A fuzzy approach for soft fault detection in analog circuits

An automatic fault diagnosis technique based on a fuzzy approach is presented and its performance is analyzed by means of two examples. The proposed technique is applied to the detection and the isolation of single soft faults in analog electronic circuits. Both faults at component and sub-system level are considered. A ‘simulation before test’ approach is followed: a fault dictionary is a priori generated by collecting signatures of different fault conditions simulated in the frequency domain. A fuzzy classifier is trained by data contained in the fault dictionary and used to process the circuit under test response during test phase. The effects of noise and of non-faulty parameter tolerance are taken into account.