Identification method of error sources of A/D measuring chain

Abstract

Measurement data in modern measuring instruments and systems is obtained using a chain which typically consists of an instrumentation amplifier, sample-and-hold circuit and an analog-to-digital converter. These elements introduce errors which influence uncertainty of the output results. To evaluate this uncertainty one must determine an error model of the chain and distributions of particular errors. The paper describes an identification method of the A/D chain errors. The method consists in operations of two kinds. First, using a function generator, one measures the output (resultant) error distribution in selected conditions. Next, probability density functions of partial errors are determined basing on the obtained resultant error distribution and using the relationship between errors given by the error model. In order to verify hypotheses concerning the shape of distribution, relations between uncertainties defined as parameters of error value sets are used. The method has been applied to determine selected errors of an A/D conversion card.