|
|
In the presence of an imperfect transmission "channel," there are physical and mathematical reasons that random jitter sometimes increases with increasing length PRBS test patterns. Essentially, the nature of the any inter-symbol interference (ISI) is such that the behavior of the total jitter (Tj) diverges significantly from the "dual-Dirac model". We rely upon that model for the basic deterministic and random jitter (Dj, Rj) breakdown. For multiple reasons, often the increase in observed Rj is not only apparent but can be verified by BERT measurements, and mathematically justified. The effect is larger than many engineers suspect. For that matter, one can ask whether long PRBS patterns serve as good test case at all.
|