Parametric Fault Detection of Analog Circuits by Utilizing the Fundamental RMS of the Supply Current Fluctuation
Vassilios D. Vassios, Argyrios T. Hatzopoulos, Ioannis G. Intzes, Dimitrios K. Papakostas · 2024
An improved method for accurately identifying and categorizing parametric or soft errors that impact analog integrated circuits is presented in this research. This method relies on the monitoring of the fluctuations of the power supply current. New metrics for the good or faulty circuit under test are derived from these measurements. The mean and fundamental rms supply current values are calculated. From these calculations three points of interest are extracted. The under/overshooting, the damping, and the relaxation point. The algorithm's general behavior in hard fault detection was investigated in the author's previous works and this paper is an extension on the parametric-hard fault detection capabilities of the algorithm. The investigated faults affect only resistor and capacitor values. The algorithm was tested both in simulations and the results are presents. The obtained results show that every CUT that was tested has a high classification average accuracy. The proposed method has achieved a high degree of classification accuracy with significantly less complexity and processing power as compared to earlier research studies.