Search

This paper proposes an innovative method for Integral Nonlinearity (INL) testing, that allows to measure the voltage corresponding only to a subset of all the codes of Digital-to-Analog Converters (DAC), shortening the duration of static characterization. The method operates, in fact, a pseudo-random selection of DAC codes to be tested but, through Compressed Sensing, it is able to recover the INL value for all the codes. The proposed method is experimentally evaluated on two DACs with different architectures and thus with different trends of the INL curve. In the performed tests, the recovered curve results very close to the INL curve obtained by the standard approach, while allowing a considerable time saving. • Innovative method of Nonlinearity testing for unknown architectures. • Reduction of DAC input codes in static characterization. • Compressed Sensing for reconstruction of Integral Nonlinearity curve. • Time saving in comparison to standard procedure. • Experimental validation on different DAC architectures. [ABSTRACT FROM AUTHOR]