To recover the analog voltage at the input of an
analog-to-digital converter (ADC) from its digital output, one
needs to know at least the ADC gain and offset. For high-accuracy
measurements, it is necessary to estimate the actual values of
these parameters since they are usually different from the ideal
values (one and zero, respectively). This estimation inevitably
has an uncertainty, which contributes to the uncertainty of any
measurement made with the ADC. Here, the precision of gain
and offset error estimators, based on the histogram method for
ADC testing is analyzed. The “terminal based” and “independently
based” definitions are compared, both through simulation
and experimental evaluation. Our conclusion is that, in typical
conditions, the “independently based” definition is more precise.