QUESTION:
What is meant by "pull-in" and "pull-out" range?
ANSWER:
The notion of pull-in range and pull-out range are terms concocted to describe the behaviour of the phase-locked-loop (or frequency-locked-loop) implemented. The scenario is that we have a loop controlling a local aoscillator as well as a reference that is allegedly good. Under normal circumstances, the loop would operate conventionally and the local clock would be "locked" to the reference or "slaved" to the reference. The pull-in range represents how far off the reference can be from "true" and yet the slave must follow. The pull-out range represents a situation where the slave must reject the reference because it is too far off from true (and raise an alarm!). G.813 (1996) sets 4.6 ppm as both the pull-in and pull-out values though the pull-out is still being debated. The debate is necessary because the intrinsic accuracy of the slave oscillator must be better than the difference of pull-in and pull-out ranges (so that it can accurately and reliably distinguish whether the reference is actually good or has gone haywire).
However, it should be noted that a pull-out range is actually a very useful requirement. Without it frequency errors can propagate very easily without any indication until everything just breaks down and we have a real mess.
|
