EMI Troubleshooting Cookbook for Product Designers provides the 'recipe' for identifying why products fail to meet EMI/EMC regulatory standards. It also outlines techniques for tracking the noise source, and discovering the coupling mechanism, that is causing the undesired effects. This title gives examples of simple, easilyimplemented, and inexpensive troubleshooting tools that can be built by the engineer or technician, and uses methods that require only a basic understanding of electromagnetic theory and a minimal background in EMI/EMC. It will show the engineer and technician how to develop a process for troubleshooting using a straightforward approach in solving what may seem like a rather complicated problem at first. It will provide guidelines on how to approach an EMI failure, things to try, how to choose the right parts and balance cost, performance, and schedule. This book tells readers trying to solve EMI problems what to do and how to do it.
Inspec keywords: electromagnetic interference; product design; electromagnetic field theory; electromagnetic compatibility
Other keywords: product designers; EMI measurements; diagnostic tool; EMI troubleshooting cookbook; EMI failure; electromagnetic theory; EMC theory; EMI-EMC
Subjects: Electric and magnetic fields; General electrical engineering topics; Electromagnetic compatibility and interference
That's right: if you've been testing and retesting at the EMC test facility, it's time to stop repeating the full test. Unless the problem is very simple, you are likely just wasting time and money. Instead of performing a full test, it might be time to look at a specific frequency or small frequency range, to focus on a specific test level, or to analyze just part of a circuit or filter. By treating your EMC test failure in a methodical and process-based manner, you'll be able to narrow down the root cause easier and quicker. Although the implementation of the fix or fixes may not be as easy, hopefully with the help of this book you will learn how to develop a range of potential solutions and to narrow down the probable causes for the test failure.
EMI requires a (1) source of energy, (2) a receptor or victim circuit or system, and (3) some coupling path for the energy to get from one place to the other. If there is no energy source, there is no EMI, and if there is no coupling path, there is no EMI. As shown in Figure 2.1, there are four primary coupling modes where energy can transfer from one place to another: inductive, capacitive, radiated, and conducted. Inductive coupling requires a time-varying current source and two “loops”or parallel wires (with return paths), which are magnetically coupled together. Examples might include a power transformer (high di/dt) in a switch-mode power supply coupling to a nearby cable or one “noisy”cable routed in proximity to another. Capacitive coupling requires a time-varying voltage source and two “plates”of metal closely coupled together; these can also be two parallel wires. An example might include a large heat sink of a switch-mode power supply (high dV/dt) that couples to a cable or adjacent PC board.
The most common instrument used in any EMC test lab is the spectrum analyzer. The spectrum analyzer is a swept-tuned superheterodyne device, although fast Fourier transform (FFT)-style analyzers are becoming more common. FFT analyzers can capture single-shot type events but often have limitations. Examples include limited frequency range it can measure, limited sensitivity to low-level signals, and limited dynamic range of the measurement. There have also been recent developments in real-time spectrum analyzers (RTSA), which use a combination of a superheterodyne receiver, FFT processing, and fast update rate.
Radiated emissions will most likely be your highest risk when performing compliance testing at the test facility. With all the high-speed digital circuitry inside electronic products today, it becomes all too easy for harmonics of the clock frequencies and other fast-edged devices to radiate EM fields. Typically, failure modes will be cable radiation or leakage from enclosure seams or apertures. This handy checklist can be used either as a pretest check prior to compliance testing or as a check following a test failure.
In most cases, conducted emissions should be easier to control and avoid than radiated emissions. Being lower frequency, it is less influenced by parasitics than higher frequency problems. However, they are still an issue and must be considered. Thus, causes of conducted emissions, and the solutions for them, are usually easier to understand than those for radiated emissions. Most conducted emissions are due to switch-mode power supplies (SMPS), and the best power supply designs are usually adequately filtered at the power input. However, many OEM power supplies are poorly designed, have horrible emissions, yet have FCC and CE markings. When these power supplies are loaded into a reactive load, instead of the nice resistive load it was designed for the power supply may start to get unstable or noisy, and additional measures are usually needed to keep it in compliance. Also, most commercial line filter modules or filter circuitry is designed to cover frequencies up to only 30 MHz. Therefore, there is the possibility that harmonics generated by the switching devices or rectifier switching transients can still make it through the filter. In addition, with all the high-speed digital circuitry inside electronic products today, it is possible for higher frequency harmonics to contaminate the system power supply and leak out through the filter and back out the power line. Therefore, while our experience demonstrates that most well-designed filters are sufficient, always be on guard for situations where the filter is compromised - either by design or by system design issues, such as poor internal cable routing, filter or power supply placement, or poor connection to chassis or signal returns. Typically, failure modes will be minimal to the product itself, but high emissions can upset sensitive measuring equipment or communications receivers nearby or connected to the same power line circuit.
One important EMC compliance test is to determine whether external RF fields can affect your product. This test is also often referred to as radiated immunity or radiated susceptibility testing and is defined by the IEC 61000-4-3 standard for commercial products. For commercial products, the test is usually performed from 80 to 1,000 MHz at E-field levels from 3 to 20 V/m depending on the product environment or application. The test is performed in a shielded semi-anechoic chamber using a broadband antenna to transmit the RF in the direction of the product under test. The shielded chamber prevents interference to other communications services. Some military, vehicular, or aerospace applications require testing to 200 to 1,000 V/m and frequencies up to 18 GHz or more. The RF signal is generally modulated by a 1,000 Hz AM sine wave modulation set to 80% for commercial testing and either square wave or short-duration (at times less than 1%) pulsed modulation, typically at 1 kHz for military and aerospace testing. The modulation is designed to test for audio rectification issues (or radar pulse interference for military testing). For example, if the RF signal is rectified by semiconductor junctions or in audio or other analog circuitry, the low-frequency modulation could cause bias upsets or otherwise disrupt sensitive analog circuitry.
One important EMC compliance test is to determine whether external low frequency radiated RF fields can couple into your product via I/O or power cables. This test is often referred to as conducted immunity or conducted susceptibility and is defined by IEC 61000-4-6 for commercial products. The test is usually performed from 150 kHz to 230 MHz at voltage levels of 1, 3, or 10 volts root mean square (RMS), depending on the product environment or application. Some military, vehicular, or aerospace applications require testing at more rigorous levels. The RF signal is generally modulated by a 1,000 Hz AM modulation set to 80% for commercial testing and a 1 kHz square wave pulsed modulation is used for military and aerospace testing. The modulation is designed to test for audio rectification issues. For example, if the modulation is rectified (in audio or other analog circuitry), it could cause bias upsets or otherwise disrupt sensitive analog circuitry. Because it is difficult to reproduce a uniform field at these frequencies in a shielded chamber, the RF is coupled directly to the product I/O or power cables through various means. For commercial products, the test is performed only on I/O cables that are typically longer than 3 meters (e.g., Ethernet) or power cables.
High-frequency transients and impulses such as electrically fast transients (EFT) are caused by light switches, relay chatter, or motor start-up transients on the power line. These transients often occur in bursts and can cause upset to your product if the power line filtering is inadequate. The focus of this chapter will be with the IEC 61000-4-4 EFT test, but the concepts apply to all high-frequency transient events. The test is performed between line and neutral, line to safety ground, and neutral to safety ground and consists of a repeating burst of pulses. The test is also performed on I/O, signal, or data cables typically longer than 3 m (e.g., Ethernet) using a capacitive coupling fixture. Several performance levels may be acceptable (refer to the EFT standard IEC 61000-4-4 for details), but loss of data, system reset, or damage is generally considered a test failure.
This chapter discusses the effect of external electrostatic discharge (ESD), or induced field and secondary discharges to a product as a part of EMC compliance testing. It includes the following topics: ESD checklist; typical failure modes; trouble shooting at the test laboratory; low-cost testing tools; and typical fixes.
The types of pulses used in surge and lightning testing include double exponential style pulses and damped sine wave pulses. Of the two, the double exponential pulse is the most difficult to deal with. The pulse durations are much longer than damped sine. Also, since they are not oscillatory, the components exposed to the pulse are experiencing a flow of charge in one direction and not able to discharge until after the threat signal is passed. Types of transient suppression devices include gas discharge tubes, transient voltage suppressors (TVS), metal-oxide varistors (MOV), diodes, thyristor, and other basic filter components. High-energy impulses are different from those of electrical fast transient (EFT) or ESD style pulses. The latter two can be treated as RF signals or highfrequency noise and typically do not require components that can handle significant energy. High-energy pulses, known as surge in the commercial electronics world and lightning or EMP in the military and aerospace electronics, are significantly slower in rise time and decay. They also have much lower source impedance than the other pulses. Lower source impedance results in greater current levels at the same voltage.
This chapter is a collection of product or application-specific EMI considerations related to EMI testing and troubleshooting. It encompasses topics that don't necessarily conform to a checklist format as in previous chapters.
The paper presents definitions, power ratios, frequencies and wavelengths, the EM spectrum, shielding effectiveness and slot length, Ohm's law, electric field from differential-mode current, electric field from common-mode current, antenna relationships, resonance, VSWR and return loss, E-field levels and transmitter power output.
Clock or crystal oscillators can generate a large number of high-order harmonics because of their typically fast edge rates. In this section, a clock harmonic analyzer spreadsheet for harmonics identification was developed. By entering the clock oscillator frequency (MHz) in the green box, all the higher order harmonics will be calculated in the second column. The first column indicates the harmonic number.
The Bode plots of simple circuits may be plotted on reactance (also known as impedance) paper to quickly indicate the resulting impedance of the circuit versus frequency. This is handy for quickly sketching out the frequency response of simple R-L-C circuits. Reactance graph paper is a log-log representation of resistance, capacitance, inductance, and frequency all on a single graph.
When going to a testing laboratory, it is often a good idea to not assume the lab will have all the tools and equipment you might need. This is especially true if your equipment needs a special tool to remove hardware, or if you might need to repair or rework a connector. In addition, here are some suggested contents of a typical EMI troubleshooting kit. The specific equipment specified is only a suggestion; many other manufacturers' products should work equally well.
The most effective filtering is placed on the PC board near the connectors. This applies to both signal lines and power input. Filters work by combining blocking action with diversion of noise currents. We block currents with high series impedance (resistors, inductors, or ferrites) and divert currents with low shunt impedance (capacitors). Generally, you can expect about 30 to 40 dB reduction in noise signal; however, you must be careful not to affect the desired signal too much. Points to keep in mind: It's always best to avoid creating the noise currents in the first place. Common-mode chokes on cables (such as clip on ferrites) are limited to about 10 dB reduction at the most. A filter is like a fence; know where the boundary should be. For I/O or power supply inputs, locate the filter at the connector. For noisy ICs, locate the filter as close to the noise source as practical. When designing filters and using filtering components, always remember the return path is part of the filter. If a filter (e.g., a capacitor) is placed close to the noise source but it then has a return path which is rather long to get back to the source, then it may not work very well even if the cap is physically close to the source of noise.
Cables or other metal (antenna-like) structures can couple to sources of common mode currents and end up radiating, causing product failures during compliance testing. During the troubleshooting process, it would be helpful to determine the resonant frequency of these cables or structures to confirm they are the source of certain harmonic signals. Often, as you probe a circuit board or measure the emissions from a product, you may notice a group of individual harmonics, which peak in amplitude over a given frequency range. This may indicate that a metal structure or cable is resonant at these peak frequencies. By analyzing whether the cable or structure is a halfwave or quarter-wave long, it might be identified and remediated in some way. To do so, you will need to convert the frequency into a corresponding quarteror half-wavelength.
Most countries regulate Electromagnetic Compatibility (EMC) of electronic products marketed or sold within its borders. Much of today's current regulations and standards development was driven by the European Union (EU) starting in 1992. In parallel with this renewed focus on EMC standards in the EU, the International Electrotechnical Commission (IEC) and the International Special Committee on Radio Interference (CISPR) started developing or updating the current emission and immunity standards we use today. Most countries worldwide have adopted some or all aspects of the international (IEC) set of EMI standards (IEC 61000series) and CISPR standards.
This paper provides detailed description of various symbols and acronyms used for the study of electromagnetic compatibility, antennas, electromagnetic interference and other fields related to electromagnetism.