Design Article
Layout guidelines help manage video equalizer's return loss (Part 1 of 3)
Mukund Krishna and Jeff Hushley, Cypress Semiconductor Corp.
8/19/2008 6:32 PM EDT
One of the critical SMPTE specifications for these devices is return loss, measured at the inputs of the equalizer and the outputs of the cable driver. When signals cross from the cable to the equalizer inputs or from the cable driver outputs to the cable, there is a change in impedance of the medium, which causes signal reflections measured as return loss.
This article shows how the prudent use of matching networks and certain board-layout methods can improve return-loss performance in a system. It starts by discussing the importance of return loss and methods to measure it. Next, it will discuss termination circuits to minimize return loss. Finally, it goes to a deeper level by discussing board-layout guidelines to manage return loss, with specific reference to the equalizer.
Return Loss
Return loss is defined at the point of intersection of media with different impedances as the measure of reflected energy from a signal transmitted into that point.
When an electrical signal is transmitted through a medium such as copper wire, a certain amount of the energy of the transmitted signal is reflected back from the point where the impedance of the transmission medium changes. A common example of such impedance changes in real-world applications are when a coaxial cable ends in a load with an impedance that differs from the characteristic impedance of the cable.
Hence, the larger the change in impedance, the more the signal is reflected back. Everyday examples of impedance discontinuities are when a cable terminates into a connector, change in trace path and/or shape on a PC board, damage to cable due to long-term physical stress and abuse, and improper installation of cables and their connectors.
Mathematically, return loss is defined as follows:
(Click on equation to enlarge)
ZO = impedance of transmission line
ZL = impedance at load
This usually results in positive dB values, since the numerator is smaller than the denominator. If the magnitude is large, it means the loss is small and vice versa.
S-Parameters
S-parameters (Scattering parameters, see Reference 1) are commonly used for high-frequency circuits to supply information about a network. Instead of using currents, S-parameters use normalized incident and reflected waves at the port of measurement.
Figure 1 represents a high-frequency signal being fed into a two-port network that processes it and transmits a signal out into a load.
Figure 1: Incident and reflected waves in a two-port network
(Click on image to enlarge)
On both the reception and transmission sides of the network, the transmission media have a characteristic impedance of ZO. Zs represents the impedance of the source and ZL represents the load impedance.
These waves are the voltage magnitudes measured at the input and output of the two-port network:
a1: Wave from source incident on two-port network
b1: Wave from input of two-port network incident on the source
a2: Part of transmitted wave reflected by load and incident on output of two-port network.
b2: Wave from output of two-port network incident on load.
Hence, the total wave traveling from the output of the two-port network to the load consists of that part of the portion of a2 reflected by the output of the port, and that part of a1 that is transmitted through the two-port network.
Similarly, the total wave traveling from the input of the two-port network back to the source is made up of that part of a1 that is reflected by the input of the port, and that part of a2 that is transmitted through the two-port network.
This translates to the following equations:
(Click on equation to enlarge)
S11 and S22 represent the input and output reflection coefficients, respectively.
Most commonly, the parameter used to denote differential return loss is S11 also sometimes known as SD11.
Therefore, in decibels:
(Click on equation to enlarge)
In a system with an equalizer, the outputs of the device are being fed to another chip or are terminated. Hence, to draw a comparison with Figure 1, a2 becomes zero since there is no reflection from the load.
A later section will discuss some important termination and layout techniques for ensuring good return loss while using differential signaling.
Importance of return loss for the video equalizer
At higher frequencies such as those associated with high definition (HD) video rates, return loss becomes a very important issue since impedance mismatches cause more signal reflections at these frequencies, which in turn distort the original signal being transmitted to a greater extent The Society of Motion Picture and Television Engineers (SMPTE) places a limit on the return loss in a system based on the data rate.
For frequencies from 0 to 1.485 GHz (HD data rate), the return loss in a system should be greater than 15 dB. From 1.485 to 2.97 GHz (2xHD data rate), the return loss in the system should exceed 10 dB. With digital video transmission and distribution transitioning to HD and 2xHD, return loss in any cable, connector or board is a very critical parameter to be understood and minimized by design.
Therefore, the return loss factors contributed by the endpoints of the serial link have to be better than the limits specified for the system. In addition to the device's port characteristics and the package characteristics, the layout of the board, proximity of components, trace geometry, power and ground planes, the return loss compensation network and connectors are all important factors that affect the final return loss.
Measuring Return Loss
This section looks at how single-ended and differential return loss is measured and how graphs such as Smith charts (see Reference 2) are interpreted.
I. Network Analyzer-based measurement:
A commonly used method of measuring the return loss on a single port is to use an S parameter-based network analyzer such as the Agilent 8722ES. This analyzer is capable of measuring and graphically displaying the return loss over frequency. One important detail to keep in mind while using these analyzers is that many of them do not have 75-Ω port interfaces, which is the characteristic impedance defined for broadcast video transmission links. Most commonly, they have a 50-Ω interface and so a 50-to-75 Ω converter is needed.
Depending on the analyzer being used, it can display the measurement in either a Smith chart or a log scale graph. The log scale graph (shown in Figure 2) is simple to read, while the Smith chart is an older method that requires some calculation to obtain return loss numbers.
Figure 2: Log scale gaph showing return loss over frequency
(Click on image to enlarge)
Figure 3 shows a typical Smith chart. The chart is a graphical representation of the mathematical properties of a network. It is widely used in the RF world to represent port impedances and to determine S parameters, standing wave ratios and in impedance matching.
Figure 3: Smith Chart (Source: http://www.tmworld.com/article/CA187342.html )
(Click on image to enlarge)
The circles in green are constant-resistance circles and the arcs in red are constant-reactance arcs. Those red arcs on the positive y-axis represent inductive reactances and those below represent capacitive reactances. A given point plotted on the chart will be at the intersection of a constant resistance and constant reactance circle. This decides the impedance represented by that point R + jX.
Measuring Return Loss for the equalizer and cable driver:
While measuring return loss (RL) contributed by the video equalizer and cable driver, it is important to ensure the following:
- The board is to be populated with the device and all other components, as it would be for full functionality.
- Power supply is provided to the board within the limits of the device or suitable for normal operation.
- In the case of the equalizer, since the RL is measured at the inputs, there is obviously no need to give an input signal to the device.
- Ensure all unused inputs (equalizer) and outputs are terminated to 75 Ω
(Parts 2 and 3 discuss time domain reflectometry, layout guidelines, PC traces, board debugging, and examine a case study. You can read Part 2 here and Part 3 here.)
References
1. "Using S-parameter data effectively," Wilfredo Rivas-Torres, Agilent Technologies, Planet Analog, March 27, 2006.
2. "The Smith chart: more vital after all these years," Bill Schweber, EDN, March 18, 1999.
3. "Return Loss Layout Guidelines for the Video Cable Equalizer," available at www.cypress.com.
About the Authors
Mukund Krishna is an Applications Engineer for the lighting solutions group (EZ-Color) at Cypress Semiconductor Corp.at San Jose, CA. His responsibilities involve development and evaluation of demo systems, supporting customer designs from initial concept to design completion, creating product collateral and new product definitions. His experience at Cypress involves being an applications engineer for physical layer devices for high-speed communication interfaces including professional video broadcast. Mukund received an MSEE from the University of Southern California. He can be reached at ukk@cypress.com.
Jeff Hushley is a Staff Applications Engineer at Cypress Semiconductor Corp. His expertise is in the characterization and systems analysis of high-speed serial digital communication devices in broadcast and consumer video applications. He has a Bachelors Degree in Computer Engineering from the University of Toronto. He can be reached at fre@cypress.com.
