When performing offset calibration for the AMC131M03-Q1 using internal registers, since each chip exhibits different offset voltages, the register values written to different chips should theoretically vary. However, the customer expects a unified calibration value that can limit the deviation within 5 μV.
Question 1:
Would it be feasible to obtain a unified calibration value by testing the offset voltages of multiple chips, deriving a normal distribution of these offsets, and then using the mean (expected value) of this distribution as the unified calibration value?
Question 2:
If feasible, how many chips should be tested for statistical distribution analysis to ensure reliability?
Question 3:
If not feasible, do you recommend any alternative solutions to achieve a unified calibration value?
