Testing Analog Circuits by PCA of Power Supply Current

Abstract

A method using principal component analysis (PCA) of dynamic power supply current was proposed for testing of analog circuits in this paper. The basic model of the proposed method and the general rule for analog fault detection were described in detail. At first, the principal component model of fault-free circuits was constructed. Then the circuits-under-test was compared with the principal component model to calculate the statistic for fault detection. The features of power supply current in both time and frequency domain were combined by PCA, and it could overcome the difficulty to determine threshold by empirical knowledge. The proposed method was applied to detect faults of the signal filtering and amplifying circuit, which is used in the ultrasonic liquid-level sensor. The results show that the power supply current contains information about the circuit’s faults, and can be used for fault detection of analog circuits by analyzing this signal.