Global parametric faults identification with the use of Differential Evolution

This paper presents a novel method to a multiple parametric faults diagnosis (global parametric faults - GPF) in analogue integrated circuits (AIC). The method is based on features of AIC time domain response to voltage step excitation, i.e. AIC response and its first order derivative maxima and minima locations. A circuit states classification is acquired with the use of linear evolutionary classifier which parameters are determined with the use of Differential Evolution. Selected AIC response features distributions are approximated with geometric figures based on polynomial functions. The proposed diagnosis method has been applied for a GPF diagnosis in an exemplary integrated circuit - operational amplifier muA4741.