time. Depending on your noise characteristic, sampling faster or slower might help you getting better results: for example if you have spurious spikes eventually, sampling time tuning might help filtering out those.
Regards,
P Yelamos