The Impact Of Signal Jumping Across Multiple Different Reference Planes On Electromagnetic Compatibility

Similar documents
Determining The Size Of Cabinet Apertures For Effectively Mitigating Radiated Emissions. By David Norte Thursday, April 7 th, 2005

Ileana-Diana Nicolae ICMET CRAIOVA UNIVERSITY OF CRAIOVA MAIN BUILDING FACULTY OF ELECTROTECHNICS

Aries Kapton CSP socket

Chapter 16 PCB Layout and Stackup

Custom Interconnects Fuzz Button with Hardhat Test Socket/Interposer 1.00 mm pitch

Signal and Noise Measurement Techniques Using Magnetic Field Probes

Designing Your EMI Filter

Differential-Mode Emissions

"Natural" Antennas. Mr. Robert Marcus, PE, NCE Dr. Bruce C. Gabrielson, NCE. Security Engineering Services, Inc. PO Box 550 Chesapeake Beach, MD 20732

Internal Model of X2Y Chip Technology

Aries QFP microstrip socket

High Speed Digital Systems Require Advanced Probing Techniques for Logic Analyzer Debug

Aries CSP microstrip socket Cycling test

Learning the Curve BEYOND DESIGN. by Barry Olney

APPLICATION SPECIFICATION

Antenna Matching Within an Enclosure Part II: Practical Techniques and Guidelines

A VIEW OF ELECTROMAGNETIC LIFE ABOVE 100 MHz

Techniques to reduce electromagnetic noise produced by wired electronic devices

Minimizing Input Filter Requirements In Military Power Supply Designs

Aries Center probe CSP socket Cycling test

CHAPTER ELECTROMAGNETIC COMPATIBILITY. Contents of this Chapter...

Relationship Between Signal Integrity and EMC

X2Y versus CM Chokes and PI Filters. Content X2Y Attenuators, LLC

OMNETICS CONNECTOR CORPORATION PART I - INTRODUCTION

Understanding and Optimizing Electromagnetic Compatibility in Switchmode Power Supplies

Designing external cabling for low EMI radiation A similar article was published in the December, 2004 issue of Planet Analog.

Design and experimental realization of the chirped microstrip line

150Hz to 1MHz magnetic field coupling to a typical shielded cable above a ground plane configuration

Coaxial Cable Protection

EMC filters. Mounting instructions. Date: January 2006

Aries Kapton CSP socket Cycling test

The number of layers The number and types of planes (power and/or ground) The ordering or sequence of the layers The spacing between the layers

The Causes and Impact of EMI in Power Systems; Part 1. Chris Swartz

Understanding the Unintended Antenna Behavior of a Product

LISN UP Application Note

SIMULATION of EMC PERFORMANCE of GRID CONNECTED PV INVERTERS

VLSI is scaling faster than number of interface pins

The shunt capacitor is the critical element

Cross Coupling Between Power and Signal Traces on Printed Circuit Boards

Physical Test Setup for Impulse Noise Testing

T est POST OFFICE BOX 1927 CUPERTINO, CA TEL E P H ONE (408) FAX (408) ARIES ELECTRONICS

Decoupling capacitor placement

TECHNICAL REQUIREMENTS FOR ELECTROMAGNETIC DISTURBANCES EMITTED FROM LIGHTING EQUIPMENT INSTALLED IN TELECOMMUNICATION CENTERS

Heat sink. Insulator. µp Package. Heatsink is shown with parasitic coupling.

Probe Considerations for Low Voltage Measurements such as Ripple

Signal/Power Integrity Analysis of High-Speed Memory Module with Meshed Reference Plane 1

ELECTROMAGNETIC COMPATIBILITY HANDBOOK 1. Chapter 8: Cable Modeling

Choosing and using filters

1) The diagrams in the section Assembly notes were provided by Rittal GmbH Co. KG, Herborn and by Invensys Systems GmbH EUROTHERM, Limburg/ Lahn.

Common myths, fallacies and misconceptions in Electromagnetic Compatibility and their correction.

Suppression Techniques using X2Y as a Broadband EMI Filter IEEE International Symposium on EMC, Boston, MA

Chapter 12 Digital Circuit Radiation. Electromagnetic Compatibility Engineering. by Henry W. Ott

Low Cost Mixer for the 10.7 to 12.8 GHz Direct Broadcast Satellite Market

G019.A (4/99) UNDERSTANDING COMMON MODE NOISE

Power Electronics. Exercise: Circuit Feedback

Technical Report Printed Circuit Board Decoupling Capacitor Performance For Optimum EMC Design

Traveling Wave Antennas

Characterization Methodology for High Density Microwave Fixtures. Dr. Brock J. LaMeres, Montana State University

Chapter 7 Design of the UWB Fractal Antenna

Effect of slots in reference planes on signal propagation in single and differential t-lines

Methods for Reducing Emissions from Switching Power Circuits. A. McDowell, C. Zhu and T. Hubing

Advanced Signal Integrity Measurements of High- Speed Differential Channels

Advanced Transmission Lines. Transmission Line 1

Improving CDM Measurements With Frequency Domain Specifications

11 Myths of EMI/EMC ORBEL.COM. Exploring common misconceptions and clarifying them. MYTH #1: EMI/EMC is black magic.

High Speed Characterization Report

Microcircuit Electrical Issues

PDN Probes. P2100A/P2101A Data Sheet. 1-Port and 2-Port 50 ohm Passive Probes

Class-D Audio Power Amplifiers: PCB Layout For Audio Quality, EMC & Thermal Success (Home Entertainment Devices)

Texas Instruments DisplayPort Design Guide

SHIELDING EFFECTIVENESS

Reducing Motor Drive Radiated Emissions

The Impact of Broadband PLC Over VDSL2 Inside The Home Environment

White Paper: Electrical Ground Rules

Improving TDR/TDT Measurements Using Normalization Application Note

FLTR100V10 Filter Module 75 Vdc Input Maximum, 10 A Maximum

Successful SATA 6 Gb/s Equipment Design and Development By Chris Cicchetti, Finisar 5/14/2009

Common myths, fallacies and misconceptions in Electromagnetic Compatibility and their correction.

IEEE CX4 Quantitative Analysis of Return-Loss

DesignCon 2003 High-Performance System Design Conference (HP3-5)

Technology in Balance

Keywords: ISM, RF, transmitter, short-range, RFIC, switching power amplifier, ETSI

KRF EMC Filters Installation, Operation and Maintenance Manual

Modeling of Conduction EMI Noise and Technology for Noise Reduction

Corcom Product Guide. Introduction

EC6011-ELECTROMAGNETICINTERFERENCEANDCOMPATIBILITY

Investigation of Electromagnetic Field Coupling from DC-DC Buck Converters to Automobile AM/FM Antennas

Bill Ham Martin Ogbuokiri. This clause specifies the electrical performance requirements for shielded and unshielded cables.

Application Note AN-13 Copyright October, 2002

Experimental Investigation of High-Speed Digital Circuit s Return Current on Electromagnetic Emission

Antennas Studies for UWB Radio

Keysight Technologies Pulsed Antenna Measurements Using PNA Network Analyzers

Chapter 12: Transmission Lines. EET-223: RF Communication Circuits Walter Lara

Electro-Magnetic Interference and Electro-Magnetic Compatibility (EMI/EMC)

EMG4066:Antennas and Propagation Exp 1:ANTENNAS MMU:FOE. To study the radiation pattern characteristics of various types of antennas.

Transient calibration of electric field sensors

Challenges and Solutions for Removing Fixture Effects in Multi-port Measurements

Exercises for the Antenna Matching Course

Application Note 5044

Analogue circuit design for RF immunity

Transcription:

Copyright by Dr. Andrew David Norte, All Rights Reserved March 18 th, 2012 The Impact Of Signal Jumping Across Multiple Different Reference Planes On Electromagnetic Compatibility David Norte, PhD www.the-signal-and-power-integrity-institute.com Abstract Oftentimes, it is necessary to route a signal between multiple, different printed circuit board reference planes in route to a receiver. In this case, it is desired to understand the radiated emissions that result from such signal propagation when the propagating signal is coupled onto an unshielded differential cable that delivers the signal to a matched receiver. In such cases, impedance discontinuities exist at the locations where the signal jumps between any two reference planes. The pairs of reference planes are typically classified as power-ground planes, or ground-ground planes. The ground-ground classification can pertain to either different ground planes or the top and bottom sides of the same ground plane. Reflections can exist between adjacent discontinuities, as well as between the source and the first discontinuity, and between the receiver and the last discontinuity. If the source and receiver are both matched to their transmission line segments, then no reflections will exist at either the source or the load. This paper discusses the electromagnetic interference and the subsequent electromagnetic compatibility that is due to the cable when the source and receiver are both matched to their transmission lines. I. INTRODUCTION When routing a signal throughout a printed circuit board, it is very possible that this signal may need to be routed between multiple different reference planes, before being coupled onto an unshielded differential cable. In this situation, impedance discontinuities exist at the locations of the signal jumping. As a result of these impedance discontinuities, reflections will occur between the discontinuities and either the source or the load. If both the source and the load are matched to their transmission line segments, then no reflections will occur at either the source or load locations. However, reflections can still occur between adjacent discontinuities, for example. This paper discusses the resulting electromagnetic interference from the case in which a 200Mbps signal jumps between different reference planes several times, and is then coupled onto an unshielded differential cable. This coupling to the cable can be accomplished through the use of a transformer, for example. This paper also ignores the reflections that occur between adjacent discontinuities. In other words, the electromagnetic interference that is addressed in this paper relates to only the primary forward travelling signal, and does not consider reflections between adjacent discontinuities that ultimately make it to the receiver. This issue will be addressed in a future paper. II. SIMULATION RESULTS Figure 1 shows the simulation results of the frequency response at the input to the cable when the signal jumping before the cable includes one case of jumping between a power plane and a ground plane, one case of jumping between two different ground planes, and one case of jumping between both sides of the same ground plane. In this simulation, the via parasitic capacitance is equal to 1.0pF for all vias, and the via inductance is 1.0nH for all vias. In addition, when the signal jumping is between a power plane and a ground plane, it is assumed that a surface mountable capacitor connects the power plane to the closest via that then connects to the nearest ground plane. This capacitor is assigned a value of 0.01 F and is characterized with a parasitic inductance of 1.0nH, as well as a parasitic resistance of 0.2. Fig. 1. Frequency response at the input to the cable, and when there is one case of jumping between a power plane and a ground plane, one case of

Copyright by Dr. Andrew David Norte, All Rights Reserved March 18 th, 2012 jumping between two different ground planes, and one case of jumping between both sides of the same ground plane (red). The blue curve On the other hand, Fig. 2 highlights the frequency response when there are two cases of jumping between a power plane and a ground plane, two cases of jumping between two different ground planes, and two cases of jumping between both sides of the same ground plane. The addition of twice the number of jumps tends to increase the signal attenuation, as well as the width and depth of the notch that occurs at 5GHz. As long as the risetime of the propagating signal, t r, is such that the associated frequency, 1/(2t r ) Hz, is significantly below the location of the notch frequency, then minimal signal degradations should occur. It is then expected that risetimes greater than about 500ps will experience little signal degradations. Fig. 3. Frequency response at the input to the cable, and when there are two cases of jumping between a power plane and a ground plane, one case of jumping between two different ground planes, and one case of jumping between both sides of the same ground plane (red). The blue curve Fig. 2. Frequency response at the input to the cable, and when there are two cases of jumping between a power plane and a ground plane, two cases of jumping between two different ground planes, and two cases of jumping between both sides of the same ground plane (red). The blue curve If the total number of signal jumps includes two cases of jumping across a power plane and a ground plane, as well as one case of jumping across two different ground planes, and one case of jumping across both sides of the same ground plane, then Fig. 3 highlights this frequency response. In this case, a resonance at around 5.8GHz emerges, and the notch frequency of 5GHz remains, however, the strength of the notch is significantly lowered. Although a resonance occurs Fig. 4. Input signal with a 500ps risetime (black), the output signal with signal jumping (red), and the output signal without signal jumping (blue). at 5.8GHz, it is really the notch frequency of 5GHz that should be avoided because it occurs at a lower frequency.

Copyright by Dr. Andrew David Norte, All Rights Reserved March 18 th, 2012 Figure 4 shows the time-domain waveforms that are associated with the frequency response from Fig. 1. Note that the signal jumping imposes negligible signal degradations on the signal that is fed to the cable. As can be seen from Fig. 4, the input, and output waveforms appear nearly identical. On the other hand, Fig. 5 shows the output waveforms associated with the frequency response from Fig. 3. Figure 5 highlights the time-domain waveforms when only four cases of jumping between a power plane and a ground plane are encountered by the propagating signal. As can be seen from Fig. 5, ringing appears on the flattop portions of the output signal with the signal jumping. If in addition to the four signal jumps between a power plane and a ground plane, the signal also jumps between two different ground planes, as well as both sides of the same ground plane, then Fig. 6 shows the results. In this situation, the two additional signal jumps relating only to groundground jumping provides some low-pass filtering of the ringing on the flattop portions of the output signal from the previous case shown in Fig. 5. Given these various scenarios of signal jumping across various reference planes, it is of interest to understand the subsequent radiated emissions for the output signals from Figs. 5-6 when these signals are coupled onto a 1m long, 100 differential transmission line through some means such as a center-tapped transformer, for example. Fig. 5. Input signal with a 500ps risetime (black), output signal with signal jumping (red), and output signal without signal jumping (blue). Figure 6 shows the radiated emissions spectra for the signal propagating along the cable without any signal jumping (blue), as well as with the four ground plane-to-power plane signal jumps (red) from Fig. 5. In this case, signal jumping between the power and ground planes places a more significant shielding requirement on the cable. At around 5.8GHz, the additional shielding requirement is about 23dB for case of signal jumping. Without signal jumping, the shielding requirement for the cable in order to just pass the radiated emissions test is about 10dB through 10GHz. Fig. 5. Input signal with a 500ps risetime (black), output signal with signal jumping (red), and output signal without signal jumping (blue).

Copyright by Dr. Andrew David Norte, All Rights Reserved March 18th, 2012 23dB Fig. 6. Radiated emissions at the output of the cable when there are four cases of jumping between a power plane and a ground plane (red) in route to the input of the cable. The blue curve corresponds to the case in which no signal jumping occurs between the source Therefore, only when jumping between power and ground planes several times, the additional ringing that appears on the flattop portion of output signal places a significant shielding requirement on the differential cable in order to just pass the radiat ed emissions test. On the other hand, Fig. 7 shows the radiated emissions spectra when one additional signal jump occurs between two different ground planes, and when a second additional signal jump occurs between both sides of the same ground plane. From Fig. 7, it is clear that the additional jumping between two different ground planes and between the two sides of the same ground plane produced negligibly different radiated emissions spectra. Therefore, it appears that if multiple signal jumps occur between a power plane and a ground plane, then the printed circuit board designer needs to add additional signal jumps between either two different ground planes or the two sides of the same ground plane in order to demonstrate compliance with a required radiated emissions test. If only one additional signal jump between two different ground planes is added to the four signal jumps between a power plane and a ground plane, then Fig. 8 shows the radiated emissions results. In this case, the additional shielding requirement is about 11dB, instead of 23dB from Fig. 6. On the other hand, if the additional signal jump is between both sides of the same ground plane, then Fig. 9 shows the results. In this situation, the shielding requirement is about 21.5dB, which is nearly equal to the shielding requirement when only the signal jumping between the power and ground plane exists, which was shown in Fig. 6, and was equal to about 23dB. From Figs. 8-9, it is clear that minimally one additional signal jump between two different ground planes and one additional signal jump between two sides of the same ground plane are needed to nearly match the radiated emissions profile when no signal jumping exits. Fig. 8. Radiated emissions at the output of the cable, and when there are four cases of jumping between a power plane and a ground plane, and one case of jumping between two different ground planes (red). The blue curve Fig. 7. Radiated emissions at the output of the cable, and when there are four cases of jumping between a power plane and a ground plane, one case of jumping between two different ground planes, and one case of jumping between both sides of the same groun d plane (red). The blue curve From Fig. 9, it is clear that adding one case of signal jumping between both sides of the same ground plane does not provide any significant low pass filtering of the ringing, as expected.

Copyright by Dr. Andrew David Norte, All Rights Reserved Fig. 9. Radiated emissions at the output of the cable, and when there are four cases of jumping between a power plane and a ground plane, and one case of jumping between both sides of the same ground plane (red). The blue curve If the two additional signal jumps are between two different ground planes, then Fig. 10 shows the results. From Fig. 10, it is clear that the radiated emissions arising from this situation is nearly identical to the radiated emissions profile from Fig. 7. For the sake of completion, Fig. 11 shows the time-domain waveforms at the input to the cable. Fig. 10. Radiated emissions at the output of the cable, and when there are four cases of jumping between a power plane and a ground plane, and two cases of jumping between two different ground planes (red). The blue curve March 18th, 2012 Fig. 11. Input signal with a 500ps risetime (black), output signal with signal jumping (red), and output signal without signal jumping (blue). Note that Fig. 11 shows only slight signal degradations with respect to the case in which signal jumping is not present. The two additional signal jumps between two different ground planes that were inserted into the signal propagation path did not appear to significantly affect the signal integrity of the signal that feeds the cable. Therefore, it appears that whenever multiple signal jumps occur between a power plane and a ground plane, at least two additional signal jumps between two different ground planes, or between two different ground planes and between both sides of the same ground plane should be inserted into the propagation path in order to minimize the shielding requirement of the cable. It is also possible that more than four signal jumps between a power plane and a ground plane are necessary. As an example of this situation, when six cases of signal jumping across a power plane and a ground plane exist along the propagation path, it was determined through computer simulations that four additional signal jumps between two different ground planes were needed in order to require no additional cable shielding relative to the case in which no signal jumping occurred. Finally, the four signal jumps between different ground planes enabled up to eight signal jumps between a power plane and a ground plane without any significant signal integrity degradations, and without any additional cable shielding requirements relative to the case in which no signal jumping occurred. Although these results were established with respect to an input signal with a 500ps risetime, it is expected that input risetimes exceeding 500ps might require fewer signal jumps between different ground

Copyright by Dr. Andrew David Norte, All Rights Reserved planes. On the other hand, if the input risetime is less than 500ps, then it is expected that more signal jumps between different ground planes will be required in order to minimize the shielding requirements of the cable. The material covered throughout this paper can be studied through the interactive March 18th, 2012 signal integrity learning environment that is available at www.the-signal-and-power-integrity-institute.com.