Serial transmission of data representing video and audio signals has become standard practice across the broadcast industry. There are many advantages, including high data rate capability, self-clocking encoding schemes and acceptance by standards-making bodies. Yet, with these advantages comes a new set of parameters that must be monitored and controlled to ensure proper data transmission. Certainly the attributes of the data waveform itself have a major impact on transmission accuracy — commonly referred to as monitoring of the signal's eye pattern.
However, time-domain monitoring of the eye pattern can only give partial information about one of the most important serial transmission parameters: jitter. Other, more advanced testing means are required for full jitter analysis.
Serial video transmission
The Society of Motion Picture and Television Engineers (SMPTE) has codified the encoding and serial transmission specifications used in the broadcast industry. These are now universally accepted by equipment manufacturers, broadcasters and all the intermediate contributors. Serial data transmission involves taking the stream of 10-bit parallel words used in uncompressed component video schemes and converting that stream to serial.
Non-Return to Zero Inverted (NRZI) is the basic encoding scheme used in video transmission. Using NRZI, any binary “1” produces a transition in the serial data stream. A binary “0” does not cause a transition. This technique is advantageous since it lowers the average transition frequency and makes the resulting data stream polarity-independent.
Further encoding is performed using a scrambling polynomial to more evenly distribute transitions in the final serial stream. A side effect of this process is that certain flat signals (often a gray or single color image) cause relatively long periods without transitions in the serial data stream. These signals, referred to as pathological, can cause up to 20-bit periods without a transition.
Parameters influencing serial transmission integrity
As effective as these techniques are, there are still numerous challenges in transmitting data at 270Mb/s or 360Mb/s. Bandwidth is even more crucial for HD serial data transmission, where the data rate is 1.485Gb/s. These links require approximately 2.1GHz of bandwidth to maintain the data waveform integrity.
Long cable lengths will attenuate the electrical signal amplitude, which can indirectly cause frequency domain effects. Every piece of equipment involved in serial data transmission uses a cable equalizer component to automatically restore the received signal to nominal amplitude. Unfortunately, these devices introduce noise, which appears as jitter in the equalized data waveform.
While overall jitter amplitude is a crucial factor in assessing a serial video transmission system, it is meaningless without knowledge of what jitter frequencies are present. Jitter of a particular amplitude will cause different system effects as its frequency varies. The two jitter quantities — amplitude and frequency — cannot be separated when analyzing a system; knowledge of one parameter is meaningless without knowledge of the other.
Common measurement techniques
Any amplitude vs. time display — such as that produced on an oscilloscope — will supply information about the amplitude parameters of a serial data electrical signal. However, using an oscilloscope to view an eye pattern is difficult for a variety of reasons. The first reason is bandwidth. For proper evaluation of HD signals, an oscilloscope with at least a 10GHz response should be used. Waveform storage is another concern; many data points need to be accumulated in order to view a proper eye pattern, and this requires substantial memory. Accurate triggering is also a critical concern; the jitter high-pass (HP) filters required by SMPTE specs will not be available in an oscilloscope.
For these and other reasons, eye patterns are typically viewed on video-specific test equipment. Many waveform monitors designed for the serial data interface (SDI) also have the ability to display eye patterns. The display is still essentially that of an oscilloscope — a voltage vs. time analog waveform. However, being a video-specific device, the waveform monitor controls are already optimized for eye pattern display.
Advanced measurement features
Some test equipment can exhibit anomalies that yield incomplete or even misleading measurements. One such area is in the most basic display — the eye pattern itself. Most existing test instruments perform adequately with basic test patterns, such as color bars. However, many of these same instruments do not display some of the serial data stream characteristics that are unique to the SDI check field (often colloquially referred to as a pathological signal).
The receiver equalizer test portion of the SDI check field is defined for HD signals in SMPTE RP178. It produces 19 consecutive bit times without level transitions, followed by one bit time of the opposite polarity. (Either starting level can occur.) This pattern produces a high DC level in the serial stream. It also produces an asymmetrical eye waveform because of this DC offset. Many instruments lack the sample rate and display persistence to properly see this asymmetrical result. An example of a properly-displayed test is shown in Figure 1.
Two things are apparent from this display. First, the single period leading to the high DC offset can clearly be seen. Second, the effect of this offset produces middle logic levels while the input coupling capacitors are charging and discharging.
As useful as eye waveform analysis is, it has numerous weaknesses when analyzing jitter. While an eye pattern display can tell us the total amount of jitter, it gives no indication of jitter frequency or jitter variances over time. Other measurement techniques are required for these analyses.
The output of the jitter demodulator gives an accurate view of jitter vs. time. This output is usually plotted as jitter amplitude (as the Y axis) and time (as the X axis) at typical video sweep rates — one or two lines or fields. This yields a plot of jitter that shows any pattern dependencies, such as jitter steps during vertical interval or due to video content. Certain test patterns (particularly the SDI check field) can show appreciable jitter steps at the point in the field where the pattern changes. Examining only the eye waveform cannot reveal this change of jitter with time.
Figure 2 shows a jitter waveform at a two video line rate. Note the large change in jitter amplitude from the active portion of the line to horizontal blanking. Similar changes can often be seen at field rates with the changes in jitter amplitude occurring at vertical blanking.
Further analysis of this waveform can yield information about the frequency content of the jitter. With additional processing of the jitter waveform, different jitter frequencies present in the jitter can be calculated and plotted as a histogram. Like a dedicated spectrum analyzer, both the total bandwidth and the resolution bandwidth can be adjusted, allowing anything from a coarse overview to a narrow and specific frequency range to be analyzed.
Analyzing jitter frequency is essential, because it is a critical part of a system's jitter tolerance. The total amount of jitter that any system can process with acceptable performance is frequency dependent. Systems can usually tolerate more absolute jitter at very low frequencies. Jitter tolerance decreases at midband frequencies (roughly 100Hz to 1kHz) before starting to decrease again at high frequencies outside the response range of the serial receiving circuitry. So, knowing the absolute jitter value is not enough to predict the response of the system; the frequency must be known as well.
The relationship between jitter amplitude and frequency in HD systems is well-defined in SMPTE RP 184 and S 292. (See Figure 3.) These specifications define a lower frequency for jitter of any type. In HD systems, this frequency is 10Hz. Frequencies below this are defined as wander. These relationships also show the overall system jitter tolerance decreasing as jitter frequencies increase.
Measuring these amplitude and frequency relationships is critical in determining system performance. Simple jitter amplitude values are meaningless without knowing the frequencies comprising that jitter. Any other measuring or monitoring means — such as alarms based on jitter values — are similarly limited without knowing the frequency of the jitter being measured. Figure 4 shows one method of examining the relationship between jitter amplitude and frequency.
A set of SMPTE Recommended Practices and an Engineering Guide provides detail on the definition and measurement of jitter in bit serial systems. One of the key points in these documents is the relationship of jitter frequencies to system tolerance. Yet this aspect is often overlooked in system analysis.
Advanced jitter analysis requires not only accurate jitter values, but knowing what frequencies comprise the jitter present in any system. The ability to actually see jitter frequencies is key in determining system performance.
However, advanced signal analysis is not limited to frequency studies. There are other key tests that need to be made to ensure a properly-operating digital system. Keep in mind that many of these tests require specialized test equipment. While a waveform monitor may show some important aspects of an SDI signal, good engineering practice dictates that further tests be conducted.
Michael L. Richardson is director of product development, Videotek test and measurement, for Harris Broadcast Communications.