Evaluation of alternative LBIST flows: A case study

The cost of manufacturing test has been growing dramatically over the years. The traditional pseudo-random pattern based Logic Built-in Self Test (LBIST) can potentially reduce the test cost by minimizing the need for the automatic test equipment. However, LBIST test coverage can be unaccept-ably low for some designs. Various methods for complementing pseudo-random patterns to increase test coverage exist, but the combined effect of these methods has not been studied. In this paper, we evaluate the effectiveness of alternative LBIST flows by a case study on a real industrial design. Our results can guide the selection of the best LBIST flow for a given set of design constraints such as test coverage, area overhead, and test time.