A discussion of methods for measuring low-amplitude jitter

There is an escalating need in current designs to control and measure the stability of critically timed signals, most notably clocks. The success of any design methods which attempt to ensure such stability hinge on the suitability and quality of the timing data upon which the design decisions are based. For clock signals, there are a variety of metrics of stability (peak, cycle-to-cycle, etc.). Furthermore, the widespread use of PLL devices for the distribution and reception of clocks necessitates other measurement types, such as modulation spectrum. This paper presents a comparative discussion of methods of measuring low-amplitude jitter, and other related behavior such as modulation spectrum. It compares “front-panel” oscilloscope-based methods, and time-interval analyzer and counter-based methods. It then describes a method of post-processing oscilloscope data in ways that are more accurate and useful to the design process than is possible from the scope´s front panel