Jitter in Professional Video Systems

Understanding the causes of deterministic and random jitter in SDI and other digital video signals, and how to minimize jitter in the design of pro video equipment meeting SMPTE specs.

By Mark Sauerwald, Applications Engineer, MTS, National Semiconductor Corp.

Page 2 of 7
Video/Imaging DesignWire
(9/11/2009 7:00:13 PM)

How does SMPTE Define and Measure Jitter?
In a broadcast video system, the Society of Motion Picture and Television Engineers (SMPTE) dictate how to measure and specify jitter. This is done in two of their ‘Recommended Practice’ or RP documents: RP 184 “Specification of Jitter in Bit-Serial Digital Systems” outlines the method for specifying jitter, while the method for measuring this jitter is described in RP 192.

Basically, SMPTE specifies jitter as the total jitter measured peak-to-peak. SMPTE then divides it into two different ‘types’ of jitter by band limiting the frequency response to some band. For the 3Gbps serial standard, ‘Timing Jitter’ includes all jitter components from a frequency of 10Hz up to 300MHz, and ‘Alignment Jitter’ includes all jitter components from 100kHz up to 300MHz.

null

Figure 2: Graphical illustration of SMPTE jitter definitions

These numbers are useful for comparing one piece of equipment to another, but not when constructing a system. Let’s look at why this is and how you can test a system to get valuable numbers that will help to predict the errors we are concerned about. In Figure 2, the brown area of the graph corresponds to the alignment jitter, and the combination of both pink and brown represent the timing jitter.

A Closer Look at Jitter — Deterministic and Random Jitter
Jitter comes from various different sources, but we can broadly divide it into two different sorts of jitter — Random Jitter and Deterministic Jitter. Each of these can be further divided into other subdivisions, but we won’t go into that for this analysis. Instead, we’ll look at jitter characteristics and give an example of what causes each type of jitter.

Deterministic jitter is not random noise. It’s a phenomenon which has the effect of moving the point in time where the transition point happens.  This occurs when you pass a signal through a band-limited system that does not allow the transition to move to full amplitude in one unit interval. In this case, if there was a transition in the previous unit interval, then the signal begins transitioning from a lower voltage than had there not been a transition, and it will pass through the detection threshold earlier - leading to a jitter error which can be predicted based on the prior bit pattern.

This can be seen by looking at the eye diagram in Figure 3, which shows a signal with pattern dependent jitter. This particular type of pattern dependent jitter is sometimes known as Inter-Symbol-Interference or ISI. The characteristics of ISI in the eye diagram include the banding of the transition portions of the waveform, which can be seen as multiple peaks — appearing in blue on the Figure 3 scope shot — in the histogram.

null
View full size

Figure 3: Eye diagram of a signal showing deterministic jitter

One of the key attributes of deterministic jitter is that it is bounded with a maximum and a minimum value. This means if your signal has only deterministic jitter, and there is sufficient timing margin, you’ll have a chance of building a system that will never have a bit error.

NEXT: Random Jitter

Page 2: next page

Pages: 1 2 3 4 5 6 7