We present a new pattern generation approach for deterministic built-in sel
f testing (BIST) of sequential circuits. Our approach is based on precomput
ed test sequences, and is especially suited to sequential circuits that con
tain a large number of flip-flops but relatively few controllable primary i
nputs. Such circuits, often encountered as embedded cores and as filters fo
r digital signal processing, are difficult to test and require long test se
quences. We show that statistical encoding of precomputed test sequences ca
n be combined with low-cost pattern decoding to provide deterministic BIST
with practical levels of overhead. Optimal Huffman codes and near-optimal C
omma codes are especially useful for test set encoding. This approach explo
its recent advances in automatic test pattern generation for sequential cir
cuits and, unlike other BIST schemes, does not require access to a gate-lev
el model of the circuit under test. It can be easily automated and integrat
ed with design automation tools. Experimental results for the ISCAS 89 benc
hmark circuits show that the proposed method provides higher fault coverage
than pseudorandom testing with shorter test application time and low to mo
derate hardware overhead.