Diagnostic Test of Robust Circuits

Abstract

Robust circuits are able to tolerate certain faults, but also pose additional challenges for test and diagnosis. To improve yield, the test must distinguish between critical faults and such faults, that could be compensated during system operation; in addition, efficient diagnosis procedures are needed to support yield ramp-up in the case of critical faults. Previous work on circuits with time redundancy has shown that "signature rollback" can distinguish critical permanent faults from uncritical transient faults. The test is partitioned into shorter sessions, and a rollback is triggered immediately after a faulty session. If the repeated session shows the correct result, then a transient fault is assumed. The reference values for the sessions are represented in a very compact format. Storing only a few bits characterizing the MISR state over time can provide the same quality as storing the complete signature. In this work the signature rollback scheme is extended to an integrated test and diagnosis procedure. It is shown that a single test run with highly compacted reference data is sufficient to reach a comparable diagnostic resolution to that of a diagnostic session without any data compaction.

Built-in self-diagnosis targeting arbitrary defects with partial pseudo-exhaustive test

“…As shown in Figure 1 the P-PET architecture proposed in [9] is combined with the window-based diagnosis of [4] [5]. To generate a P-PET sequence for a given bound b, the multiple-polynomial LFSR (MP-LFSR) switches between several polynomials stored on chip, each one exercising a subset of I b = {I(o) | |I(o)| b}, such that overall each input set in I b receives all possible input combinations.…”

“…While the test continues normally with the next window, the shadow MISR runs in autonomous mode as long as the first pattern is applied. This way fault effects are distributed over the MISR randomly, and it is sufficient to observe only l bits of the shadow-MISR [5]. The observed bits are compared to the respective reference data stored in the response memory.…”

“…To support a unique solution, the dimensions of the system of equations must be properly set. As shown in [4] [5], the number of patterns in a window should be less than or equal to the number of bits in the MISR signature. A straightforward application of the scheme to long P-PET sequences would therefore result in an extremely large response memory.…”

“…As the sequences for P-PET are typically much longer than the pseudo-random sequences in LBIST or mixed-mode BIST, it is even more challenging to limit the storage requirements for response data. This paper shows how the diagnosis approach in [5] can be adapted to efficiently deal with P-PET sequences while maintaining the high defect coverage.…”