Fault modeling and fault sampling for isolating faults in analog and mixed-signal circuits

Expensive specification testing of analog and mixed-signal circuits is currently being replaced by fault-oriented test techniques, similar to the ones used for digital circuits. However, the large number of possible failure modes in analog and mixed-signal circuits presents a major bottleneck in terms of fault simulation complexity and size of fault dictionary. For a given set of possible faults, it has been shown that only a small fraction of the total number of faults contribute to diagnostic accuracy. In this paper, we present a fault sampling methodology that identifies those faults by simulating only a minimal fraction of the entire set of possible faults. The proposed sampling algorithm is based on statistical fault simulation and leads to major reductions in fault simulation complexity for hierarchical analog and mixed-signal circuits.