Electromagnetic Compatibility (EMC) Design and Test Case Analysis

Preface

The majority of domestic electromagnetic capacity books have a common defect, which is the lack of connections between design and testing. The discussion of the approach and techniques of EMC design should be based on EMC testing, not only because the first challenge of EMC design is the EMC test but also because those key factors like interference source, receiving antenna, and equivalent radiated antenna, which are critical to EMC analysis, will only exist during the EMC test. Taking the conducted emission test as an example, its essence is the voltage across a resistor in the line impedance stabilization network (LISN), when the resistance is fixed, the level of conducted disturbance depends on the current passing through the LISN resistor. EMC design is to reduce the current flow through the resistor. Possible tests include the typical immunity test, electrical fast transient/burst (EFT/B) test, big current injection (BCI) test, and electrostatic discharge (ESD) test, which is a typical common mode immunity test. The source of disturbance is a common‐mode disturbance, referred to the reference ground plane, i.e. the reference point of these disturbance sources is the reference ground plane used in the test, which means that the current generated by the disturbance will eventually return to the reference ground plate. This is the basic starting point to analyze such disturbance problems.

Imagine, for the above‐mentioned conducted disturbance test, that during the product testing, that the disturbance current does not flow through the LISN resistor, and at the same time, for the immunity test, that this disturbance current never passes through the product circuit, it is certainly very favorable for this product to pass the EMC tests, and this is what product design needs to consider. Therefore, the EMC design must be started from the EMC test. Electromagnetic Compatibility (EMC) Design and Test Case Analysis, as a project reference book, makes a close connection with the EMC test substance, EMC design principles, and specific product design to narrate EMC design methodology. Highly integrating the practical and theoretical contents is the biggest characteristic of this book.

The book is divided into seven chapters, in which the basic EMC knowledge is described in Chapter 1, mainly served for the 2–7 chapters. When readers read those later sections, if some basic concept is vague and not clearly explained, it can be easily consulted and checked from Chapter 1. Chapters 2–7 includes cases, which are typical and representative. Case descriptions use the same format: [Symptoms], [Analyses], [Solutions], and [Inspirations]. By analyzing each case, we introduce the practical information about EMC design and diagnostic technology to the designers to reduce the mistakes made by the designer in the product design and the diagnostic of EMC problems, and achieve good product EMC performance. At the same time, illustrating the design principles through EMC cases enables readers to achieve better understanding on the origin of the design. [Inspirations] section actually sums up the problem and highlights related issues. It can be used as a checklist of product EMC design. The cases are divided into the following six categories:

Products’ structural framing, shielding, and grounding versus EMC. For most devices, shielding is necessary. Especially with the increasing frequency in the circuit, relying solely on the circuit board design often fails to meet EMC standards. Proper shielding can greatly strengthen EMC performance, but an unreasonable shielding design can not only fail to play its desired effect but also oppositely cause additional EMC problems. In addition, grounding will not only help solve the safety problem but is also very important for EMC. Many EMC problems are caused by an unreasonable grounding design, as the ground potential is a reference potential of the entire circuit. If the ground is not properly designed, the ground potential may be unstable, which leads to failed circuits. It may also generate additional EMI problems. The purpose of the grounding design is to ensure that the ground potential is as stable as possible, to reduce the voltage drop on the ground, thereby eliminating the interference.

Cables of products, connectors, and interface circuit versus EMC. Cable is always the path, which gives rise to radiation or bringing in the major disturbance. Because of their length, the cable is not only the transmitting antenna but also a good receiving antenna. And the cable has the most direct relationship, with the connector and interface circuit. Good interface circuit design not only can make the internal circuit noise well suppressed, so that there is no driving source for the transmitting antenna, but can also filter out the cable disturbance signal received from outside. Proper connector design of cable and interface circuit provides a good matching path.

Filtering and suppressing. For any devices, filtering and suppressing are key techniques to resolve electromagnetic interference (EMI). This is because the conductor of the device is acting as a highly efficient receiving and radiating antenna, and therefore, most of the radiation generated by the device is achieved through a variety of wires, while the external disturbance is often received by the conductor first, then brought into the device. The goal of filtering and suppressing is to eliminate these interfering signals on the wire, to prevent circuit interference signals being transferred onto the wire and then radiated through the wire, and also to prevent the conductors receiving the disturbance and taking them into the circuit.

Bypass, decoupling, and energy storage. When the device is operating, the signal level of the clock and data signals pins changes periodically. In this case, decoupling will provide enough dynamic voltage and current for the components when the clock and data are changing in normal operation. Decoupling is accomplished by providing a low‐impedance power supply between the signal and power planes. As the frequency increases, before reaching the resonant point, the impedance of the decoupling capacitors will decrease, so that the high‐frequency noise is effectively discharged from the signal line. Then the remained low‐frequency RF energy will not be affected. Best results can be achieved through storage capacitors, bypass capacitors, and decoupling capacitors. These capacitance values can be calculated and obtained by specific formula. In addition, the capacitor insulation material must be correctly selected, rather than randomly selected based on the past usage and experience.

PCB design versus EMC. Whether the device emits electromagnetic interference or is affected by outside disturbance, or generates mutual interference between the elects, PCB is the core of the problem (the component layout or the circuit routing of the PCB), and will have an impact on the nature of the product overall EMC performance. For example, a simulated interface connector position will affect the direction of common mode current flows in, and the path of the routing will affect the size of the circuit loop, these are the key factors of EMC. Therefore, a properly designed PCB is important to ensure good EMC performance for the product. The purpose of PCB design is to reduce the electromagnetic radiation generated by the circuit on the PCB and susceptibility to outside interference, and to reduce the interaction between the PCB circuits.

Components, software, and frequency jittering technique. Circuits are composed of components, but the EMC performance of the components is often overlooked. In fact, the packaging, the rising edge, the pinout of the component, and the ESD immunity of the device itself have a huge impact on the performance of a product’s EMC performance. Although the software does not belong to EMC academic areas, in some cases software fault‐tolerant technique can be used to avoid the impact on the products from the outside interference. Frequency Jittering is a popular technique to reduce the conducted and radiated emission from circuits in recent years, but the technology is not foolproof. This chapter will give details to the substance of the case and precautions for frequency jitter technique.

In fact, EMC design rules are just like traffic regulations. Noncompliance will certainly not result in a traffic accident, but the risk is bound to increase. EMC design is in accordance, noncompliance of some rules may also be able to pass the test, but the risk is bound to be increased. So there is an urgent need to introduce the product design risk awareness to the industry. The purpose of EMC design is to minimize the risk of EMC test, as only for those products complying with all the EMC, and traffic rules have the lowest EMC risk. Most of the listed problems in this book are originated from EMC problems encountered in practical work, each case is originated from the experience of these cases, which come from the accumulation of a large number of typical EMC cases the author encountered. For those classic cases, there are more detailed theoretical analysis. Each of the results of those cases is formed with one or more of EMC design rules, and it is worth learning and referring. As the engagement of the author is limited, this book may not contain all kinds of EMC issues in electronics and electrical products.

If readers discover any mistakes due to the author's incomprehensive knowledge, leading to unreasonable or inaccurate descriptions or even critical mistakes, please feel free to contact me. In addition, it must be mentioned that cases in this book are taken from specific products, for the convenience to readers, including the component symbols, codes, graphics, etc., and are not normalized in accordance with GB Standard.

I would like to express my gratitude to those who mentioned comments and suggestions for the book, beginning with Professor Wu Qinqin, PhD; I would also like to thank Professor Alain Charoy, Renzo Piccolo, Beniamino Gorini who is chairman of IEC/CISPRA, as well as those of my colleagues who mentioned comments about this book. Great thanks to the Electronics Industry Publishing House deputy editor Zhang Rong, Niu Pinyue and the Wiley Publishing House editors.

Junqi Zheng

Vice Chairman of IEC/CISPR

Exordium

Electromagnetic Compatibility (EMC) Design and Test Case Analysis has received more attention from readers since first publication in 2006. A number of defects in the version first published in the PHEI edition have been modified in this book, and on the basis of the original case analysis, the following important principles of design key points and specific treatment measures of EMC have been clarified by adding more cases.

Regarding the essence of EMC testing, the essence of various EMC test items defined in the standard is analyzed.

The design method of the filtering circuit on the power supply port has been clarified, including the selection of the filtering circuit and the parameters of the filtering components.

The EMC design method for digital‐analog mixed circuit has been clarified, including not only the crosstalk between analog circuit and digital circuit but also how to consider the EMC problem from a system perspective − especially how to deal with the analog ground and the digital ground, which puzzles the majority of designers.

The advantages and disadvantages of segregating the ground plane in the printed circuit board (PCB) have been clarified.

The method and principle of the interconnection between various working ground in PCB and the metallic casing for the product with metallic casing have been clarified, concerning whether the interconnection between them is needed; how to connect; and where the connection shall be positioned and other problems.

The sensitive traces, sensitive components, clock traces, or clock components and so on have been more fully described, including why they can't be positioned at the edge of the PCB. Concrete solution and remedy measures have been detailed.

The relationship between the stack‐up design and the EMC problems in the design of multi‐layer PCB has been clarified.

The magnitude of the differential mode radiation induced by loop has been clarified.

In China, EMC design got off to a late start but is developing rapidly. More and more companies and engineers have come to understand EMC, also gradually mastering a number of EMC design rules and applying them to guide the product design after several years of development. However, in China, with the rapid development of electronic design technique, there are many misconceptions about the nature of EMC in the process of product design. Eliminating these misunderstandings can help the reader to solve the inevitable EMC problem. These misconceptions are mainly reflected in the following aspects.

1 Grounding

The word grounding has entered the vision of the vast number of electronic product designers. Before we get to EMC, the most familiar ground is the earth. For safety reasons, certain metal conductors in electronic and electrical products must eventually be connected to the earth (called protective earth), generally connected to the earth through the building structure or a dedicated ground bar. For EMC, EMI radiation of the product can be reduced to the maximum extent by grounding, and the external disturbance going into the product can also be reduced to a maximum extent through grounding. However, is it necessary to connect the product to the earth? How can we correctly understand the grounding in EMC? These issues are addressed in Case 13, The Metallic Casing Oppositely Causes the EMI Test Failed, Case 14, Whether Directly Connecting the PCB Reference Ground to the Metallic Casing Will Lead to ESD, and Case 69, Detailed Analysis Case for the PCB Design of Analog‐Digital Mixed Circuit.

To a certain extent, the answer to the above question is given, to control product EMC, it is not necessary to connect the product to the earth of the natural world, grounding is to guide the flow of common mode current for EMC. Actually, for EMI, the reference point of the disturbance source of disturbance is the ground of the PCB. In order to avoid the disturbance source, through various channels, going into the antenna (such as cable in product), the correct grounding point should be a certain point in the ground of the PCB. So, this grounding from the flow of EMI disturbance current, should occur before the antenna (such as the cable); for most of the product's high frequency immunity, the reference point of disturbance source is the ground reference plate in the test, the correct grounding position should be the ground reference plate, to the purpose of this grounding is not to allow the external common mode current to be injected into the product. So, this grounding, from the perspective of the disturbance flow, occurs before the product's circuit. Regarding the grounding design of a product, the first thing needs to be considered is not selecting or designing single point grounding or multipoint grounding, but the location of the grounding and the grounding ways. If a product has a metallic casing, the above two kinds of grounding ways can be well realized with the help of the metallic casing or other parasitic parameters, this is why the equipment with metallic casing is more easily to pass the EMC tests, these two kinds of grounding are more difficult for the equipment with nonmetallic casing to pass the EMC tests.

[Loop and differential mode radiation] How is the radiation produced in the product? This is an indisputable fact that a PCB signal loop will produce the differential mode radiation.

The formula E(μV/m) = 1.3 × I(A)S(cm²)F(MHz)/D(m) gives the order of magnitude of the differential mode radiation.

E(μV/m): Radiation field strength, in unit of μV m−1

I(A): Current at a certain frequency in the loop, in unit of A

S(cm²): Signal loop area, in unit of cm²

F(MHz): Signal frequency of radiation in the loop, in unit of MHz

D(m): The distance from the test point to the loop, in unit of m

According to this formula, if the voltage amplitude of a clock signal is 3.3V, its frequency is 20 MHz (the working current is 3.3V/100Ω = 33mA. The virtual value of the third‐harmonic current at 60 MHz is 0.005A), the loop area is 1 × 1 cm (it is generally considered that this is a very poor PCB design). At the harmonic frequency 60 MHz, the radiation field strength at 3 m test distance from this clock circuit is 7 μV m−1. This value is a far below the standard radiated emission limit. That is to say, when the loop area is not large (with the popularization of multilayer PCB board technology, the loop area can be designed to be smaller), this radiation will not exceed the radiated emission limit specified in the current EMC standards. But it is worth noting that the resulting common‐mode radiation problem by loop area is increased, and Case 25, The Excessive Radiated Emission Caused by the Loop, gives an analysis. So, do not place undue emphasis on the differential mode radiation and neglect the more important common‐mode radiation.

2 Shielding

Assuming the differential‐mode radiation already mentioned is more than the standard limit (or the equivalent antenna, which causes the radiation to exceed the standard is inside the shield), as long as a metallic casing with an opening that is not large is used for shielding, the radiation problem can be solved. At this time, the metallic casing does not need to be connected with the PCB. However, with the elimination of the above misunderstanding, and with the equivalent antenna that causes the radiation to exceed the standard also outside the shield (such as the cable), at this time, the necessity of shielding with the metal casing has gradually declined. Case 13, The Metallic Casing Oppositely Causes the EMI Test Failed, is a typical case of this misunderstanding. Using metallic casing to achieve better EMC performance, it is because the metallic casing provides a better grounding path or bypassing path. If you want this path to become more direct, a reasonable interconnection between the PCB and the metallic casing will be needed. Designers must get rid of this misunderstanding when you want to add a shield to the product, and you have to be responsible for the consequences of this. We must take into account the physical location of the equivalent antenna of the radiation for product shielding design. If you cannot put them inside the shield, we must consider making a reasonable interconnection between the PCB and the metallic casing to achieve the transformation between shielding and bypassing.

3 Filtering

Capacitance and inductance are the basic components of the filtering circuit. The inductor behaves as inductive reactance and the inductive reactance increases with the increased frequency. The capacitor behaves as capacitive reactance and the capacitive reactance decreases with the increasing frequency. When the original circuit is in series with an inductor or parallel with a capacitor, the voltage divider network formed by the inductance and capacitance can reduce the interference voltage on the load. It seems that we could say, It would be good to add more inductors in series or more capacitors in parallel. In fact, as inductor and capacitor are energy storage components, there is a phase difference between the voltage and current on them, an extreme performance of the filtering network consisting of an inductor and a capacitor creates resonance. An LC filtering circuit occurs when the resonance occurs. The interference signal is not attenuated and oppositely amplified; it's horrible. This misunderstanding must be eliminated to design a good filtering circuit, and the resonant point of the filtering circuit must be far from the EMC test frequency. In the case of filters, more is not better.

By Junqi Zheng

Introduction

In this book, we use electromagnetic capacity case analyses as the main avenue for discussing EMC issues and introduce EMC technique in product design by describing and analyzing the cases. Our purpose is to explain the related EMC practical design techniques and diagnostic techniques during product design processes and to reduce designers’ mistakes when designing the products and diagnosing the EMC problems. The EMC cases in this book refer to various aspects such as structure, shielding and grounding, filtering and suppressing, cabling, routing, connectors and interface circuits, bypassing, decoupling and energy storage, PCB layout, components, software, frequency jittering technique, and more.

This book is based on practical applications, to explain the complicated principles with representative cases and avoid unnecessarily long theory. This book can be used as a necessary EMC reference book in the electronic product design departments, as well as a basic EMC training and reference material for electrical and electronics engineers, EMC engineers, and EMC counselors.

1

The EMC Basic Knowledge and the Essence of the EMC Test

1.1 What Is EMC?

Electromagnetic compatibility (EMC) is the capability of electronic and electrical equipment and systems to operate as designed in the forecasted electromagnetic (EM) environment, which is an important technique performance of the electronic and electrical equipment and the systems. EMC has the following two implications:

Electromagnetic interference (EMI), that is, the disturbance produced by the equipment and systems that operate in a certain environment, shall not exceed the limit required in their corresponding standard regulations. And the corresponding test items depend on the product classifications and the standards, for residential devices, ISM (industrial, scientific, and medical) devices and railway devices. The basic EMI test items include the following:

Conducted Emission (CE) Test on Power Line

Conducted Emission (CE) Test on Signal and Control Line

Radiated Emission (RE) Test

Harmonic Current Test

Voltage Fluctuation and Flicker Test

For military device, basic EMI test items include the following:

CE101: Conducted Emissions, Power Leads, 30 Hz to 10 kHz

CE102: Conducted Emissions, Power Leads, 10 kHz to 10 MHz

CE106: Conducted Emissions, Antenna Terminals, 10 kHz to 40 GHz

CE107: Conducted Emissions, Power Leads, Spike, Time Domain

RE101: Radiated Emission, Magnetic Field, 30 Hz to 100 kHz

RE102: Radiated Emission, Electric Field, 10 kHz to 18 GHz

RE103: Radiated Emission, Antenna Spurious and Harmonic Outputs, 10 kHz to 40 GHz

For vehicles, as well as vehicle electronic and electrical products, the basic EMI test items include the following:

Vehicle Radiated Emission Test

Conducted Emission Test for Vehicle Electronic and Electrical Parts/Modules

Radiated Emission Test for Vehicle Electronic and Electrical Parts/Modules

Transient Emission Test Vehicle Electronic and Electrical Parts/Modules

Electromagnetic susceptibility (EMS), that is, in normal operation, the devices and system in a certain environment can withstand the EM disturbance specified in their corresponding standard regulations. Similarly, according to the product classification and standards for residential devices, ISM (industrial, scientific, and medical) devices, and railway devices, the basic EMC test items include the following:

Electronic Static Discharge (ESD) Immunity Test

Electrical Fast Transient Burst (EFT/B) Immunity Test

SURGE Immunity Test

Radiated Susceptibility Immunity Test (RS)

Conducted Susceptibility Immunity Test (CS)

Voltage Dip and Voltage Interruption Test

For military device, the basic EMI test items include the following:

CS101: Conducted Susceptibility, Power Leads, 30 Hz to 50 kHz

CS103: Conducted Susceptibility, Antenna Port, Inter–modulation, 15 kHz to 10 GHz

CS104: Conducted Susceptibility, Antenna Port, Rejection of Undesired Signals, 30 Hz to 20 GHz

CS105: Conducted Susceptibility, Antenna Port, Cross‐modulation, 30 Hz to 20 GHz

CS106: Conducted Susceptibility, Power Leads, Spike, Time Domain

CS114: Conducted Susceptibility, Bulk Current Injection, 10 kHz to 40 MHz

CS115: Conducted Susceptibility, Bulk Current Injection, Impulse Excitation

CS116: Conducted Susceptibility, Damped Sinusoidal Transients, Cables and Power Leads, 10 kHz to 100 MHz

RS101: Radiated Susceptibility, Magnetic Field, 30 Hz to 100 kHz

RS103: Radiated Susceptibility, Electric Field, 10 kHz to 40 GHz

RS105: Radiated Susceptibility, Transient Electromagnetic Field

For vehicles, and vehicle electronic and electrical products, the basic EMS test items include:

Conducted Coupling/Transient Immunity Test at Power Port Conforming to ISO7637‐1/2

Coupling/Transient Susceptibility Test on Sensor Cable and Control Cable Conforming to ISO7637‐3

RF Conducted Susceptibility Test Conforming to ISO11452‐7

Radiated Susceptibility Test conforming to ISO11452‐2

Radiated Susceptibility Test in Transverse Electric and Magnetic (TEM) Cell Conforming to ISO1145‐3

Big Current Injection (BCI) Susceptibility Test Conforming to ISO11452‐4

ISO11452‐5: Strip Line Susceptibility Test

Parallel Plate Susceptibility Test Conforming to ISO11452‐6

Electrostatic Discharge Immunity Test Conforming to ISO10605

Electromagnetic environment: the operation environment for the systems or devices.

1.2 Conduction, Radiation, and Transient

The moment an air conditioner is switched on, a nearby indoor fluorescent lamp dims transiently. This is because a large amount of current flows to the air conditioner and the supply voltage drops quickly, affecting the fluorescent lamp supplied by the same power network. Also, when someone turns on a vacuum cleaner, the radio will make a para‐para noise because the weak (low amplitude but high frequency) voltage/current change generated by the motor of the cleaner is transferred into the radio through the power line. So, when the voltage/current generated by a device is transferred through the power line or signal line and then affects other devices, this voltage/current change is called the conducted disturbance. In order to solve this problem, the general method is to install a filter for the noise source and the power line of the victim to prevent the transmission of the conducted disturbance. Alternatively, when the noise appears on the signal line, the coupling path can be cut off by changing the signal cable to the optical fiber.

When using a cellphone, a nearby computer’s LED display might flash. The reason is that, in normal operation, the EM signal generated by the cellphone is transferred into the LED display through space. When a motorcycle drives past a house, the TV screen image might be affected because the pulse current generated by the motorcycle's ignition device transmits EM waves, which will propagate into space, then to the nearby television antenna, and then to the circuit, thus disturbing the voltage/current. In situations like this, the harmful disturbance, which is transferred through the space and causes the undesired voltage/current for the other device's circuit, is called radiated disturbance. The occurrence of the radiation phenomenon must be accompanied with the antenna and the source. Since the route of this transmission is space, shielding is the effective way to solve this problem.

So to summarize, the root cause of the disturbance is the undesired voltage/current change. If the change is transferred directly through the wire into the other devices, it is called conducted disturbance. If the change is transferred through the space in the form of EM waves into the other devices and causes the undesired voltage/current on the circuit or the line, this harmful disturbance is called radiated disturbance. However, in reality, the categorization is not so simple.

For example, for the disturbance source of the computing devices, such as computers, although the voltage/current of the digital signals are transferred through the circuit inside the devices, the interference is conducted out of the device through the power line or signal line and directly transferred to the other devices. At the same time, the EM wave generated from these wires can be harmful for the nearby devices in the form of the radiated interference. In addition, the circuit inside the computing device can also generate the EM wave and affect the other devices.

The occurrence of the radiated disturbance phenomenon is always inseparable from the antenna. According to the operation principle of the antenna, if the length of a wire is equal to the wavelength, it easily generates the EM wave. For example, a several‐meter‐length power line can radiate the EM wave at VHF frequency band (about 30–300 MHz). At frequencies lower than this frequency band, because of the longer wavelength of those frequencies, when the same current flows through the power line, strong EM wave cannot be generated and radiated. So, at frequency bands below 30 MHz, conducted interference is the primary problem. However, the disturbing magnetic field can be generated around the power line affected by the conducted disturbance and can then disturb amplitude moderation (AM) radio and others. In addition, as mentioned above, since the leakage disturbance from the power line at the very high frequency (VHF) band can be transformed into the EM wave and the scattered into space, the radiated disturbance becomes a more importance problem than the conducted disturbance. At higher frequencies, the circuit inside the devices of which the size is smaller than that of the power line can generate the radiated disturbance, which is harmful for other devices.

Thus, on the one hand, when the size of a wire and device is smaller than the wavelength, the main problem is the conducted disturbance; On the other hand, when the size is bigger than the wavelength, the main problem is the radiated disturbance.

There are some transient high‐energy impulse disturbances in the operation environment that are harmful for electrical devices. Generally, this kind of disturbance is called transient disturbance. Transient disturbance can be transferred into the devices not only through the wire in the conducted mode but also through the space in the broadband radiated mode, such as the disturbance to the radio, which is generated from the motor ignition circuit and the brushes of the DC motor. The main root cause of the transient disturbances generally includes lightning, electrostatic discharge, switching on or off the load (especially the inductive load) of the power line, and the nuclear electromagnetic pulse, for example. It can be seen that the transient disturbance is the EM disturbance with high amplitude but short duration. Common transient disturbance (the immunity of the devices needs to be verified by test) is divided into three groups, various electrical fast transient/burst (EFT/B), various surge, and various electrostatic discharge (ESD).

1.3 Theoretical Basis

1.3.1 Time Domain and Frequency Domain

Any signal can be converted from time domain to frequency domain through Fourier transform, as shown in the following formula:

(1.1) equation

where x(t) is the time‐domain waveform function of the electrical signal; H(f) is the frequency function of this signal; 2πf = ω, ω is angular frequency and f is the frequency.

The frequency spectrum of a trapezoidal pulse function, as shown in Figure 1.1, comprises the main lobe and countless minor lobes. Although every minor lobe has its maximum value, the general trend is to decline linearly with the increased frequency. The peak value curve in the frequency spectrum of the trapezoidal pulse, which has the rising time tr and the pulse width t, includes two turning points, 1/πt and 1/πtr. In the frequency spectrum, the peak value is constant at low frequency, but it drops with the slope of −20 dB/10 dec after the first turning point and then drops with −40 dB/10 dec slope after the second turning point.

Figure 1.1 Spectrum of trapezoidal pulse function.

Thus, when designing the circuit, under the condition that the normal logic function is ensured, we should increase the rising time and the falling time as much as possible, which contributes to reducing high‐frequency noise. But, because of the existence of the first turning point, the periodic signal with a steep rising edge and low frequency does not include higher‐order harmonic noise with a high level (for the calculation of each harmonic amplitude, you can refer to my book Electronic Product Design – EMC Risk Evaluation, published by PHEI in 2008).

Since at every sampling of the periodic signal, the frequency spectrum is the same, the frequency spectrum is dispersed but its amplitude is high at each frequency, which generally is the narrow‐band noise. For a nonperiodic signal, since the frequency spectrum at every sampling is different, the frequency spectrum is very wide and its amplitude is small, which generally is the wide‐band noise. In an ordinary system, the clock signal is a periodic signal, but the signal on the data line and the address line generally is nonperiodic, so the root cause for the radiated emission beyond the standard limit is always the clock signal. The frequency spectrum of the clock noise and the data noise is shown in Figure 1.2.

Figure 1.2 Spectrum of clock noise and data noise.

1.3.2 The Concept of the Unit for Electromagnetic Disturbance, dB

Electromagnetic disturbance is generally measured in decibels. The original definition of decibel is the ratio of two powers. As shown in Figure 1.3, the decibel is a logarithmic unit used to express the ratio of two powers.

Figure 1.3 The concept of dB.

The unit dBW is often used to denote a ratio with 1 W reference power, and similarly dBm for the 1 mW reference power, as shown in Figure 1.4.

Figure 1.4 The dB of power value.

The decibel scale of the voltage can be calculated from the decibel scale of the power, as shown in Figure 1.5. (The precondition is, R1 = R2, and they are generally 50 Ω.)

Figure 1.5 The concept of decibels‐voltage.

In EMC domain, generally, dBμV is used directly to express the voltage, which is the voltage relative to 1 μV, as shown is Figure 1.6.

Figure 1.6 The dB of voltage value.

For example, the amplitude of the electrical field is used to evaluate the radiated disturbance, and its unit is V m−1. In EMC domain, the decibel unit is generally used, which is dBμV m−1. When combining the antenna and the disturbance measurement instrument together to measure the amplitude of the disturbance field strength, what the instrument can measure is the voltage on the antenna port. The voltage with the addition of the antenna coefficient is the field strength of the measured disturbance.

Note that the cable attenuation is not taken into account.

1.3.3 The True Meaning of Decibel

When the EM disturbance from the devices cannot meet the limit requirement specified in the relative EMC standard regulations, we must analyze the root cause of the excessive emission and clarify the trouble. Many engineers have tried to accomplish this, but troubles still exist. One reason is that the diagnosing work can drop into an endless loop. The following example illustrates this situation.

Assume that the conducted emission is excessive when testing a system, and it means that the system cannot meet the CLASS B limit required in EMC standard CISPR22, as shown in Figure 1.7. After preliminary analysis, there are four possible reasons:

Conducted emission produced by the transformer

Conducted emission produced by the switching transistors in the switching mode power supply

Conducted emission produced by the PCB design defects

Conducted emission produced by the auxiliary equipment

Figure 1.7 The composition and level of the conduction noise of a product's power port.

When locating the problem, the factors related to the transformer are removed first to reduce the conducted emission. The result is that there is no obvious reduction, and the constitution and the amplitude of the conducted emission at the power port after those factors are removed is shown in Figure 1.8. So it is concluded that the transformer is not the main reason for the excessive conducted emission but we request the change for the transformer. Then we dispose of the switching transistor in the switching mode power supply. Remove its adverse factors to reduce the conducted emission at the power port, and then it is found that the test result is not obviously improved. The constitution and the amplitude of the conducted emission at the power port after removing these factors is shown in Figure 1.9. It is then concluded that the switching transistor is not the main reason, either.

Figure 1.8 The composition and level of the conduction noise of a product's power port without a transformer.

Figure 1.9 The composition and level of the conduction noise of a product's power port without the switching tube.

Then check the printed circuit board (PCB). If we improve the PCB to remove the original defects, we discover that the signal amplitude on the frequency spectrum is barely decreased. The constitution and the amplitude of the conducted emission at the power port after improving the PCB is shown in Figure 1.10. So, it is concluded that PCB is not the main reason, either. From the change of the relative amplitude, it seems that the PCB can be ignored.

Figure 1.10 The composition and level of the conduction noise of a product's power port without the PCB.

Hereto, the conducted emission problem of this product still exists because the test engineers ignored the signal amplitude on the frequency spectrum, expressed in dB scale. Let's have a look at the cause of this phenomenon. Assume the amplitude of the conducted emission generated by the transformer is Vn, the amplitude of the conducted emission generated by the switching transistor in the switching mode power supply is 0.7Vn, the amplitude of the conducted emission generated by PCB design defects is 0.1Vn, and the amplitude of the conducted emission generated by the auxiliary equipment is 0.01Vn. On this condition, if we remove the factors related to the transformer and the switching transistors at the same time, the test results show an obvious improvement, as shown in Figure 1.11. On this basis, we remove the PCB relevant factors, which were considered to have nothing to do with this problem formerly, and then the result show a big change, too. The constitution and the amplitude of the conducted emission at the power port after removing the factors related with the transformer, the switching transistors, and the PCB is shown in Figure 1.12.

Figure 1.11 The composition and level of the conduction noise of a product's power port without transformer and switching tube.

Figure 1.12 The composition and level of the conduction noise of a product's power port without transformer, PCB and switching tube.

Actually, although the absolute value of the PCB contribution is only 0.1Vn, which is a small value compared with the amplitude of the conducted emission generated by transformer Vn and that of the switching transistors 0.7Vn, it is a high value relevant to the amplitude of the conducted emission generated by the auxiliary equipment 0.01Vn. Thus, when the factors relevant to the transformer and the switching transistors exist, the effectiveness of the PCB factors is insignificant. However, when the factors relevant to the transformer and the switching transistors are removed, the effectiveness of the PCB factor becomes decisive.

In conclusion, the correct EMI diagnostic method is that, even if there is no obvious improvement, keep the former suppression measure to a possible disturbance source, and add more suppression measures to the other possible sources. If the disturbance amplitude declines a lot and passes the test when a certain measure is taken, it does not mean that this source is the main reason but only that this source is big in magnitude relative to the previous sources, and can be the last one to be solved.

Above all, assume that when a countermeasure to a certain disturbance source is taken, all sources in this product are removed in 100%, and then the reduction of the EM disturbance should be infinitely great after the last source is removed. Actually, this is impossible. Any countermeasure cannot remove 100% of the disturbance. It can remove 99, 99.9, or even above 99.99%, but not 100%, ever. So after the last disturbance source is removed, the improvement is really good but still limited.

Finally, when the device meets the relevant regulations completely, to reduce the product cost and remove the unnecessary components or designs, we can take out the measures one by one. The first one to be considered should be the high‐cost components or material, as well as the measures that are not easy to be implemented. If the radiated emission is not excessive after removing the factors, then this measure can be removed in order to minimize the product cost.

1.3.4 Electric Field, Magnetic Field, and Antennas

1.3.4.1 Electric Field and Magnetic Field

Electric field (E‐field) exists between two conductors with different electric potential. Its unit is V m−1. The electric field strength is proportional to the voltage between conductors and inversely proportional to the distance between conductors. A magnetic field (H field) exits around current‐carrying conductors and the unit is A m−1. Magnetic field is proportional to the current and inversely proportional to the distance away from the conductors. When alternating current (AC) is generated by the alternating voltage on conductors, the EM wave will be produced. E field and H field will propagate mutually and orthogonally at the same time, as shown in Figure 1.13.

Figure 1.13 The E field and the H field propagate at the same time and are orthogonal to each other.

The propagation velocity of the EM field is decided by the propagation medium. In free space, the speed is equal to the velocity of light 3 × 10⁸ m s−1. Near the radiation source, the geometric distribution and strength of the EM field is decided by the characteristics of the interference source, and only at the far distance, the EM field is mutual orthogonal. When the frequency of the interference source is high and the wavelength is smaller than the structure size of the victim, or the distance from the interference source to the victim meets r ≫ λ/2π, the disturbance can be regarded as a radiated field − namely, the far field. It can radiate the EM energy in the form of the plane wave and get into the circuit of the victims. Through the insulation supports (including air), the disturbance can be coupled into circuits, devices, and systems; when it passes through the common impedance of the circuit, the circuit may be disturbed. If the frequency of the disturbance source is low and the wavelength λ is bigger than the structure size of the victim, or the distance from the interference source to the victim meets r ≪̸ λ/2π, the disturbance can be regarded as near field interference, which can be induced into the circuit. The near‐field coupling can be expressed by the capacitance and the inductance in the circuit. The capacitance represents the electric field coupling and the inductance represents the magnetic field coupling. So, the radiated disturbance can be directly conducted into the circuits, the devices, and the systems. Figure 1.14 shows the relationship between the far field, the near field, the magnetic field, the electric field, and the wave impendence, versus the distance away from the disturbance source.

Figure 1.14 The relationship between near field, far field, magnetic field, electric field, and wave impedance in the radiation field.

At 30 MHz, the turning point of the plane wave is 1.5 m, at 300 MHz, the turning point of the plane wave is 150 mm, and at 900 MHz, it is 50 mm.

1.3.4.2 Using an Antenna to Detect Signals

An antenna has two transforming functions: one is to transform the EM wave to the usable voltage and current in the circuit, the other is to transform the voltage and the current to the EM wave propagating to the space. Signal is transmitted to the space in the form of EM wave, and the EM wave comprises the magnetic field and the electric field, which are measured, respectively, by A m−1 and V m−1. The structure of an antenna depends on the type of the measured field. As shown in Figure 1.15a, the antenna used to detect the electric field is constituted by a rod and a metallic plate, and in Figure 1.15b, the antenna used to detect the magnetic field is a wire loop.

Figure 1.15 The shape of the antenna for picking up electric and magnetic fields.

Sometimes a part of the electronic and electrical products (such as the cable, the PCB trace, etc.) possesses these characteristics unconsciously and becomes the antenna. One of the important tasks in EMC design is to find and remove these unconscious antennas.

The electric field (V m−1) around the antenna will induce a voltage to ground (m ⋅ V/m = V) along the direction of the antenna length. The receiver connected to the antenna can detect the voltage between the antenna and the ground. This antenna model can also be equivalent as a lead wire of the voltmeter, which is used to measure the potential in the space, and the other lead wire of the voltmeter is the ground of the circuit.

1.3.4.3 The Significance of the Antenna's Shape

Some antennas are made up of coils. These antennas are used to detect the magnetic field rather than the electric field, and they are H‐field antennas. The current will be induced while the magnetic field penetrates the coil, just as the current flowing through the coil can produce a magnetic field that will pass through the coil. The two ends of the magnetic field antenna are fixed on a receiving circuit, so that the magnetic field can be detected by measuring the current of the loop antenna. The magnetic field is generally perpendicular to the propagation direction of the field, so the torus should be parallel with the propagation direction of the wave to detect the field. The antenna of the radiated electric field has two mutually insulated units. The simplest electric field antenna is a dipole antenna and its name hints that it has two units. Two conductor elements act as the two plates of the capacitor; only the field between the two plates of the capacitor is radiated to the space but not confined between these two plates. Furthermore, the magnetic‐field coil is similar as the inductor, and its field is radiated to the space but not confined to a closed magnetic loop.

1.3.4.4 Formation of Antennas and the Radiation of EM Field

As mentioned earlier, the electric field antenna can be associated with the capacitor. Figure 1.16a shows a simple parallel plate capacitor. The electric field is generated between the plates while the electric charge is accumulated on the plates. If the plate is spread and placed on the same plane, the electric field between the plates will be extended to the space. As shown in Figure 1.16b, the same situation occurs on the electric field dipole antenna. The electric charge on each part of the antenna will generate a field, which will be transmitted to the space, and there is an inherent capacitor between the two poles of the dipole antenna, as shown in Figure 1.16c. The dipole arms need to be charged by current, which flows on each portion of the antenna in the same direction, and this electric current is called as an antenna mode current. This condition is quite special, because it leads to the generation of the radiation. When the signal applied on the antenna poles oscillates, the E field is kept continuously alternating and transmitted to the space.

Figure 1.16 The principle of electric field antenna.

The electric charge and the E field generated by the electric current are perpendicular to each other. Applying the voltage on the antenna, the direction of the electric field E is from the positive pole to the negative pole. As shown in Figure 1.17a, the alternating current on the antenna generates a magnetic field H, of which the direction surrounds the metallic conductor and meets the right‐hand law, as shown in Figure 1.17b. God created this law; when the electrons move along the metallic conductor, the magnetic flux that surrounds the metallic conductor is created. If the right‐hand thumb directs the direction of the current, the direction indicated by the fingers surrounding the metallic conductor is the direction of the magnetic field. The inductance of the antenna is caused by the surrounded magnetic field. So, the antenna behaves as a capacitance formed with the charge distribution and an inductance formed with the current distribution.

Figure 1.17 Schematic diagram of electric field antenna radiation.

As shown in Figure 1.17c, the E field and the H field are perpendicular to each other. They are linked with each other and extended from the antenna to the space. A plane wave will be formed while the signal oscillates on the antenna. The transverse electromagnetic wave (TEM) is generated while the E and H are perpendicular to each other. The antenna can also transform a TEM wave through the reciprocal theory back to the current and the voltage, as the antenna has the complementary nature of transmitting and receiving. The radiation situation of the antenna is shown in Figure 1.18. The antenna's reactance sections store the energy in the electric field and magnetic field around the antenna. The reactive power in the antenna is exchanged backward and forward between the power supply of the antenna and the reactive components of the antenna.

Figure 1.18 The power flow of radiation.

As there is a 90° phase difference between the voltage and the current in the L–C circuit, a 90° phase difference exists between the E field (generated by the voltage) and the H field (generated by the current) of the antenna if the antenna's resistance can be ignored. In a circuit, active power can only be consumed when the load impedance has an active component and the current and the voltage are in phase. This situation also applies to the antennas. The antennas have some little resistance, so the power consumption caused by these active power components exists in the antenna. In order to produce the radiation, E field and H field must be in phase, as shown in Figure 1.17c. For the antennas behaving as a capacitance and inductance, how is the radiation produced? The in‐phase