A parallel testing algorithm for pattern sensitive faults in random access memory
[Thesis]
S. Lu
M. Ashtijou
Texas A&M University - Kingsville
1994
154
M.S.
Texas A&M University - Kingsville
1994
A pattern sensitive fault recognition algorithm (PSFRA) is presented that not only detects and locates the neighborhood pattern sensitive faults (NPSF) but also it recognizes the types of the fault. A parallel testing structure for NPSF in Random Access Memory (RAM) has been developed in (1) to reduce overall test time and cost. It is designed for a broad class of pattern-sensitive faults and is significantly more efficient than previous approaches. The algorithm is modified to provide a fault type recognition capability that helps in differentiating between active, passive, and static NPSF for Type-1 and Type-2 neighborhood. The PSFRA needs 321 N write operations and 480 N read operations for Type-1 neighborhood, and it needs 129 N write operations and 192 N read operations for Type-2 neighborhood (N is the memory size). The recognition of fault is in expense of more test time and it is still more efficient than sequential algorithms.