Analog fault diagnosis: a fault clustering approach

A novel analog circuit fault diagnosis method is proposed. This method uses a neural network paradigm to cluster different faults. It is capable of dealing with the common fault models in analog circuits, namely the catastrophic and parametric faults. The proposed technique is independent of the linearity or nonlinearity of the circuit. The process parameter drifts and component tolerance effects of the circuit are well taken care of. Several fault diagnosis strategies for different problem complexities are described. The proposed methodology is illustrated by means of an operational transconductance amplifier (OTA) example