Study On Analog Fault Diagnosis Based On Circuit Simulation

With the rapid development of large scale integrated circuits, the reliability of the circuit is highly demanded. Thus, diagnosis of the circuit boards is becoming more and more important. However, research in the areas of analog circuit fault simulation and test generation has not achieved the same degree of success as its digital counterpart owing to the difficulty in modeling the more complex analog behavior, the allowance of component parameter, the lack of accessible test points, etc.This thesis is to give a brief discussion of the test and diagnosis methods of analog circuits. Considering that traditional functionality test is limited to the range of the circuit and hard to get the repository, it is comparatively easier to use circuit simulation to help us diagnose analog fault, with the advantage that we can make intelligentized analysis of the simulation results.So, this theis proposes the simulation based method of diagnosing analog circuits, discusses its advantages and then gives a clear flowchart of this method, as well as the key principles among every stage, including theory of simulator, fault modeling process, test point selection, and fault dictionary construction. We choose SPICE as our simulation engine, which is developed by UC Berkeley. Later part of this thesis gives a detailed discussion about the 3 steps of analog fault modeling: defect simulation, formation of the fault model and verification of the fault model. Then, it proposed 3 methods and 3 strategies in test point selection stage. Three methods are: inclusion method, exclusion method and sort method. The combination of exclusion and sort method could select the minimum test points set, which is demonstrated in the thesis.The last chapter is composed of several technical implements of our SIMULATION PLATFORM, including automatic netlist generation, fault injection, fault simulation, etc.