DFT Digest

When I last left off in this series, we were talking about the broadcasting of scan data from a few external scan chains across many internal parallel scan chains, and some of the ways it is implemented. Not too much detail, but enough to present the concept. So far, all high-level DFT.

Well, once the data has propagated through these much shorter chains, the data must be compressed back into the width of the external scan data paths. Like in the case of decompressing the data, there are a couple of alternatives here, some combinational, some sequential. The combinational implementations are XOR-trees, and the sequential implementations are as you might expect, some sort of MISR. Each has it’s advantages and disadvantages.

Some of the concerns in compressing data at the end of the internal chains are how to avoid aliasing (a common problem when using XOR-tree type structures) , how to prevent unknowns from masking good data, and how to be able to debug when failures occur. Again, this is where the hardware and software team up. The first two problems, if not fully addressed, affect the efficiency of the pattern generation, In other words, you won’t get the TDV compression you think you should.

The third problem, in my mind is worse, because it’s a real time waster. Ultimately, when faced with failures on the tester, one takes the datalogs, reads them back into the tool that generated the vectors, and the tool spits out the failing scan cells. If the tool is not that capable, it may require using a bypass mode, if implemented, to find the failing cell(s).

A bypass mode reconfigures the internal scan chain into one or more external scan chains by daisy-chaining them. But as you may have already guessed, you have to now maintain a separate set of patterns, which don’t necessarily correlate well with the normal patterns, and as such, may not fail at all, even though the normal patterns do!

When we visit next in this series, we’ll try to put it all together, and explore the compression results, and what they mean to test time and memory.

Are we having fun yet?