A Novel Implicit Parametric Fault Detection Method for Analog/Mixed Signal Circuits Using Wavelets

Abstract Analog ICs in general have many performance metrics that depends on multitude of device parameters. Parametric yield loss is often a problem in analog IC designs, where a significant portion of ICs manufactured fail specification test due to variation in process parameters. Parametric faults are mainly caused by statistical fluctuations in the manufacturing process. It causes the analog circuit’s performance parameters to deviate from their specified limits. In the conventional testing techniques used for testing analog and mixed signal circuits, the fault-free and faulty circuits are classified according to the satisfaction and violation of their specifications. In this work, fault model based testing technique is used to test analog and mixed signal circuits. A fault list is constructed based on fault models. Then test signals are applied to the circuit under test and the output response is measured .It is compared with the fault-free response. Circuits are classified based on the comparison results. Faults are detected using MAHALANOBIS distance metric. The proposed technique is validated with an eight bit R2R Digital to Analog Converter (DAC) circuit.