Test points selection process and diagnosability analysis of analog integrated circuits

Fault diagnosability analysis is an approach to enhancing the diagnosability of a circuit. Whereas the design for diagnosability ensures that the test points are properly selected and the generation of diagnostic tests is considerably simplified, diagnosability analysis is used to locate sections of a circuit having poor diagnosability. The information allows estimation of a circuit's diagnosability before the fault diagnosis is attempted. Hence any potential problem can be located early on the design phase, allowing modifications to be introduced to improve the final diagnosability. This paper presents a simple diagnosability analysis process in which the diagnosability of a circuit is measured from a graph that describes the circuit topology and a given set of test points, where no circuit simulation is needed. In addition, a simple algorithm is also presented to select a minimum set of test points and their locations for achieving the desired diagnosability.