| Due to rapid advances in communications and high speed integrated circuits, there has been a resurgence of analog and mixed-signal circuits and systems. However, research in the area of analog and mixed-signal test has not achieved the same degree of success as its digital counterpart. The lack of suitable analog fault models has been the prime reason for restricting the problem of analog test to the functional domain, thereby precluding a structured solution. In this dissertation, a comprehensive test framework for analog and mixed-signal circuits, which entails a three step process consisting of fault modeling, fault simulation and test generation, is developed. A hierarchical fault modeling methodology based on functional error characterization is used to derive comprehensive analog fault models, which permit the behavioral representation of faulty circuits. The fault-free and faulty circuits are transformed to the discrete Z-domain, giving rise to an efficient fault simulation technique. The fault simulator forms the core of an integrated test generation system. This automated fault-based technique generates efficient, cost-effective tests and provides a quantitative measure which is lacking in traditional techniques. Since test generation is performed at the behavioral level instead of the circuit level, it can handle complex circuits. This approach also forms the basis of a unified framework for mixed-signal testing. The presence of continuous-valued analog signals alongside binary digital signals is handled by the discretized representation of the analog circuit along with the digital circuit in the Z-domain, thereby enabling simulation of the hybrid circuit. In addition to automatic test generation, this work also includes a built-in self test scheme for analog and mixed-signal circuits. The scheme comprises of a multisine test input and a novel signature analyzer for test response compression of the imprecise analog outputs. |
