Three different MCMs, three different test strategies
Andrew Flint · 2002
This paper looks at three quite different MCMs and presents, for each design, how and why a test strategy was chosen. The first MCM examined is a very complex design implementing a RISC subsystem including a processor, cache controller, and large memory. This design has the complexity of a circuit board, including the requirement to support repair of failing components, so a test strategy closely paralleling a board test strategy was chosen. The second MCM is far simpler: four identical ASICs are mounted in one small package. The test strategy adopted was clearly a chip test strategy. The final design, an automotive engine control application, dictated a somewhat hybrid approach-a limited functional test implemented through a tester assisted self-test.