Electromagnetic Compatibility

Knowledge Base

The following chapters are thought as free education for those of you out there, who have to deal with EMC and electronics design.

We wrote down some of the most essential theoretical knowledge, which you need for mastering EMC. If you wish that we add a specific topic or if you have any questions: Please write us info@academyofemc.com.

 

1  The Decibel

The decibel is defined as a ratio of two quantities, typically power, voltage or current. In the field of EMC, the decibel must be understood. Here the most important points:

Gain [dB] and Loss [dB].

Let's have a look at the amplifier or damping network below.

 

The power, voltage and current gain of this network can be expressed in [dB] as [1.1]:

 

If R1 and R2 are equal (typically 50Ω), then the following term is equal 0:

 

And we can write the following for power/voltage/current gain:

 

Points to remember when it comes to gain calculations in decibel:

  • Amplification. If P2 is bigger than P1, the gain value in [dB] is positive. This means if there is an amplification, the power gain in [dB] is positive.

  • Damping. If P2 is smaller than P1, the gain value in [dB] is negative. This means if there is a power loss, the power gain in [dB] is negative.

  • Cut-off Frequency. At the cut-off frequency, the output power (P2) is half the input power (P1). And the power/voltage/current gains are all -3dB.

  • Factor to [dB]. If power increases by factor 2, the power/voltage/current gains increase by +3dB. If power increases by factor 10, the power/voltage/current gains increase by +10dB. To get an overview, have a look at our power/voltage/current ratios table below.

 

 

Absolute Levels [dBm, dBµV, dBµA].

The most common absolute power, voltage and current levels in EMC are [dBm], [dBµV] and [dBµA]. They are calculated like this [1.1]:

 

 

The most common absolute power, voltage and current levels in EMC are [dBm], [dBµV] and [dBµA]. They are calculated like this:

 

 

For example, a negative dBm-value means that the power level is <1mW. 0dBm = 1mW and a dBm-value bigger than 0 means that the power is higher than 1mW. The same for dBµV and dBµA: 0dBµV = 1µV (0dBµA = 1µA), a negative dBµV-value means <1µV, and a positive dBµV-value means >1µV.

 

 

 

Conversion formulas and tables.

Find below tables and formulas to convert between the different dB-units and from linear values to dB-values and vice versa. In addition, you can download a Excel sheet, which contains conversion calculations and more [1.2].

 
 
 
 
 

2  Frequency & Wavelength

It is fundamental that professional engineers in the field of EMC understand the basics about signal frequency f and wavelength λ. We therefore summarized the topic as follows:

EMC and Frequency.

In general, EMC issues occur with signals of frequency > 9kHz. This is the reason why most EMC Standards do not consider signals with < 9kHz.

  • Conducted Emission. Conducted emissions tend to occur at f < 30MHz.

  • Radiated Emission. Radiated emissions tend to occur at f > 30MHz. 

Wavelength: Calculation.

The frequency f of a sinusoidal signal and its wavelength λ [m] have the following relationship [2.1]:

 

Whereas v [m/sec] is the propagation velocity of the signal and f [Hz] the frequency of the signal.

 

 

Wavelength: Blank Wires vs. Cables & PCB traces.

It is important to understand that the signal propagation velocity v depends on the transport medium through which the electromagnetic field is travelling. Therefore, the same signal with the same frequency has a different wavelength λ in a blank wire (surrounded by air) than in a cable or PCB trace (surrounded by an insulation).​ The wavelength λ [m] is calculated the following way [2.1]:

 

where v is the signal propagation velocity in [m/sec], c is the speed of light (3E8 [m/sec]), f is the frequency of the sinusoidal signal in [Hz], εr is the relative permittivity [1] and μr is the relative permittivity [1] of the media through which the electromagnetic field is propagating. VF is called the velocity factor.

  • Wavelength in a blank wire. The wavelength λ of a signal with frequency f in a blank wire (or antenna surrounded by air) depends only on the speed of light c and the signal frequency f (v=c, because εr = 1 and μr =1 and therefore VF=1) [2.1]:

  • Wavelength in cables and PCB traces. The wavelength λ of a signal with frequency f in an insulated copper wire or a cable or a Printed Circuit Board (PCB) trace is [2.1]:

 

Where c is the speed of light (3E8 [m/sec]), f is the signal frequency [Hz], εreff the effective dielectric constant (relative permittivity) [1] through which the electromagnetic wave is propagating. The effective dielectric constant εreff is defined as the uniform equivalent dielectric constant for a transmission line, even in presence of different dielectrics (e.g. FR-4 and air for a microstrip line, see picture below).
The relative permeability μr is assumed to be equal 1 for cables and PCBs, because the insulation materials are non-magnetic. Thus, the velocity factor VF depends primarily on the effective relative permittivity (also called effective dielectric constant) εreff of the insulation or PCB material.

The calculation of the effective dielectric constant εreff [1] depends on the insulation material and the geometry of the transmission line (e.g. ribbon cable, microstrip, coplanar waveguide, etc.), because the amount of the electric field lines in the different medias depend on the geometry of the transmission line (e.g. see the microstrip line below).

The Excel sheet below contains a calculator for calculating the effective dielectric constant εreff [1] (effective permittivity) for some of the most common transmission lines:

  • PCB-traces. Microstrip, stripline, coplanar waveguide with reference plane.

  • Cables. Ribbon cable, twisted pair.

 

 

The velocity factor VF [1] of a transmission medium is the ratio of the velocity v [m/sec] at which a wavefront of an electromagnetic signal passes through the medium, compared to the speed of light in vacuum c [3E8m/sec]: VF=v/c. Thus, the smaller the velocity factor VF [1], the smaller the wavelength λ [m]. The table below shows the approximate velocity factors for different insulation and PCB materials and different transmission line types [2.2, 2.3, 2.4].

 

The table below shows some rough approximations of wavelengths in different conductors (cables, PCBs) compared to free-space (air). Possible assumptions for wavelength λ in PCBs (FR-4) and cables are:

  • PCB. λPCB≈0.5*λair (assumption εreff ≈3.0...4.5 → VF≈0.5).

  • Cables. λCable≈0.7*λair (assumption εreff ≈1.5...3.0 → VF≈0.7).

 
 
 

 

3  Time vs. Frequency Domain

Electrical signals - periodic or non-periodic - can be measured in the time domain (e.g. with an oscilloscope) or in the frequency domain (e.g. with a spectrum analyzer). This means that an electrical signal can be described either in the time domain or in the frequency domain. The time domain representation helps you to determine the signal integrity (ringing, reflection), whereas the frequency domain representation helps you to determine at which frequencies a signal potentially leads to radiated emissions.

Fourier Analysis.

Fourier analysis refers to the mathematical principle that every signal can be represented by the sum of simple trigonometric functions (sine, cosine, etc.). The Fourier analysis enables a transformation of a signal in the time domain x(t) to a signal in the frequency domain X(ω), where ω=2𝝅f (notation: x(t)○─●X(ω)). In other words, a Fourier analysis is a mathematical operation for calculating the frequency domain representation (frequency spectrum) of a signal in the time domain.

The picture below shows a representation of a square wave signal (1V amplitude) with the sum of only 4 harmonic sine waves and a direct current (DC) component of 0.5V. For the representation of an ideal square wave, an indefinite number of sine waves would be necessary (because the rise- and fall-time of an ideal square wave is 0sec).

The different variants of the Fourier analysis are presented in the list below. All variants have in common that they present the necessary math for converting a signal from the time to the frequency domain and vice versa: X(ω)─○x(t). Every Fourier analysis variant has its own field of application.

  • Fourier series. Time signal = continuous, periodic. Frequency spectrum = discrete. Formulas for the calculation of the complex Fourier coefficients cn of a signal x(t) are presented below [3.1]:

 

T is the period length of the signal x(t) in [sec], f0 is the first harmonic of the signal x(t) in [Hz], ω0 is the first harmonic of the signal x(t) in [rad], j=√(-1) is the complex number and n stands for the n-th harmonic frequency [0, 1, 2, ...].

  • Fourier transform. Time signal = continuous, non-periodic. Frequency spectrum = continuous. Formulas of the Fourier transform of a signal x(t) (left) and the inverse-Fourier transform (right) are given below [3.2]:

 

X(ω) is the Fourier transform (spectrum) of x(t), ω is the frequency in [rad], j=√(-1) is the complex number and t the time in [sec].

  • Discrete Fourier transform (DFT). Time signal = discrete, periodic. Frequency spectrum = discrete. The DFT is also applied to non-periodic signals in the time domain by periodically continuing the non-periodic signals in order to make them computable with the DFT. The DFT is by far the most common method of modern Fourier analysis. The Fast Fourier Transformation (FFT) is a fast algorithm for calculating the DFT (in case the bock length N is a power of two) [3.3].

 

X is the spectrum of x and X[k] is the k-th sample of the spectrum at ωk. x[n] is the n-th sample of the discrete time signal, which is sampled with the sampling time Ts [sec] and the sampling frequency fs [Hz].

  • Discrete-time Fourier transform (DTFT). Time signal = discrete, non-periodic. Frequency spectrum = continuous. The DTFT can be viewed as the form of the DFT when its length N approaches infinity [3.4].

 

X is the spectrum of xx[n] is the n-th sample of the discrete time signal, which is sampled with the sampling time Ts [sec] and the sampling frequency fs [Hz].  denotes the continuous normalized radian frequency variable [-π...+π].  is the product of the radian frequency ω [rad] and the sampling time Ts. ​

  • Laplace transform. The Laplace transformation belongs - like the Fourier analyses - to the group of integral transformations. It is mentioned here for the sake of completeness. The Laplace transform is used for system analysis (e.g. control systems, filters), whereas the Fourier transform is used for signal analysis [3.5].

 

X is the Laplace transform of x. s=σ+ is a complex variable with a real part σ and a imaginary part with the radian frequency ω [rad]. In case of σ=0, the Laplace transform reduces to the Fourier transform.

  • Z-transform. The Z-transform is the time-discrete counterpart to the Laplace transform with z = e^(sTsampling) = e^(+jω)Tsampling) [3.6].

 

X is the Z-transform of x. z is a complex variable. x[n] is the n-th sample of the discrete time signal, which is sampled with the sampling time Ts [sec] and the sampling frequency fs [Hz].

Spectra of digital waveforms.

Waveforms of primary importance in electronic circuits are clock and data signals. Digital signals in the time domain can be represented by trapezoid-shaped pulses with a period time T [sec], a pulse width tpw [sec], a rise-time tr [sec] and a fall-time tf [sec]. The two pictures below show an extract of a digital waveform and the corresponding Fourier series [3.7].

The frequency spectrum envelope of a trapezoid-waveform signal compared to an ideal square wave signal is shown below. It can be seen, that there is no -40dB drop-off for the ideal square wave. Instead, the frequency spectrum envelope of an ideal square wave drops constantly with -20dB/decade.

The picture below compares the frequency spectrum envelope curves of two trapezoid-waveforms with different pulse-width tpw [sec], but with identical periodicity T [sec], rise- and fall-time tr and tf [sec] and amplitude A [V]. It can be seen how the duty-cycle D=tpw/T of a digital signal influences the frequency spectrum.

The picture below compares the frequency spectrum envelope curves of two trapezoid-waveforms with different rise- and fall-time tr and tf [sec], but with identical periodicity T [sec], pulse-width tpw [sec] and amplitude A [V]. It can be seen that a reduction of the rise- and fall-time leads to lower amplitudes of the higher frequencies in the spectrum (and therefore to potentially lower radiated emissions).

Bandwidth of digital signals.

Bandwidth [Hz] of a digital signal means: What is the highest significant sine-wave frequency component in the digital signal? Significant in this case means that the power [W] in the frequency component is bigger than 50% of the power in an ideal square wave's signal with the same amplitude A [V] and duty-cycle D=tpw/T  [%]. A drop in 50% of the power [W] is the same as a drop of 70% in amplitude [V] or a drop of 3dB.

The rule of thumb for calculating the bandwidth [Hz] - or the highest significant sine wave frequency - of a trapezoid digital signal is [3.8]:

Where t10%-90% is the rising- or fall-time (whichever is smaller) from 10% to 90% of the slope of a digital signal in [sec].

The picture below shows the frequency spectrum envelop curves of an ideal square wave (t10%-90%=0) and a real waveform (trapezoid with t10%-90%0). The bandwidth of the real waveform can be found at f=0.35/t10%-90%. Note: It is assumed that there is no ringing in the real waveform. In case of ringing, the frequency spectrum envelope for f>1/(πt10%-90%would not drop off with -40dB/decade.

 
 
 
 

 

4  Radio-Frequency Parameters

 

This section introduces some of the most common Radio-Frequency (RF) parameters used in the field of EMC:

 

 

Reflection Coefficient Γ.

We speak of matched impedances in case the load impedance Zload is the complex conjugate of the source impedance Zsource. In radiated emission and immunity EMC testing, it is important to understand the term matching and how to quantify it. All receiver and/or transmitter antennas must be matched to their receiver and/or transmitter equipment impedance (typical Z= 50Ω).

The reflection coefficient Γ (=s11 in case of 1-port networks!) is defined as [4.4]:

 

All variables are complex numbers. Vforward is the forward voltage wave to the load and Vreflection is the reflected voltage wave by the load. Zsource is the complex source impedance and Zload is the complex load impedance. Zsource is typically the characteristic impedance Z0 or the transmission line impedance. The reflection coefficient Γ is often given in [dB]:

 

 

VSWR.

VSWR [1] means Voltage Standing Wave Ratio. The VSWR expresses the ratio of maximum and minimum voltage of a standing wave pattern on a transmission line. The VSWR can be calculated by using the reflection coefficient from above [4.5]:

 

 

Return Loss.

The return loss [dB] is the dB-value of the loss of power in the signal reflected (Preflected) by a discontinuity in a transmission line or due to a impedance mismatch. Return loss [dB] is the negative value of the reflection coefficient Γ in [dB] [4.6].

Insertion Loss [dB].

The term Insertion Loss (IL) is generally used for describing the amount of power loss due to the insertion of one or several of the following components (passive 2-port networks):

  • Transmission Line (cable, PCB trace)

  • Connector

  • Passive Filter

 

The insertion loss (IL) represents the power ratio in [dB] of the power P1 and the power P2  of the picture above. P1 is the power, which would be transferred to the load in case the source is directly connected to the load. The power P2 represents the power which is transferred to the load in case the Passive 2-Port Network is inserted between the source and the load [4.2], [4.3].

 

 

Impedance Matching Summary.

The table below shows how to convert between VSWR [1], return loss [dB] and the reflection coefficient [1]. Z0 is the "system impedance" (typical Z= 50Ω or Z= 75Ω).

 

In order to give you an idea what a good match means in terms of VSWR, reflection coefficient or return loss: we summarized all the values in this table below.

 
 
 
 
 
 

 

5  Transmission Lines

 

In EMC, it is essential to understand transmission lines and when to apply the concept of transmission lines. Why? Because once a signal interconnection line exceeds a critical length lcritical [m], impedance matching (Zsource, ZlineZload) MUST be considered to prevent reflections and ringing, thus preventing unwanted radiated emissions and bad signal quality - a signal integrity topic [5.2].

This chapter introduces you to transmission lines and their most important properties.

What is a transmission line?

A transmission line is a series of conductors, often but not necessarily two, used to guide electromagnetic energy form one place to the other [5.3]. It's that simple. The more complicated part is the math behind it (Maxwell's equations), because we do no longer consider a conductor as a lumped element (e.g. a simple R-L series circuit to emulate a electrical interconnection). Rather more, we consider the signal conductor as a transmission line through which an electromagnetic field is moved from one point to another. Transmission lines are characterized or described by its characteristic impedance Z0.

Here some common transmission line geometries [5.3]:

  • Coax: Coaxial cable.

  • Microstrip. Transmission line where the signal conductor is on the top or bottom layer of a PCB with an adjusted return path conductor (e.g. ground plane or power supply plane).

  • Stripline. Transmission line where the signal conductor is on a PCB embedded between two signal return path conductors (e.g. ground or power supply plane).

  • Balanced line. Two conductors of the same size and shape with equal impedance to ground and all other conductors (e.g. ethernet cable).

  • Waveguide. A waveguide consists of a single hollow conductor used to guide the electromagnetic energy. Waveguides are used in the gigahertz frequency range and they cannot pass direct currents (DC) signals.

 

 

When to consider a signal path as a transmission line?

Every signal interconnection is a transmission line. However, it is not necessary to threat every signal path as transmission line. Rules of thumb - if a conductor should be treated as transmission line or not - are mentioned below: one rule of thumb for the frequency domain and one for the time domain.

 

Frequency domain. A common rule of thumb, when working in the frequency domain, is the following [5.3].

  • Interconnection length llong  ≥ λ/10. Consider the signal path as transmission line, in order to minimize signal distortions and ringing due to reflections and in order to minimize radiated emissions and electromagnetic interference (EMI).

  • Interconnection length lshortλ/10. Consider the signal path as simple conductor. If an interconnection length l [m] is short with respect to the signal wavelength λ [m], it is good practice that the interconnection is considered as simple conductor with lumped-element parameters (e.g. resistor R [Ω] in series with inductance L [H]).

When determine the shortest wavelength λmin [m] in a digital signal (e.g. clock), it is necessary to know the maximum frequency fmax [Hz] of the signal by considering the rising- and falling-times (rather than the fundamental frequency). The rule of thumb for calculating the bandwidth [Hz] - or the highest significant sine wave frequency - of a rectangular digital signal is [5.1]:

Where t10%-90% is the rising- or fall-time (whichever is smaller) from 10% to 90% of the slope of a digital signal in [sec]. The corresponding minimum wavelength λmin [m] is

Where v [m/sec] is the propagation velocity and fmax [Hz] is the highest significant sinusoidal frequency in a digital signal. The frequency domain approximation for the critical length lcritical-fd [m] can be calculated with respect to the digital signal rise-/fall-time t10%-90%:

Where c [m/sec] is the speed of light (3E8m/sec), fmax [Hz] is the highest significant sinusoidal frequency in the digital signal, t10%-90% [sec] is the rising- or fall-time (whichever is smaller) from 10% to 90% of the slope of the digital signal and εreff [1] is the effective dielectric constant (relative permittivity) through which the electromagnetic wave is propagating.

Time domain. There is also a rule of thumb for the time domain [5.3]:

  • t10-90% ≤ 2*tpd. If the rise- or fall-time t10-90% [sec] of a digital signal is smaller than twice the propagation delay tpd [sec] (along the signal line), then the signal path should be considered as transmission line, in order to minimize signal distortions and ringing due to reflections and in order to minimize radiated emissions and electromagnetic interference (EMI).

  • t10-90% > 2*tpd. If the rise- and fall-time t10-90% [sec] of a digital signal are bigger than twice the propagation-delay tpd [sec] of the signal across the conductor, the signal path may be considered as simple conductor.

The time domain approximation for the critical length lcritical-td [m] can be calculated with respect to the digital signal rise-/fall-time t10%-90%:

Where c [m/sec] is the speed of light (3E8m/sec), t10%-90% [sec] is the rising- or fall-time (whichever is smaller) from 10% to 90% of the slope of the digital signal and εreff [1] is the effective dielectric constant (relative permittivity) through which the electromagnetic wave is propagating.

 

Critical length lcritical. The two rules above lead to similar values for lcritical [m], where the frequency domain lcritical-fd is smaller than the time domain lcritical-td. It is therefore recommended to go with the rule of thumb of the frequency domain: Consider an interconnection as transmission line if it is longer than λ/10.

 

Characteristic Impedance.

The characteristic impedance Z0, also called the wave impedance, is an important property of a transmission line. First, have a look at the equivalent circuit of a transmission line (the distribution parameter model of a two-conductor transmission line) [5.3]:

 

The picture shows transmission line segments of an infinitely small length dz in [m]. The parameters are defined per unit length (e.g [m]):

  • R' = Resistance per unit length [Ω/m]

  • L' = Inductance per unit length [H/m], mostly in [mH/m] or [nH/m]

  • C' = Capacitance per unit length [F/m], mostly in [pF/m]

  • G' = Conductance per unit length [S/m]

The formula for the characteristic impedance Z0 of a transmission line is defined as [5.3]:

 

where ω is the angular frequency [rad] and all other parameters are explained above. In practice, it is often adequate to describe transmission lines as lossless (R' = 0, G' = 0). In that case, the equivalent circuit is simplified to this:

 

The formula for the characteristic impedance Z0 of lossless transmission lines is defined as [5.3]:

 

Enough theory, let's think about how to calculate the characteristic impedance for your projects.

  • Cables. In case your cable contains signals which require impedance matching (e.g. CAN or Ethernet), you should only use cables which match with the required impedance (e.g. 100Ω or 50Ω). In general, it is not necessary to calculate the characteristic impedance of cables, this is already done for you by the cable manufacturer.

  • PCBs. Because every PCB design is individually configurable (type of dielectric material, thickness of copper and dielectric media etc.), it is usually necessary to calculate the characteristic impedance by yourself. Therefore, you should know the following data:

    • Dielectric media. Thickness, dielectric constant εr and loss tangents tan(δ) of the PCB substrate. Have a look at our PCB stackup and dielectric material collection here.

    • Trace. PCB trace width, thickness, height above neighbor planes (and separation distance for differential lines).

    • Solder resist. For PCB traces at top or bottom, you must know the thickness and dielectric constant of the solder resist.

There are plenty of free online and offline tools available for calculating the characteristic impedance of arbitrary PCB traces or signal pairs. However, be aware that these free tools usually use approximations and the results are only accurate in a certain parameter range. If you need accurate values use a solver, which calculates the transmission line impedance based on Maxwell's Equations (e.g. HyperLynx or Si8000). To help you even more, we wrote down some information on PCB stackups and properties of PCB materials.

 

 

Propagation Constant.

The propagation constant γ [1/m] is an important property of a transmission line. It describes the attenuation and phase shift of the signal as it propagates through the transmission line. Propagation constant γ is calculated like this [5.3]:

 

where R'=resistance per unit length [Ω/m], L'=inductance per unit length [H/m], C'=capacitance per unit length [F/m], G'=conductance per unit length [S/m],  ω=angular frequency [rad] of the signal, μ=μrμ0μ0=4πE-7 [H/m] μr=relative permeability [1], εrε0, ε0=8.85E-12 [F/m] εr=relative permittivity[1].

Propagation constant γ in can also be written as [5.3]:

 

where the real part α in [1/m] is the attenuation constant and the imaginary part β in [rad/m] is the phase constant. α and β are calculated like this [5.3]:

 

For a lossless line (σ=0) we get:

  • No attenuation α [1/m]:

  • Phase shift β [rad/m]:

 

High-Frequency Loss.

As already mentioned, the lossless line model is good enough in most cases (up to several 100MHz). However, with increasing frequency (1GHz and higher) high frequency losses may not be neglected anymore. High frequency losses are a result of [5.3]:

  • Ohmic Loss. Resulting from the resistance of the conductors. Ohmic losses are a function of the frequency f as well as of the geometry of the conductor (skin-effect!).

  • Dielectric Loss. Resulting from the dielectric material absorbing energy from the propagating electric field (heating the material). Dielectric losses are a function of frequency f, the dissipation factor (tan(𝛿)) and the dielectric constant (ε).
    NOTE: Dielectric losses do NOT depend on the geometry of the transmission line, only on the dielectric material.

 
 
 
 
 
 
 
 
 

 

 

6  Near-Field vs. Far-Field

EMI often happens due to electromagnetic radiation (see coupling paths). In order to lower the coupling of these kind of interference, it is important to unterstand the near-field and far-field.

 

Electromagnetic Field Characteristics.

An electromagnetic field is characterized by three things [6.4]:

  1. Source. The antenna (e.g. dipole/horn/loop antenna, cable, PCB trace).

  2. Media. The media surrounding the source (e.g. air, plastics, metal).

  3. Distance. The distance between the source and the observation point. Close (compared to the wavelength λ) to the source, the field properties are determined primarily by the source characteristics (wave impedance Zw). Far from the source, the field depends mainly on the medium through which the field is propagating.
    Therefore, the space around a source of radiation can be split into two regions: The near field and the far field.

For EMC emission measurements, it is important to know if a certain measurement takes place in the near-field or the far-field. Different probes and antennas have to be used accordingly.

  • Near-field. Special near-field probes are used to measure the H- or E-field.

  • Far-field. Log-periodic, biconical or horn antennas are used to measure the EM-field.

 

 

Wave Impedance.

Before we dive into the near- and far-field topic, we have to talk about the wave impedance. The wave impedance is the characteristic of a medium (air, transmission line dielectric medium etc.) in which a wave propagates.

Why is it important to know the characteristic impedance? Because the ratio of reflected and transmitted amplitude of the wave at an interface, from one medium to another, is determined by the characteristic impedance of the two media.

For ANY electromagnetic wave, the wave impedance ZW in [Ω] is defined as [6.1]:

 

E is the electric field in [V/m] and H is the magnetic field in [A/m]. 

The characteristic impedance Z0 [Ω] of a medium (NOT a transmission line, here we talk about electromagnetic waves travelling through a medium like air, insulator, metal shield, etc.) is defined as [6.1]:

 

ω=angular frequency [rad/sec] of the signal, μ=permeability of the medium [H/m], ε=dielectric constant of the medium [F/m], σ=conductivity of the medium [S/m].

For insulators and air (σ << jωε) the characteristic impedance is independent of the signal frequency [6.1]:

 

μ=permeability of the medium [H/m], ε=dielectric constant of the medium [F/m], μ0=permeability of free-space 4πE-7 [H/m], μr=relative permeability of the media [1], ε0=dielectric constant of free-space 8,85E-12 [F/m], εr=relative dielectric constant of the media.

For free-space (and approximately for air) are: μr=1, εr=1. Therefore, the characteristic impedance of free-space (air) is defined as [6.1]:

 

 

 

Wave Impedance vs. Distance.

The graph below shows the wave impedance ZW [Ω] in dependency of the distance d [m] to the source (normalized with λ/(2*π)) for an ideal Hertzian dipole (an infinitesimal element of length dl). The graph is a simplification and it should just illustrate how the magnetic field H and electric field E differ in their wave impedance ZW in the near field and that the electromagnetic field (EM-field) has a constant wave impedance ZW [6.1].

 

 

Near-Field to Far-Field Boundary.

The electromagnetic field around an antenna can be divided in three regions [6.2]:

  1. Reactive near-field

  2. Radiating near-field
  3. ​Far-field

The regions depend on the maximum linear dimension of the antenna D [m] and the wavelength of the signal λ [m]:

  • D<λ/2. Electrically small antennas. The reactive near-field is significant. For electrically small antennas, the radiating near-field and the far-field are minimal if they exist at all.

  • D>λ/2. Electrically large antennas. All three regions are significant: the reactive near-field, the radiating near-field and the far-field.

 

 

Near Field.

In the near field, the characteristic impedance depends primarily on the source and the electric and magnetic fields have to be considered separately (because the ratio of E/H is not constant).

  • The electric field E is predominant in these cases [6.3]:

    • The source voltage is high compared to the source current (E/H > 377Ω).​

    • The source impedance is high (e.g. dipole or straight wire antennas).

  • As a consequence of a predominant E-field:​​​​

    • The wave impedance near the antenna is high.

    • E-field attenuates with a rate of 1/d^3 in the near field (d = distance to source).

    • H-field attenuates with a rate of 1/d^2 in the near field (d = distance to source).

  • The magnetic field H is predominant in these cases [6.3]:

    • The source voltage is low compared to the source current (E/H < 377Ω).​

    • The source impedance is low (e.g. loop antennas).

  • ​As a consequence of a predominant H-field:

    • The wave impedance near the antenna is low.

    • E-field attenuates with a rate of 1/d^2 in the near field (d = distance to source).

    • H-field attenuates with a rate of 1/d^3 in the near field (d = distance to source).

As mentioned above, the near-field can be divided into the following two regions:

  • Reactive Near-Field. In the reactive near-field, energy is stored in the electric and magnetic fields very close to the source, but not radiated from them. Instead, energy is exchanged between the signal source and the fields. In case of D<λ/2, the reactive near-field extends until the distance d [m] from the antenna by [6.2]:

where λ is the wavelength [m] of the signal. In case of D>λ/2, the reactive near-field extends until the distance d [m] from the antenna by [6.2]:

 

where λ is the wavelength [m] of the signal and D the maximum linear dimension of the antenna [m].

  • Radiating Near-Field (Fresnel Zone). In the radiative or radiating near-field, the angular field distribution depends on distance d from the antenna, unlike in the far-field where it does not depend on the distance. Energy is radiated as well as exchanged between the source and a reactive near-field. In case of D>λ/2, the radiating near-field extends until the distance d [m] from the antenna by [6.2]:

where λ is the wavelength [m] of the signal and D the maximum linear dimension of the antenna [m]. Beyond the radiating near-field, the far-field begins.

 

 

Far-Field (Fraunhofer Region).

In the far-field, the E- and H-fields move perpendicular (orthogonal) and in phase to each other and form a plane wave.

  • E- and H-field attenuate with a rate of 1/d in the far field (d = distance to source).

  • The wave impedance in free-space (air) is 377Ω.

There is usually only a far-field region in case of D>λ/2, where λ is the wavelength [m] of the signal and D the maximum linear dimension of the antenna [m]. However, if there is a far-field, it starts roughly at the following distance d [m] from the antenna [6.2]:

 

 
 
 
 

7  Antennas & Radiation

This chapter is a brief introduction to the topic of antennas and electromagnetic radiation. We skip the math intense part around the Maxwell Equations. The formulas and statements in this chapter are applicable to the far field / free-space (not the near field), matched impedances (of antennas and receiver/transmitter equipment) and matched polarization (of the electromagnetic waves and the antenna polarization).

 

 

Antenna Fundamentals.

Phil Smith (inventor of the Smith Chart) explained antennas like this: ”A component that matches the (transmission) line to space”.

 


Isotropic Radiator and Power Density S.

First, we need to introduce two terms: isotropic radiator and power density S [W/m^2]. An isotropic radiator (isotropic antenna) radiates equally in all directions (from one single point) and therefore has no directivity. Such an antenna does only exist in theory. S represents the power density of the field around an antenna at a given distance d. The power density S of a isotropic radiator is simply the radiated power Prad by the antenna [W] divided by the surface area of the sphere [m^2] with the distance d [m] from the center of the radiator [7.1].

 

The power density Sw [W/m^2] of a plane wave is defined as product of the E-field [V/m] and H-Field [A/m]. ZW is the wave impedance 120π = 377 [Ω].

Antenna Gain G.

In EMC testing, we use directional antennas (e.g. biconical dipole antennas, horn antennas) which have an antenna gain G. The antenna gain G [1] ([dB]=log10(G[1])) is defined as the ratio of the power radiated in the desired direction of an antenna compared to the power radiated from a reference antenna (e.g. isotropic radiator or dipole) with the SAME power input (this means the antenna efficiency factor η=Prad/Pt, which considers the antenna losses, is already taken into account in gain G) [7.2]:

 

The antenna gain Gi of an ideal halve-wave (λ/2) dipole is 1.64 [1] (2.15 [dBi]), whereas the antenna gain Gd of an ideal halve-wave (λ/2) dipole is 1 (0 [dBd]) [7.2]:

 

The antenna gain compared to an isotropic radiator Gi and the antenna gain compared to a dipole antenna Gd are in the following relation to each other [7.2]:

 

Now, we are ready to calculate the often used terms: effective isotropic radiated power (EIRP, referred to an isotropic radiator) and effective radiated power (ERP, referred to a λ/2-dipole) [7.3]:

 

P= transmitter antenna input power [W], Gi = antenna gain referred to a isotropic radiator [1], Gd = antenna gain referred to an ideal λ/2-dipole [1]. The power density S [W/m^2]  at a distance d [m] from the transmitter antenna can now be written as:

 

Effective Aperture Ae.

Another useful parameter to know is the effective area of an antenna, also called effective aperture Ae [m^2]. The effective aperture represents the ratio of power Pr [W] (output power of the receiver antenna) to the power density SW [W/m^2] (power density of the plane wave which reaches the antenna) [7.2]:

 

The maximum effective aperture Aem [m^2] for any antenna can be calculated out of the signal wavelength λ [m] and the antenna gain referred to an isotropic radiator Gi [1]. The maximum effective aperture Aem refferes to the circumstance where the load impedance is the conjugate of the antenna impedance (matched impedances, which means that the maximum power transfer to the load takes place) and the incident wave and the polarization of the antenna are matched. 

 

Antenna Factor AF.

The antenna factor AF [1/m] or [dB/m] is the term which is often needed during EMC emissions testing. AF is used to calculate the received field strength E [V/m] based on the measured voltage Vr [V] at the receiver's antenna terminals. We focus on the antenna factor of E-field antennas. Therefore the antenna factor is the ratio of the strength of the electric field E [V/m] at the antenna to the voltage Vr [V] at the terminals of the receiver antenna [7.4]:

 

Now, we have everything ready to express AF [1/m] or [dB/m] as a function of only the plane wave frequency f [Hz] and the receiver antenna gain Gir [1] (in case of matched impedances and aligned antenna to the polarization of the incident wave):

 

V= voltage at receiver antenna terminals [V], Z= impedance of receiver [Ω] (typically 50 [Ω])= field strength of plane wave [V/m], λ = wavelength of plane wave signal, Gir = receiver antenna gain referred to an isotropic antenna [1], Z= plane wave impedance [Ω] (377 [Ω] for far field),  Aem = maximum effective aperture [m^2] 
If we rearrange the formula from above and replace λ=c/f (c=3E8[m/sec]), we get AF as a function of f, Zr and Gir:

 

AF is often used in [dB/m]:

 

Replace the wavelength λ=c/f, where c=3E8 [m/sec]. Now, we can express AF [dB/m] as a function of signal frequency f in [MHz] (!), receiver antenna gain Gir, and impedance at receiver measurement instrument Zr [Ω]:

 

 

Antenna Formulas: Emission Testing.

We focus here on EMC emission testing like CISPR 11 or CISPR 32. In other words: the physical quantity of interest is the E-field [dBμV/m] and the measured physical quantity is the voltage Vmeasure [dBμV] at the EMI receiver or spectrum analyzer (for [dBm] to [dBμV] conversion: look here).

When we set the receiver impedance Zr to 50 [Ω] and wavelength λ=c/f (c=3E8 [m/sec]), we get this for the receiver antenna factor AF:

 

 

The field strength E at the antenna can be calculated based on AF (which is frequency dependent!) and Vr (output voltage from the receiver antenna):

 

Let's consider the cable loss [dB] from the antenna to the measurement unit (EMI receiver, spectrum analyzer) and call the measured voltage at the receiver Vmeasure [dBμV]:

 

 

Antenna Formulas: Immunity Testing.

We focus here on EMC immunity testing like IEC 61000-4-3. In other words: the physical quantity of interest is the E-field [V/m] at a certain distance d [m] for a given transmitter antenna input power [W].

From the antenna fundamentals from above, we know how to calculate the power density S [W/m^2] at a distance d [m] for a transmitter antenna with an antenna gain Git [1] and the power Pt [W] at the transmitter antenna terminals:

 

Furthermore, we know the power density S of a plane wave in free space (far field characteristic impedance of Zw=120π [Ω]) is given by [7.1]:

 

If we combine these two formulas above, we can determine the field strength E [V/m] at a given distance d [m] from the antenna for a given transmitter antenna input power Pt [W]:

 

The required power input power of a transmitting antenna Pt [W] for achieving a desired field strengt E [V/m] at a given distance d [m] is:

 

 

Free-Space Path Loss (FSPL).

The term free-space loss (FSL) or free-space path loss (FSPL) is important to understand for radiated emission and immunity testing. The FSPL is the attenuation of the electromagnetic field (EM-field) between a transmitter and a receiver antenna. It is assumed that the space between the antennas is free of obstacles and a line-of-sight path through free-space. As mentioned at the beginning of this chapter: all the formulas are valid for the far field or in other words: free-space. In addition to that, the Friis transmission equation assumes matched impedances and matched polarizations of the transmit and receiver antennas. The Friis transmssion equation is stated as [7.5]:

 

P= receiver antenna output power [W], P= transmitter antenna input power, Git = antenna gain of transmitter antenna, Gir = antenna gain of receiver antenna, λ = wave length [m] and = distance between transmitter and receiver antennas [m]. The free-space path loss FSPL is now defined as:

 

= distance between transmitter and receiver antennas [m], λ = wave length [m], = speed of light 3E8[m/sec], = frequency [Hz].

 

Link Budget.

For radiated EMC testing, the term link budget is often used. If you want to calculate the received power Pr [dBm] for a given receiver-transmitter-setup (transmitter, frequency, antennas, distance, etc.), you calculate the link budget. In simple terms, this means:

Received Power Pr [dBm] = Transmitted Power [dBm] + Gains [dB] - Losses [dB].

 

If we rewrite the link budget formula above in a little more detailed way, we get:

 

P= receiver antenna output power [dBm], P= transmitter antenna input power [dBm], L= transmitter losses (coaxial cable, connectors, etc.) [dB], Git = transmitter antenna gain [dB], FSPL = free-space path loss [dB], Lmisc = miscellaneous losses (fading, polarization mismatch, etc.) [dB], Gir = receiver antenna gain [dB], L= receiver losses (coaxial cable, connectors, etc.) [dB].

 
 
 
 
 
 

Introduction and definitions of the term skin depth δ [m]:

  • Conductors: The skin depth δ [m] is defined as the distance from the conductor edge where the current density has fallen to 37% (37% = 1/e = 1/2.72) of the current density at the surface of the conductor Js [A/m^2]. The current density J0  [A/m^2] at distance d [m] from the conductor surface is defined as [8.2]:

 

8  Skin Effect

The resistance per unit length [Ω/m] for direct current (RDC') and for alternating current (RAC' ) of any conductor can be written as [8.1]:

resistance per unit formula

 

ρ = specific electrical resistivity of the conductor material [Ωm]. σ = specific electrical conductivity of the conductor material [S/m]. A = cross-sectional area of the conductor [m^2], Aeff = effective cross-sectional area of the conductor through which the current effectively flows [m^2].

For direct current (DC, 0 Hz), the area A is equal to the complete conductor diameter. However, for alternating current (AC) with frequency f, the magnetic fields produced by current in the conductor forces the current flow towards the outer surface of the conductor. The higher the signal frequency f, the smaller the cross-section Aeff through which the current effectively flows.

 

As a consequence of the skin effect, the resistance of a conductor increases with increasing frequency.

The diagram below shows RAC' of a round copper wire with outer diameter D=1mm and a copper PCB trace with width w=0.25mm and height h=0.035mm (1oz). The calculations for RAC' are approximations and ignore the return current path (proximity effect) and assume a single conductor surrounded only by air. However, the diagram gives an idea how the skin effect influences the resistance at higher frequencies.

  • Shielding: Imagine an electromagnetic plane wave of field strength E0 and H0 entering an absorbing material (shield). The skin depth δ [m] is the distance an electromagnetic wave has to travel through that absorbing material until its field strength is reduced to 37% of E0 or H0 (37% = 1/e = 1/2.72). This means that the power of the plane electromagnetic wave is lowered by 20·log10(0.37) = 9dB after it traveled the distance δ [m]. The attenuation of an electromagnetic plane wave is defined like this [8.2]:

 

Ed = remaining electric field [V/m] strength of a plane wave with field strength E0, after travelling distance d [m] through a medium with attenuation constant α [1/m]. Hd = remaining magnetic field strength [A/m] of a plane wave with field strength H0, after travelling distance d [m] through a medium with attenuation constant α [1/m]. Remember that γ = α+ is the so called propagation constant.

 

From above, you know that the skin depth δ [m] is defined as the inverse of the attenuation constant α [1/m] [8.2]:

 

For good conductors (with σ>>ωε), the skin depth formula can be simplified to [8.2]:

 

Where f [Hz] is the frequency, µr [1] is the relative permeability of the conductor, µ0 [H/m] is the absolute permeability and σ [S/m] is the specific electrical conductivity of the conductor material.

Below some example values of skin depths for silver, copper, gold, aluminum, nickel, iron and stainless steel 316. NOTE: Important to know for calculation of skin depth of nickel, iron, stainless steel and any other ferromagnetic metal:

  • Relative permeability μr [1] depends on the specific material and alloy (therefore, be careful when reading our table and graphic below).

  • Relative permeability μr [1] depends on the frequency f [Hz] (neglected in the table and graphic below).

 

 

 

9  Shielding

Shielding Effectiveness (SE).

Shielding effectiveness describes how good a shield blocks an incident wave (electrical field strength Ei [V/m], magnetic field strength Hi [A/m]) from transmitting through the shield. After passing through the shield, the remaining wave has a field strength of Et and Ht. The reflected wave has field strength Er and Hr.

Shielding effectiveness can also be calculated like this:

t represents the shield thickness [m], f represents the frequency [Hz], μr represents the relative permeability of the shield [1], σr represents the relative conductivity of the shield material to copper [1] (σCu = 5.8E7 S/m), δ represents the skin depth [m].

σ represents the conductivity of the shield material [S/m], σr represents the relative conductivity of the shield material to copper [1] (σCu = 5.8E7 S/m), ω and f represent the frequency [1/rad, Hz], μr represents the relative permeability of the shield [1], ε0 represents the relative permittivity (dielectric constant) [1].

Example. Shielding Effectiveness PCB (Far Field!).

The following pictures show the (far field!) shielding effectiveness of PCB ground planes with thickness 35 μm and 17.5 μm. It can be seen that a thinner ground plane shields as good as a thicker ground plane for frequencies below 10 MHz.

Example. Shielding Effectiveness Cable (Far Field!).

The following pictures show the (far field!) shielding effectiveness of a typical cable shield with thickness of 1.5 mil. It can be seen that the shielding effectiveness SE for silver and aluminum is better than compared to tin at frequencies f > 1MHz.

Near Field Shielding.

Let's assume a shield is placed at a distance d [m] in the near field of a noise source and you would like to know the shielding effectiveness. Calculating the shielding effectiveness in the near field area of a source is much more difficult, than in the far field. This is due to the fact, that the wave impedance in the near field is difficult to determine and it changes significantly with distance d (either with factor 1/d^2 or 1/d^3, more details here).

In order to calculate the shielding effectiveness in the near field of a source, it must be known if the source is a ...:

  • Electric field source. High wave impedance Zw, Hertzian electric dipole. Examples: wires, PCB traces, cables, spark gaps (e.g. DC motors).

  • Magnetic field source. Low wave impedance Zw, magnetic loop dipole. Examples: current loops, transformers, wireless charging devices.

From above, we know that the shielding effectiveness is the sum of reflection loss R [dB], absorption loss A [dB] and multiple-reflection loss M [dB]. In the following paragraphs the absorption and reflection loss of near field shielding are discussed in more detail.

 

Absorption Loss.

The absorption loss A is unaffected by the type of source (near field, far field, electric/magnetic field source).

 

Reflection Loss.

The reflection loss R is calculated differently to the far field and depends on the type of source.

  • Reflection Re [dB] loss of near field, electric field sources:

  • Reflection Rm [dB] loss of near field, magnetic field sources:

σr is the relative conductivity to copper [1], μr is the relative permeability [1], f is the signal frequency [Hz], d is the distance between source and shield [m].

The graphic below shows some interesting facts:

  • Reflection loss R​[dB] of electric field sources in near field increases with decreasing distance d [m] to the source. Re is considerably higher compared to the plane wave reflection loss R.

  • Reflection loss R​[dB] of magnetic field sources in near field decreases with decreasing distance d [m] to the source. Rm is considerably lower compared to the plane wave reflection loss R.

Magnetic Field Shielding @ Low-Frequencies.

From above we learnt that the most difficult field to shield is a low-frequency magnetic field. This is due to the fact that the reflection loss R and the absorption loss A are low for low-frequency magnetic fields.

There are these two methods for shielding against low-frequency magnetic fields:

  • Shield with μr >> 1. Use of a low-reluctance shield material with a high magnetic permeability μr >> 1 (e.g. nickel, Mu-metal).
    NOTE: The permeability decreases with increasing frequency f [Hz] and with increasing magnetic field strength H [A/m] or magnetic flux Φ [Wb] respectively.

  • Shorted turn. Use a loop conductor which is placed in the magnetic field. The induced current iind in the loop conductor will generate a counter magnetic field which will then lower the magnetic field in the vicinity of the loop.

Slots and Apertures.

Above, we always assumed a perfect solid shield. Slots and apertures are efficient radiators (yes, radiators!) when their maximum linear dimension (not area!) l [m] is equal to λ/2. Therefore, if a slot or aperture has a linear dimension l of λ/2, the shielding effectiveness SE = 0 [dB]. Shielding effectiveness of a single slot with maximum linear dimension l [m] (which is equal or less than λ/2) and for a signal with wavelength λ [m] is:

If there are multiple apertures, the shielding effectiveness will be reduced even more. We can calculate the shielding effectiveness of a linear (not multi-dimensional!) array of equally, closely spaced apertures n [1] of length l [m], where the total array length larray [m] is less than λ/2:

In case of a multi-dimensional array of m rows (and m<!), only the n apertures of he first row have to be considered in case of shielding effectiveness reduction. In other words, the additional rows (2nd, 3rd, ...) will not lower the shield effectiveness significantly. The shielding effectiveness of a multi-dimensional array of equal size apertures will be the shielding effectiveness of one single hole, minus the shielding effectiveness reduction of the first row of n apertures (-20log10(√n)).

The shielding effectiveness reduction due to a linear array of n apertures (relative to a single aperture) is shown in the graph below. Be aware that the array length larray must be smaller than λ/2, otherwise the SE is 0 [dB].

Apertures located on different surfaces, which all look in different directions, do NOT decrease the overall shielding effectiveness, because they radiate in different directions.

The purposes of shielding against electromagnetic radiation are:

  • Lower emission. Prevent radiated electromagnetic emissions of your product.

  • Increased immunity. Protect you product from external electromagnetic radiation.

 

Before we jump into the theory of shielding, here two practical advice:

  1. Cables and wires. Every single signal which enters and/or leaves a shielded enclosure must be filtered and/or shielded. In case the cable is shielded: contact the cable shield 360º with the shielded enclosure.

  2. Slots and apertures. Slots and apertures lower the shield effectiveness. E.g. the shielding effectiveness (SE) is 0dB in case the maximum linear dimension l [m] of a slot/aperture is equal λ/2 of the radiated signal [9.1] [9.2].

The following topics are covered in this chapter:

 

 

Shielding Theory.

Electromagnetic waves are either generated by low-impedance or high-impedance sources.

  • High impedance sources. The electric field E [V/m] is predominant in the near-field. Examples: wires, PCB traces, cables, wireless applications (WiFi, FM, GPS, radar).

  • Low impedance sources. The magnetic field H [A/m] is predominant in the near-field. Examples: current loops, transformers, motors, wireless charging devices and other devices with a large number of ampere turns.

 

Shielding will only exist if the characteristic impedance of the shield Zs [Ω] is lower than the wave impedance Zw [9.3].

The lower the shield impedance Zs [Ω] compared to the wave impedance Zw [Ω], the higher the shielding effectiveness (SE). Why? Because, then the induced surface current density J0 [A/m] at the surface of the shield is getting closer to the value of the incident magnetic field strength H i [A/m] [9.3].

Assuming Zs<<Zw, the induced surface current density J0 [A/m] (at the surface of the shield barrier) is equal to the incident magnetic field Hi [A/m]. This means that for Zs<<Z we can write [9.3]:

Where H0 [A/m] is the induced magnetic field at the surface of the shielding, E0 [V/m] is the induced electric field at the surface of the shielding and Zs [Ω] is the characteristic impedance of the shielding.

While travelling through the shielding barrier, the values of E and H are attenuated due to the skin-effect (the same happens to the current density J). At a certain distance d [m] from the surface of the shield, the values of Jd, Hd and Jd can be calculated like this [9.3]:

δ is the skin-depth in [m]:

 

Where f [Hz] is the frequency, µr [1] is the relative permeability of the conductor, µ0 [H/m] is the absolute permeability and σ [S/m] is the specific electrical conductivity of the conductor material. The graph below shows how the attenuation of the electric field E and magnetic field H with respect to the distance from the shielding surface.

 
 
 
 
 

10  RLC @ High-Frequency

In order to choose the right components for EMI filters, the behavior of passive components at high-frequencies is important to understand. Therefore, we discus here the high-frequency behavior of passive electronic components.

Resistors R.

There are, roughly said, three basic classes of resistors. All of them are designed for different applications and have their own advantages and drawbacks when it comes to high frequency:

  • Wirewound. For high power applications. High inductance L.

  • Film type. General purpose low-power resistors. Low inductance L.

  • Carbon composition. High energy surge applications. Low inductance L.

Here the equivalent circuit of a resistor with the parasitic elements C (between the solder pads and internally) and L (series lead inductance).

The following graph shows a typical frequency behavior (impedance vs. frequency) of an actual resistor. With increasing frequency, the parallel capacitance C starts to dominate and lowers the impedance of the resistor. At the resonant frequency of the parasitic capacitance C and the parasitic inductance L, the impedance reaches its minimum. For frequencies higher than the resonant frequency, the inductance dominates and the impedance starts to increase again.

Inductors L.

Inductors can be categorized in:

  • Non-magnetic core. E.g. air inductors.

  • Magnetic core. E.g. ferrite beads (open loop magnetic core) and toroidal inductors (closed loop magnetic core inductors).

Non-ideal inductors have not only inductance, they have a series resistance (wire resistance) and a distributed capacitance between the windings. Here the equivalent circuit of an inductor:

The following graph shows a typical frequency behavior (impedance vs. frequency) of an actual inductor. At DC and low frequencies, the impedance is purely resisitive (Z = R). With increasing frequency, the impedance changes and the inductance dominates up to the parallel resonant frequency of L and C. For frequencies higher than the parallel resonant frequency, the capacitance dominates and the inductor becomes a capacitor.

Capacitors C.

Capacitors can be categorized by its dielectric material:

  • Ceramic. Multilayer ceramic capacitors (MLCCs) are the most widely used capacitors today. They have relatively low equivalent series inductance (ESL) and low equivalent series resistance (ESR). They are used up to several GHz (dielectric material C0G or NP0).

  • Electrolytic. Electrolytic aluminum and tantalum capacitor have high capacitance-to-volume ratio and they have quite high ESRs. They are usually used up to 25kHz...100kHz.

  • Paper. Film and paper capacitors have considerably lower ESR than electrolytic capacitors but still moderately large inductance. They are usually used up to several MHz.

An actual capacitor is not a pure capacitance C. The series inductance L (ESL) is from the leads and form the internal structure. The series resistor Rs (ESR) is also from the leads as well as from internal dielectric losses. The parallel resistor Rp represents the non-ideal dielectric material (leakage current).

The following graph shows a typical frequency behavior (impedance vs. frequency) of an actual capacitor. From DC up to the serial resonant frequency of L and C: the capacitance dominates. However, for frequencies higher than the resonant frequency, the inductance L dominates and the capacitor becomes an inductor.

 
 
 
 

11  Noise Coupling

Noise coupling is one of the basic and constantly recurring topics in EMC. First of all, we discuss the coupling paths. Then it is shown which coupling path leads typically to differential mode noise and which coupling path leads typically to common mode noise. 

Coupling Paths.

One important concept in EMC is the concept of coupling paths. To start off, let's see what parts are involved when Electromagnetic Interference (EMI) happens and why focusing on coupling paths is so important:

  1. Source. In real world, there are sources of unwanted electric or electromagnetic noise (during EMC testing, these sources of noise are artificial and as close as possible to the real world, e.g. ESD generators, burst generators, surge generators, antennas etc.).

  2. Coupling Path. The noise needs a path from the source to the victim to affect the victim. This path is called the coupling path or coupling channel.

  3. Victim. The victim is the receiver or receptor of the noise.

 

The picture above shows us, that if there are issues with EMI and EMC, you have the following three options:

  1. Lower the noise level of the noise source.

    • Emission testing. When testing the emission of your product (meaning: the product is the Source Of Noise) you have the possibility to lower the noise level by adding filters or apply the right guidelines to your electronics design (PCB, cables).

    • Immunity testing. When we talk about EMC immunity testing, it is not possible to influence the noise level, where the Source Of Noise is e.g. an ESD generator or a surge generator or an antenna radiating with a defined noise level (e.g. defined in the applicable EMC standard).

  2. Remove or make changes to the coupling path. This is where you usually have to focus if you do not pass EMC testing. Here you have the possibility to make changes and lower the emission or increase the immunity of your product, without redesigning the whole product (e.g. by improve shielding, add filters to cables and PCBs of your product).

  3. Increase the victim's immunity level with software. This may be realized with additional features in the software of your product, which make your product more robust. Examples could be software/digital filters (e.g. median), spike removers (remove spikes in signals) and sanity checks.

Let's have a closer look. There are different kind of coupling paths, some of them are conductive (galvanic) and some of them are non-conductive (radiated). The picture below shows you the four different types of coupling:

  1. Conductive Coupling (galvanic)

  2. Capacitive Coupling (near-field)

  3. Inductive Coupling (near-field)

  4. Radiated Coupling (far-field)

If you don't want to go too much into detail, here the summary. If you would like to dig deeper into the topic: just continue reading.

Conductive Coupling / Common Impedance Coupling (galvanic coupling).

Conductive coupling happens e.g. when two circuits share a common path / trace to ground or to another reference plane. Why could this lead to EMC problems?
Here an example: If now one of these circuits experiences an ESD, burst or surge pulse, a high current may flow for a short time through this common PCB trace and introduce a noise voltage to the second circuit is introduced. This is the reason why it is important having low-impedance ground planes to earth / chassis. Because with low-impedance ground planes, the conductive coupling can be minimized.
The circumstance of common ground path or power supply path often leads to a bad signal integrity as well. Let's assume the first circuit drives some power electronics and the second is a sensitive measurement circuit. The high currents of the power electronics (noise currents) introduce a noise voltage in the said common path / trace and will lead to interference on the measurement circuit.

Capacitive Coupling (near-field coupling).

Capacitive coupling is a near field coupling, meaning: the noise source coupling structure and the victim receiving structure are somehow close together (on a PCB or in a cable harness) compared to the wavelength of the interference signal. The field of concern for capacitive near-field coupling is E-field (electric field). The energy of the radiated E-field noise typically falls off with 1/r^3 or 1/r^2 (1/r3 if the E-field is predominant, 1/r2 if the H-field is predominant) in the near-field area, where r is the distance between the emitting noise source structure (typically cables or metal plates) and the receiving structure (typically cables or PCB traces) of the victim.
Generally speaking, capacitive coupling is an issue where you have noise sources (typically cables or metal plates) with fast transient signals or high frequency signals and victims with high impedance circuits (e.g. Analog to Digital Conversion (ADC) inputs).
Why could capacitive coupling lead to problems during EMC testing? Here an example: several wires are together in the same cable which means that each wire is capacitively coupled to the other wires inside that cable (the capacitance is the bigger the longer the cable and the closer the wires to each other). One of the wires drives the reset signal of your controller system. During EMC testing, an ESD pulse happens to a connector pin which is connected to a wire in that cable. If there are no measures against ESD at the connector pin, this pulse may couple capacitively into the other wires in the same cable and potentially reset your controller (in case the controller reset signal is not filtered properly).
Capacitive coupling is also a topic when it comes to signal integrity: A typical example of capacitive coupling is a clock trace (noise source) laying parallel over e.g. several cm to a sensitive analog sensor signal trace (victim). The shorter the rise- and fall-time of the clock signal, the better (worse for your circuit) is the coupling.

Inductive Coupling (near-field coupling).

Inductive coupling is a near field coupling, meaning: the noise source coupling structure and the victim receiving structure are somehow located close together (on a PCB or in a cable harness) compared to the wavelength of the interference signal. The field of concern for inductive near-field coupling is H-field (magnetic field). The energy of the radiated H-field noise typically falls off with 1/r^3 or 1/r^2 (1/r3 if the H-field is predominant, 1/r2 if the E-field is predominant) in the near-filed area, where r is the distance between the emitting noise source structure (typically current loops, coils) and the receiving structure (typically loops) of the victim.
Generally speaking, inductive coupling is an issue where you have noise sources with high-current traces / wires or large loop areas and victims with low impedance or large loop area structures (e.g. sensor signal in a cable which builds a current loop with the chassis structure as the sensor signal cable is not laid close the chassis).
Inductive coupling could lead to issues during EMC testing, e.g. because of high currents flowing through cables and PCB traces during ESD testing or surge testing where currents up to several kA could occur.
Inductive coupling may also lead to bad signal integrity as well. A typical case is a high current Pulse Width Modulation (PWM) signal line (and therefore a high energy H-field), laid parallel to a sensor signal with 4...20mA-signaling, whereas the sensor signal and its return current path build a large loop structure (where the magnetic field of the PWM signal can couple into).

Radiated Coupling (far-field coupling).

Radiated coupling is a far field coupling, meaning: the noise source coupling structure and the victim receiving structure are located far away to each other compared to the wavelength. The field of concern for radiated coupling is the electromagnetic field (EM-field), where the H-field and the E-field energy both fall off with 1/r.
During radiated immunity / susceptibility EMC testing (IEC 61000-4-3), the testing equipment antenna radiates a predefined EM-field to the Equipment Under Test (EUT). When the EUT is tested regarding radiated emissions (CISPR 11, EN 55011), the EMC test equipment antenna is placed at a predefined distance (e.g. 3m, 10m) in order to measure radiated EM-field by the EUT.

Coupling paths overview and summary.

For the quick reader, here a summary about EMC coupling paths and the corresponding physical coupling mechanisms.

 

 

Differential Mode vs. Common Mode.

Differential mode noise and common mode noise are fundamental concepts in EMC. This section here shows which coupling path or coupling mechanism leads to differential or common mode noise. First of all, here an explanation of differential mode vs. common mode on the basis of cables:

 

  • A differential mode noise current flows in different directions through a cable.

  • A common mode noise current flows in the same direction - or other words: a common direction - through a cable.

 

The pictures below, show differential mode noise and common mode noise on the basis of a singel current loop.

 

 

Differential and common mode noise sources can't be found in any bill of material (BOM) or any schematic. These noise sources are unwanted and unintended.

Typical differential and common mode noise sources are listed below:

  • Differential mode.

    • Inductive coupling. A magnetic field induces a noise current in a circuit loop.

    • Impedance coupling. Two or more circuits share a common current path (e.g. a common return current path) and the voltage drop of one of these circuits introduces a noise voltage in another circuit and vice versa.

    • Common mode to differential mode conversion. A common mode noise current can lead to a differential noise voltage in case of not 100% balanced lines - in other words: forward and return current lines of a differential signal have different impedances (see figure below: line-impedance+ and stray-capacitance+ are different to line-impedance- and stray-capacitance-).

  • Common mode.

    • Capacitive coupling. A noise current is coupled capacitively into both lines of a differential signal.

    • Electromagnetic coupling. A cable or a part of an electronic circuit acts like an antenna and receives wireless radiation e.g. from radio stations or smart phones etc.

    • Reference point noise. Raise or reduction of the voltage potential of the reference point (earth) of a circuit or a voltage potential difference of spatially divided circuits.

The pictures below illustrate the above mentioned differential and common mode noise sources and its physical coupling mechanisms.

 
 
 
 
 
 
 
 
 

12  Galvanic Series

Knowing the galvanic series is important when designing interconnections between different metals (e.g. for bonding  of shields). If you don't take care which two different metals you use, the interconnection may corrode within short time (and the shielding effect is not given any more).

The galvanic series helps us to choose the right combination of metals for an interconnection, in terms of corrosion.

  • Corrosion. The less noble metal (anode) of an interconnection of two metals experiences galvanic corrosion, in case:

    1. The two metals have a galvanic incompatibility (voltage difference too high). A difference of hundreds of millivolts is likely to result in galvanic corrosion, but only a few tens of millivolts is unlikely to be a problem.

    2. An electrolyte (e.g. water, moisture) is present.

    3. The two metals have an electrical conducting connection.

  • Rate of corrosion. The rate of corrosion depends on the moisture of the environment, how far apart the metals are in the galvanic series (further apart leads to faster corrosion, because the ion transfer is faster) and other parameters like the type of electrolyte (pH, concentration, flow rate), temperature (rate of corrosion increased with higher temperatures), degree of aeration, humidity, pressure and even the geometry of the interconnection.

  • Recommendations. Here recommendations on how to choose metals depending on their potential difference.

    • Harsh (outdoor, high humidity, salt laden, military).
      Choose metals where electrode potential difference is ≤0.15V.

    • Normal (non-temperature/humidity controlled, consumer product, indoor):
      ​Choose metals where electrode potential difference is ≤0.25V.

    • Controlled (temperature/humidity controlled, indoor):
      Choose metals where electrode potential difference is ≤0.50V.

The table below shows the standard electrode potential (E°), which is defined by measuring the potential relative to a standard hydrogen electrode (SHE) using 1mol solution at 25°C and at the pressure of 1atm.

[Sources: Taschenbuch der Physik, Horst Kuchling, 17. Edition (2001)

ATTENTION: A galvanic series can be derived for metals in any electrolyte solution. Be careful! The real world corrosion rate depends on the solution conditions like: electrolyte concentration, pH, flow rate, aeration, temperature, humidity and pressure. Therefore, it is common to use the sea water electrode potential table, rather than the standard hydrogen electrode potential table!

The sea water galvanic series is often used to approximate the probable galvanic effects in other environments for which there are no data. For example: from the standard electrode potentials shown in the table from above, it can be seen that aluminum (Al) should behave anodically toward zinc (Zn) and presumably would retard the corrosion of zinc in a usual coupled situation. However, the reverse is true as can be seen below from the established galvanic series of metals in sea water table.

In the table below, metals are grouped. All metals, alloys and platings of the same group have common electro-motive forces (EMF) within 0.05V when coupled with a saturated calomel electrode in sea water at room temperature. All members of a group regardless of metallurgical similarity or dissimilarity are considered compatible. Compatible couples between groups have been specified in the table (green areas) based on a potential difference of 0.25V maximum.

[Sources: MIL-DTL-14072F (2013-Aug-13); NASA-STD-6012 (2012-Mar-08)].

In the figure below, you see the galvanic series of selected metals in sea water. This series can be used as a reference to minimize galvanic corrosion when selecting metals that will be in direct contact. Generally said, the closer the metals in the series, the less galvanic corrosion is expected. In a galvanic couple, the metal higher in the series represents the anode, and will corrode preferentially in the environment to the cathode which is lower in the series. [Sources: MIL-STD-889C (2016-Aug-22)].

Here some free information about of galvanic corrosion by the U.S. Army and NASA (free download, unlimited distribution). You will find there concrete and helpful data about metals and suitable coatings and treatments to prevent corrosion (e.g. MIL-STD-889).

 

13  Triboelectric Series

The triboelectric effect is the reason why certain materials get electrically charged when they are rubbed against each other (e.g. if you rub a balloon on your sweater or shirt). Therefore, the triboelectric effect is one of the causes for electrostatic charges, which then lead to electrostatic discharges (ESD).

Some materials tend to give electrons up easily, whereas others readily absorb them. The triboelectric series is a list of materials with a ranking from the most positive (materials that have tendency to lose electrons) to the most negative (materials that tend to gain electrons). The strength of the produced charge (also called static electricity) due to friction depends on the type of materials involved, their surface roughness, temperature, strain, humidity and other items. Therefore, the table below does not tell you something about the magnitude of the static charge, just its polarity.

[Source of the triboelectric series: Introduction to Electromagnetic Compatibility, Clayton R. Paul, 2. edition (2006)]

Positive

Air

Human skin

Asbestos

Glass

Mica

Human hair

Nylon

Wool

Fur

Lead

Silk

Aluminum

Paper

Cotton

Wood

Steel

Sealing wax

Hard rubber

Mylar

Epoxy glass

Nickel, copper

Brass, silver

Gold, platinum

Polystyrene (PS)

Acrylic

Polyester

Celluloid

Orlon

Polyurethane (PU)

Polyethylene (PE)

Polypropylene (PP)

Polyvinylchloride (PVC)

Silicon

Teflon

Negative

 
 

14  Material Properties

Find here the most important material properties and physical constants which you will need for your work as EMC design engineer. All values are typical values and are without warranties.

Physical Constants.

  • c = 2.998E8 [m/sec] ≈ 3E8 [m/sec]. Speed of light.

  • e = 1.602E−19 [C]. Elementary charge.

  • m = 9.109E−31 [kg]. Electron mass.

  • ε0 = 8.854E−12 [F/m]. Permittivity of vacuum, absolute permittivity.

  • μ0 = 4πE−7 [H/m] = 12.57E-7 [H/m]. Permeability of vacuum, absolute permeability.

  • Z0 = √(μ0/ε0) = 376.7303 [Ω] ≈377 [Ω]. Characteristic impedance of vacuum.

Metals - Electrical Conductivity σ & Resistivity ϱ.

The material properties like specific conductivity [S/m] and specific resistivity [Ω/m] are important factors when it comes to the calculation of e.g. the skin-depth. The values in the table below apply for zero frequency (0 Hz) and room temperature (25ºC).

[Sources: Reference data for engineers, Newnes, 9. edition (2002); Electrical resistivity and thermal conductivity of nine selected AISI stainless steels, C. Y. Ho and T.K. Chu, Cindas Report 45 (1977); Temperature dependence of electrical resistivity of metals, Lazarus Weiner and Premo Chiotti and H. A. Willhelm, Ames Laboratory (1952)].

Metals - Magnetic Permeability μr.

In the field of EMC, knowing the relative permeability μr [1] of materials is important when it comes to shielding of low-frequency (f<20kHz) magnetic fields. The relative magnetic permeability μr [1] of a material tells us how much better this material is able to "conduct" the magnetic flux, or in other words, how big the flux-density B [T] in a material is compared to vacuum (where μr = 1), for a given field strength H [A/m].

First of all, let us have a quick look how magnetic materials are classified. Our main interest lies in ferromagnetic materials (soft magnetic), because they can be used for shielding of low-frequency magnetic fields.

  • Anti-Ferromagnetic. μr [1] is slightly bigger than 1. The only pure metal which is anti-ferromagnetic is chromium (Cr). 

  • Diamagnetic. μr [1] is slightly smaller than one. Diamagnetic materials are weakly repelled by a magnet. They can not be magnetized.

  • Paramagnetic. μr [1] is slightly bigger than one. Paramagnetic materials are weakly attracted by a magnet. They can not be magnetized.

  • Ferrimagnetic. μr [1] is bigger than 1, but much smaller compared to the μr of ferromagnets. Ferrimagnetic materials can be weakly magnetized ("weak magnets").

  • Ferromagnetic. μr [1] is much bigger than 1. Ferromagnetic materials can be magnetized and used as shielding material (against low-frequency magnetic fields) or for building permanent magnets. Ferromagnetic materials can be categorized by their coercivity (Hc [A/m]). A high coercivity of a material means that the external magnetic field must be high to change the polarization of the magnet.

    • Soft magnetic.​ Coercivity Hc [A/m] is low (typ. <1kA/m). Example applications: shielding, transformer and ferrite cores.

    • Hard magnetic. Coercivity Hc [A/m] is high (typ. >10kA/m). Example application: permanent magnets. 

The μr [1] data in the table below apply to room temperature and direct current (DC, f=0Hz).

NOTE: μr is temperature and frequency dependent. μr of a material may increase or decrease with increasing temperature, until a certain temperature (curie temperature) where μr=1. With increasing signal frequency, the value of μis getting smaller. E.g. Mumetal has a μr of over 10'000 at f=0Hz (DC), but similar to steel at f=20kHz!

The initial relative magnetic permeability μri [1] describes the relative permeability for low flux densities B [T]. The maximum relative magnetic permeability μrm [1] is usually by factor 2....5 (or even more) higher than μri [1]. μrm is valid at a single point in the H-B-Hysteresis diagram (at this point the change of the magnetig field H [A/m] results in the biggest change in the flux density B [T]).

[Sources: Reference Data For Engineers, 9. Edition (2002); Taschenbuch der Physik, Horst Kuchling, 17. Edition (2001); Engineering Electromagnetics, William H Hayt Jr., 5. edition (1988)]

Insulators - Dielectric Constant εr Loss Tangent tan(δ).

The dielectric constants (relative permittivity εr [1] or Dk [1]) in the table below should be considered representative for each material and they apply to room temperature (25ºC) and humidity (unless otherwiese noted).

[Sources: Reference Data For Engineers, 9. Edition (2002)].

PCB Materials - Dielectric Constant εr Loss Tangent tan(δ).

The dielectric constant (relative permittivity εr or Dk [1]) of the printed circuit board (PCB) materials determine the characteristic impedance of the PCB traces and is therefore an important parameter in the field of EMC and signal integrity. Besides the dielectric constant εr, the loss tangent tan(δ) [1] or dissipation factor Df [1] of a PCB material is also of interest, because it influences the loss of high frequency signals (together with the resistive loss, which increases with increasing frequency due to the skin effect).

The arrangement of the copper and insulation layers of a PCB is called the PCB layer stackup, or just: stackup. Let's have a look at an example of a PCB stackup of a 6 layer board.