ADC test methods using an impure stimulus: A survey

Abstract

Testing high-resolution ADCs with impure test signals can be done by applying first the original test signal to the ADC, and then the same signal with a DC offset. The two output data sets then have a lot of redundancy that can be used to calculate the ADC's performance parameters, e.g. INL and DNL. While there are a number of algorithms and proposals to perform such tests, they are until now only very rarely used by ATE vendors or ADC IP suppliers. This is astonishing since they can substantially reduce production test costs with little effort. This paper summarizes the state-of-the-art of the different methods, depicts their possibilities and drawbacks, and tries to answer why all of these methods are so far from being widely used.