A Closer Look at Random Jitter
Random jitter is caused by various random processes such as thermal noise at the transmitter. Imagine a graph in which the X axis denotes time, and the Y axis denotes the probability of seeing a transition at that point in time. For a signal with no jitter, the graph would be 0 everywhere except at the transition point, where the value would be unbounded.
If random jitter is introduced, then the shape of this curve will be a Gaussian curve, and this characteristic can be seen in the scope shot in Figure 3 alone, and again in the transition distribution seen in the blue histogram in the scope shot of Figure 5.
Figure 4: Gaussian distribution curve
What’s significant about random jitter is that it is unbounded. If you are expecting a signal transition in 5ns on a signal, the effect of random jitter is that there is a chance it will not occur until next Tuesday, although the probability of a large deviation drops off fairly quickly.
Still, the effect of random jitter is that for a system with random jitter, there will always be bit errors, although if the jitter is low enough, this BER can be arbitrarily low. For this reason, random jitter is often specified in terms of the standard deviation (σ) of the distribution.
Figure 5: An eye diagram showing random jitter
NEXT: Separating Random and Deterministic Jitters and Determining Sources
Page 3: next page
