The authors present test algorithms for go/no-go and diagnostic test of memories, covering neighborhood pattern-sensitive faults (NPSFs). The proposed test algorithms are March based, which have linear time complexity and result in a simple built-in self-test (BIST) implementation. Although conventional March algorithms do not generate all neighborhood patterns to test the NPSFs, they can be modified by using multiple data backgrounds such that all neighborhood patterns can be generated. The proposed multibackground March algorithms have shorter test lengths than previously reported ones, and the diagnostic test algorithm guarantees 100% diagnostic resolution for NPSFs and conventional RAM faults. Based on the proposed algorithms, the authors also present a cost-effective BIST design. The BIST circuit is programmable, and it supports March algorithms, including the proposed multibackground one.

We present two memory test algorithms for neighborhood pattern sensitive faults (NPSFs), including static NPSF (SNPSF), passive NPSF (PNPSF) and active NPSF (ANPSF). March algorithms are widely used in memory testing because of their linear time complexity and ease in built-in self-test (BIST) implementation. Although conventional March algorithms do not generate all neighborhood patterns for testing the NPSFs, they can be modified by using multiple data backgrounds such that all neighborhood patterns can be generated. The proposed multi-background March algorithms have shorter test length than previously proposed ones (68N for detecting SNPSFs and PNPSFs and 96N for detecting all NPSFs). Also, based on the proposed algorithms, linear-time BIST circuit can be implemented with low area overhead.