PCB Design and Layout Fundamentals for EMC

Table of Contents

1 Introduction

1.1 Cost of EMC

1.2 PCB Design and Layout Fundamentals for EMC

1.3 Successful PCB Design and Layout is Both a Science and Art

2 System Components to Reduce Electromagnetic Interference

2.2.3 Equivalent Circuit Model

2.2.4 Use Capacitors to Filter out System Noise

2.2.5 Common Types of Capacitors and Trade-offs

2.2.6 Select Capacitor’s Value - Impedance Curve of Bypass Capacitors

2.2.7 Bypassing a System with Wide Bandwidth

3.2 Digital Circuit Layout

3.3 Analog Circuits Layout

3.4 Switching Mode Power Supply (SMPS) DC/DC Converter Placement

3.5 I/O Connectors Placement

3.6 PCB Component Placement and Routing Guideline

4 PCB Stack-up

4.1 Determining PCB Thickness

4.2 Determining PCB Layer Count

4.3 Determining PCB Stack-up

4.4 Four Layer PCB Stack-up

4.5 Six Layer PCB Stack-up

4.6 Eight Layer PCB Stack-up

4.7 Ten Layer PCB Stack-up

5 PCB Ground and Power System

5.1 PCB Ground System

5.1.1 Ground Grid for a 2-Layer PCB

5.1.2 Ground Plane(s) for a Multi-Layer PCB

5.1.3 Guard Ring on PCB-Edges (Faraday Cage) 5.1.4 Reduce Power/Ground Loop Area

5.1.5 Local Device Ground

5.1.6 Single-Point and Multi-Point Ground System 5.1.7 Ground Pin Assignment on a Connector

5.1.8 Ground Plane Boundary

5.2 PCB Power System

5.2.1 Power Decoupling and Bypass Capacitors 5.2.2 Switching Mode Power Supply (SMPS)

6 PCB Trace Layout and Routing

6.1 PCB Trace Length and Ground Plane Cut

6.1.1 Identifying Potential EMI Sources and Victims 6.1.2 Identifying Current Paths

6.1.3 Identifying Antennas

6.1.4 Identifying Coupling Mechanisms

6.2 Low Layer Count PCB Input Power Routing

6.3 Power and Ground Plane Inset

6.4 Connecting a Component to Power and Ground Planes 6.5 PCB Traces and Stripe Line

6.6 Vias and Microvias

6.7 Current Flow When Trace Change Layer

6.8 SerDes and PCIe Layout

7 Antenna Design and RF Layout

7.1 Antenna Basics

7.2 Antenna Types

7.4.3 RF Trace Layout Considerations

8 I/O Filtering, Electromagnetic Shielding, and Cabling

8.1 System I/O Filtering

8.2 Electromagnetic Shielding

8.2.1 Materials for Shielding

8.2.2 Shielding Requirements

8.3 Cable Coupling

Appendix 1 Pre-Layout Review Checklist

Appendix 2 Placement Review Checklist

Appendix 3 Routing Review Checklist

Appendix 4 EMC Design Guidelines

Appendix 5 Shielding Guidelines

Figure 1 Four Electromagnetic Interference Coupling Modes

Figure 2 EMC Cost for Different Product Stage

Figure 3 Series Termination Circuit

Figure 4 Parallel Termination Circuit

Figure 5 RC Termination

Figure 6 Thevenin Termination Circuit

Figure 7 The Capacitor’s structure

Figure 8 Equivalent Circuit Model for a Capacitor

Figure 9 TDK C1608JB0J106M Impedance vs Frequency Characteristics Curve Figure 10 Impedance of a Real Capacitor (Not drawn to Scale)

Figure 11 Impedance curves for a variety of capacitance values

Figure 12 Impedance of an actual capacitor in different SMD package

Figure 13 Impedance of three capacitors, same SMD packages

Figure 14 Impedance of three capacitors, different SMD packages

Figure 15 Feed Through capacitor

Figure 16 L-Circuit Filter

Figure 17 PI-Circuit Filter

Figure 18 T-Circuit Filter

Figure 19 Relay Transient Suppression

Figure 20 DC Switch Transient Suppression

Figure 21 Transformer DC Transient Suppression

Figure 22 A multi-layer printed circuit board

Figure 23 An Example of Segregation on a Single PCB

Figure 24 Crystal/Oscillator placement

Figure 25 Minimizing Microprocessor System Bus Length

Figure 26 Resistance and Inductance as Functions of Frequency

Figure 27 System Power Segregation

Figure 28 Use PCB as Heatsink

Figure 29 Thermal Vias

Figure 30 Thermal Relief Pad

Figure 31 Transistor Circuit Routing

Figure 32 Place All Connectors on One Side of the Board

Figure 33 A four Layer PCB Board Structure

Figure 34 Copper Balance on an 8-layer PCB

Figure 35 A Traditional 4-Layer PCB Board

Figure 36 A Traditional 6-Layer PCB Board

Figure 37 A Traditional 8-Layer PCB Board

Figure 38 A Traditional 10-Layer PCB Board

Figure 39 Ground Grid for a 2-Layer PCB

Figure 40 Guard ring in a 4-layer PCB (Side View)

Figure 41 Creating a ’Faraday’s Cage’

Figure 42 2-Layer PCB Power/GND Trace Route to Reduce Loop Area

Figure 43 Ground Area under High Speed IC

Figure 44 Single-Point Ground Scheme

Figure 45 Multi-Point Ground Scheme

Figure 46 Use of Interleaved Ground Pins – Differential Signals

Figure 47 Use of Interleaved Ground Pins – Single Ended Signals

Figure 48 Establishing Ground Plane Boundary

Figure 49 Power System’s Star Point

Figure 50 Primary Loop Area

Figure 51 Secondary Loop Area

Figure 52 Layout and Routing Scenarios – Minimum Trace Length (Not Drawn to Scale) Figure 53 Layout and Routing Scenarios – Plane Cut with Minimum Trace Length

Figure 54 Layout and Routing Scenarios – Plane Cut with Longer Trace Length

Figure 55 Signal Return Current Path without Plane Cut – High and Low Frequencies Figure 56 Signal Return Current Path with Minimum Signal Length – High Frequency and Plane Cut

Figure 57 Signal Return Current Path with Longer Signal Length – High Frequency

Figure 58 Printed circuit board trace layout under a heat sink

Figure 59 Example of current-driven coupling on a circuit board

Figure 60 Low Layer Count PCB Input Power and I/O Signals Routing (Wyatt, 2019) Figure 61 Mini-Plane under Intelligent Device and Trace Routing in Triplets (Beeker, 2017) Figure 62 PCB Power Plane (Vcc) and Ground Plane (GND) Inset

Figure 63 Connecting Component to Power Plane (VDD) and Ground Plane (0V)

Figure 64 Examples of Trace Impedance Discontinuities

Figure 65 A Shield Stripe-Line on PCB

Figure 66 Microvias used in High density Interconnect Technology (Rumy, 2019)

Figure 67 Current flow when changing layers for one side of the reference plane to the other Figure 68 PCIe Interconnect for Transmitter and Receiver

Figure 69 A RF Radio System Building Blocks and RF Characteristics

Figure 70 Dipole Antenna Basic (courtesy from DesignNews)

Figure 71 Quarter-Wave Antenna

Figure 72 PCB Trace Antenna (Courtesy from IoTBits.com)

Figure 73 A Bluetooth BLE Module with Chip Antenna

Figure 74 Whip Antenna (courtesy from Wikiwand)

Figure 75 IFA Antenna

Figure 76 IFA Layout (Courtesy from Cypress Semiconductor)

http://www.cypress.com/go/AN91445

Figure 77 S11 of the IFA (Return Loss = -S11)

Figure 78 MIFA Layout (Courtesy from Cypress Semiconductor)

Figure 79 S11 of the MIFA (Return Loss = -S11) (Tapan Pattnayak, Guhapriyan

Thanikachalam)

Figure 80 Chip Antenna 2450AT18B100E (courtesy from Johanson Technology)

Figure 81 Layout Guideline for 2450AT42B100E Chip Antenna(courtesy from Johanson Technology)

Figure 82 Cross-Sectional View of Microstrip Line

Figure 83 Cross-Sectional View of a Strip-Line with Bottom Ground Plane

Figure 84 System I/O Filtering

Figure 85 Electromagnetic Shielding Inside a Mobile Phone (Aimonen, 2009)

Figure 86 Shield Effectiveness Curve of Copper with a thickness of 0.02 in (Sweeney, 2012) Figure 87 Skin Depth

Figure 88 Skin Depth Chart (courtesy from Wikipedia)

Figure 89 Electric Field Coupling/Shielding

Figure 90 Two aperture patterns in a shielded enclosure (Right side pattern is better) Figure 91 Seams in shielded enclosures

Figure 92 Common Mode Noise

Figure 93 The Interface Ground Reference

Tables

Table 1 Summary of Line Termination Methods

Table 2 Capacitor Dielectric Constant for Different Materials

Table 3 Capacitor Self-Resonant Frequencies

Table 4 Capacitor Types and Characteristics

Table 5 EMI Filter Selection Based on Source/Load Impedance

Table 6 Diode Characteristics

Table 7 An Example for 4-Layer PCB Stack-up and Thickness

Table 8 An Example for 6-Layer PCB Stack-Up and Thickness

Table 9 An Example for 8-Layer PCB Stack Up and Thickness

Table 10 An Example for 10 Layer PCB Stack up and Thickness

Table 11 Printed Circuit Board Objects That May be Parts of a Good Antenna

Table 12 IFA Trace Width (W) vs Thickness between Antenna Layer and Ground Layer Table 13 Impedance Control for Traces

Table 14 Skin Depth at High Frequencies

Equations

Equation 1 The Capacitance Equation

Equation 2 The Capacitor Charge Equation

Equation 3 The Capacitor Current Equation

Equation 4 Impedance of a Real Capacitor

Equation 5 Inductance Equation

Equation 6 Ohm’s Law for an Inductor

1 Introduction

Electromagnetic compatibility is defined as a state that when all electronicdevices in the system are able to function properly without interferenceeach other or causing malfunction while the malfunction is due to EMIinside the system or from outside environment There are three essentialelements to an Electromagnetic compatibility (EMC) problem There must

be a source of electromagnetic energy, a receptor that cannot functionproperly due to the electromagnetic energy, and a path between them thatcouples the energy from the source to the receptor Each of these threeelements must be present although they may not be readily identified inevery situation Electromagnetic compatibility problems are generallysolved by identifying at least two of these elements and eliminating (orattenuating) one of them

Sources of electromagnetic interference (EMI) include radio transmitters,electronic circuits, power lines, lightning, electric motors, arc welders, solarflares and just about anything that utilizes or creates electromagneticenergy Potential receptors of EMI include radio receivers, electroniccircuits, appliances, and just about anything that utilizes or can detectelectromagnetic energy

Methods of coupling electromagnetic energy from a source to a receptor fallinto one of four categories:

Conducted (electric current)

Capacitively coupled (electric field)

Inductively coupled (magnetic field)

Radiated (electromagnetic field)

The four electromagnetic interference coupling modes are shown in

following Figure

Figure 1 Four Electromagnetic Interference Coupling Modes

EMI coupling mechanism often utilizes one or a combination of these fourmodes making the coupling path difficult to identify even when the sourceand receptor are known There may be multiple coupling paths and stepstaken to attenuate one path may enhance another

1.1 Cost of EMC

The most cost-effective way to design for EMC is to consider the EMCrequirement at the early stages of the design, please refer to Figure 2

Figure 2 EMC Cost for Different Product Stage

It is unlikely that EMC will be the primary concern when the designers startdesign the circuit, chooses the components, and performing the PCB layout.But if the concepts and suggestions in this book are practiced into thedesign at early stage, the possibility of high cost causing by a non-compliant hardware design, bad component choice, and poor PCB layoutcan be reduced dramatically

1.2 PCB Design and Layout Fundamentals for EMC

When designing an electronic circuit, board/PCB designers need take anumber of precautions to ensure that its EMC performance requirementscan be met Trying to fix the EMC performance once the board has beenbuilt will be far more difficult and costly There are a number of areas thatcan be addressed during the board design and PCB layout stage to ensurethat the EMC performance is optimized:

Series resistors placed at the output driver increases the output impedance,

as seen by the trace and the receiver input pin Resistor value can bearranged such as the driver output impedance plus resistor value matchestrace impedance to reduce signal reflection They are the preferred methodfor microcomputer-based systems to reduce system noise and EMI

Bypassing the system noise at the IC’s power pin is a critical aspect of thePCB design process Put a group of capacitors with proper values canensure that the power supply pins see low ac impedance across a wide band

of frequencies

One of the main area that need to be concerned for EMC compliance is the

RF radiated emissions arising from connecting cables to system I/Oconnectors and the susceptibility to receiving interference from outsideworld It is found that I/O cabling form the major coupling path forinterference in any product Often these cables need to carry high frequency

clocks, differential pair of data, and this can present some challenges interms of improving their EMI performance.

Any cable will radiate and receive high speed signals energy, especiallywhen cable length approaches a quarter wavelengths, or odd multiple of thesignal, because it forms a resonant circuit However, even when the cabledoes not approach these lengths, electromagnetic compatibility still can be aproblem

One solution is to filter the cables which carry signals entering and leavingthe electronic unit While this does reduce the level of EMI, it may alsodegrade the performance of the circuit If high speed data needs to becarried, then any sharp edges shall be removed by EMI filters In the worstcase, the signal may be attenuated to such a degree that the system does notwork Thus a careful balance shall be made for the filter between theequipment performance and the electromagnetic compatibilityrequirements

PCB Circuit partitioning

Circuit partitioning is important to ensure that the PCB can pass its EMCtest for production release It must be accomplished at the very earlieststages of the design to make sure that it governs the whole topology of thecircuit and the mechanical construction

The first step of the circuit partitioning process is to segregate andseparation of PCB board into EMC critical and non-critical areas The EMCcritical areas are those areas which contain sources of radiation, or may besusceptible to radiation These areas may include circuits containing highfrequency digital circuits, low level analogue circuits and high speed logicincluding microprocessor circuits

The EMC non-critical areas are those which contain areas that are unlikely

to radiate signals or be susceptible to radiation, circuits including linearregulators (not switch mode power supplies), slow speed circuits, etc

Once the circuit segregation and separation has been completed, the layoutfor the design can be started The critical or sensitive regions can beshielded or filters added as necessary at the interfaces to prevent EMI beingradiated, or to protect these circuits from the effects of EMI

After the EMC critical areas are isolated, it is easier to add the relevantmeasures at the initial stages of the design An interface can be added tooptimize the overall performance to pass its EMC test

PCB Grounding

The grounding scheme within a PCB is very important for its EMCperformance Poor grounding can lead to earth loops that can in turn lead tohigh speed signals being radiated, or picked up within the unit and hencepoor electromagnetic compatibility performance results

To help ensure that PCB grounding system works satisfactorily, it is worthbearing in mind its function It can be said to be a path that enables acurrent to return to its source It should obviously have low impedance, and

it should also be direct Any loops or deviations may give rise to spuriouseffects that can give rise to EMC problems

Designing a good PCB grounding system is not a trivial task It is morechallenging than it appears, but it is essential for the EMC performance.PCB Trace lengths must be kept to a minimum, because even if frequencies

is more than only a few kilohertz, the trace impedance may be dominated

by stray inductance and lengths of a few centimeters longer will make asignificant difference To overcome these effects, thick wires should beused if possible,

Ground planes should be used on printed circuit boards Critical tracks must

be run above the ground plane, and they should be routed so that they donot encounter any breaks in the ground plane Sometimes it is necessary to

have a slot or break in a ground plane, and if this occurs a critical trackmust be routed over the plane, even if it makes it slightly longer These andother approaches can be adopted to ensure that the grounding system is able

to reduce the EMI problems to a minimum Considerable thought should begiven to the grounding, as it may not be easy to change at a later time

EMI filters

In order for a PCB or electronic device to pass its EMC testing and gain itsEMC compliance, it is necessary to incorporate various EMI preventiveelements into the design By designing the circuit to meet the EMCrequirements it is possible to significantly reduce the levels of unwantedsignals entering and leaving the unit One of the major ways is to use aboard mount EMI filter or a series of filters

The EMI filters may categorized into two main types One is where theunwanted energy is absorbed by the EMI filter The other type of filterrejects the unwanted signal and in this case it is reflected back For EMIfiltering applications, the absorptive type is preferred

In order to enable the unwanted signals and noise to be removed, EMIfilters need to be placed in the circuit on proper place The idea is that theinterfering signals generally have a frequency higher than that of the signalsnormally travelling along the wire By having a low pass filter as the EMCfilter, only the low frequency signals are allowed to pass, and the highfrequency interference signals or noise are removed

Although integrated circuits may be well screened to prevent any signalradiated or being picked up by the circuit itself, there are alwaysinterconnections to and from the integrated circuit This interconnection canconduct unwanted signals into and out of the IC unit If the device is to beable to meet its electromagnetic compatibility, and pass its EMC testing, it

is necessary to reduce the levels of unwanted signals that can enter or leavethe unit via its interconnections

There are other ways in which panel mount EMI filters can be incorporatedinto a system from a mechanical viewpoint They may exist as standaloneEMI filters to be fixed near to the extremities of the unit They may bemounted on the edge of the electronics board One popular method ofincorporating an EMC filter into a unit is to incorporate the filter into theconnector itself This has many advantages in terms of convenience andperformance Whatever the method used, a filter is often necessary if theEMC requirements are to be met.

PCB Shielding

From the cost point of view, shielding may be an option to a PCB board if it

is not necessary But placing the unit in a conductive enclosure that isgrounded will significantly improve the EMI performance All filtering can

be undertaken at this interface and the conductive wall will provide abarrier to radiation, thereby improving both the emissions and susceptibilityelements of the EMC performance

Where cost and possibly aesthetics are important it is possible to spray theinside of cabinets with conductive paint, although the level of screeningprovided will not be nearly as good as if a fully conductive metal case isused Where high levels of EMC performance are required care should betaken to choose a case where the continuity of the screen is not breached.The case should ideally be made of as few elements as possible At eachjoint there will be the possibility of radiation passing through Where joints

to occur they should be as tight as possible and they should have goodcontinuity between them

Some metal cases using a prefabricated style of construction with anodizedaluminum panels do not offer good EMC performance, although they areaesthetically more pleasing than some RF tight cases A balance has to bemade dependent upon the performance required and the EMC tests thatneed to be pass

1.3 Successful PCB Design and Layout is Both a Science and Art

Designing for a successful PCB layout can be straightforward for a simplecircuit, but for an extreme PCB layout, it can get complex when driven byproduct requirements, multiple layers, high count and various components,and different types of signals (e.g., high-speed, low voltage, high voltage,digital, analog, etc.) that must successfully co-exist on the same board.Also, regulations and environmental standards are also part of the PCBdesign process Creating well-behaved PCBs is both a science and an art

2 System Components to Reduce Electromagnetic Interference

Electronic component selection and PCB circuit design are two majorfactors that will affect board level EMC performance This chapterdiscusses how different type of passive devices, for example, resistors,capacitors, inductors, and diodes, can help to reduce EMI Using EMI filters

to screen/filter out EMI are also discussed in this section Each type ofpassive electronic components has its own characteristics, and thereforerequires careful design considerations

There are basically two types of packages for all electronic components:leaded and leadless Leaded components have parasitic effects, especially athigh frequencies The lead forms a low value inductor, about 1nH/mm perlead The end terminations can also produce a small capacitive effect, in theregion of 4pF Therefore, it is usually the lead length that should be reduced

as much as possible

Surface mount devices (SMD) and leadless components produce lessparasitic effects compared with leaded components Typically, surfacemount parts carries a 0.5nH of parasitic inductance and parasiticcapacitance of about 0.3pF with a small end termination From an EMCviewpoint, surface mount components is preferred, followed by radialleaded, and then axial leaded

2.1 Resistor

2.1.1 Resistor Basics

The resistor is a passive electronic component to create resistance in theflow of electric current Resistors can be found in almost all electrical

networks and electronic circuits The resistance value is measured in ohms

Add Resistor for Signal Integrity

The two common reasons that resistors are added after a driver are signalintegrity and current limiting If a resistor is placed at the driver output, itwill increase the output impedance, as seen by the trace and the receiverinput pin, thus matching the high impedance of the input pin

A mismatch in transmission line impedance formed by a PCB componentpin and PCB trace can cause reflections If these reflections are allowed tobounce back and forth along the trace for many cycles until they die out, thesignals’ "ring" may be misinterpreted either by voltage level or asadditional edge transitions, this will cause circuit operation error

An IC output pin usually has lower impedance than the trace and an ICinput pin has higher impedance If the design put a series resistor of valuematching the transmission line impedance on the output pin, this willinstantaneously form a voltage divider and the voltage of the wave fronttraveling down the line will be half the output voltage At the receiving end,the higher impedance of the input essentially looks like an open circuit,which will produce an in-phase reflection doubling the instantaneousvoltage back to the original But if this reflection is allowed to reach back tothe low-impedance output of the driver it would reflect out of phase andconstructively interfere, subtracting again and producing ringing Instead it

is absorbed by the series resistor which its value is selected to match theline impedance This source termination methodology works very well inpoint-to-point connections

Add Resistors for Current Limiting

Current limiting is another reason for adding a serious resistor in a circuit.CMOS IC technologies of different generations have different optimaloperating voltages, and may have damage limits set by the tiny physicalsize of the transistors Additionally, they cannot natively tolerate having aninput at a higher voltage than their supply, so most chips are built with tinydiodes from the inputs to the supply to protect against overvoltage Ifdriving a 1.8v part from a 3.3v source, it's logical to rely on those diodes toclamp the signal voltage to a safe range However, they often cannot handlethe current that can potentially be sourced by the higher voltage output, so aseries resistor is used to limit the current through the diode.

2.1.2 Resistor Type, Material and Placement

Resistors can be classified into several types based on different materials.These different materials all have their own characteristics One of the mostused resistors nowadays is the carbon film resistor, other material resistorsinclude: carbon composition resistors, metal film resistors, metal oxide filmresistors, foil resistors and wire wound resistors Surface mount resistors arealways preferred over leaded types because of their low parasitic elements

Metal film resistors are one of the most used axial resistors Usually, it willexhibit parasitic elements at relatively low frequencies; it is suitable forhigh power density or high accuracy circuits

The wire wound resistor is made by winding the metal wire around a metalcore Metal wire is used as the resistance element and metal core is used asthe non-conductive material The wire wound resistor is highly inductive Itshould be avoided in frequency sensitive applications It is best for highpower handling circuits In RC filter networks the inductive effect from theresistor must be considered because the parasitic inductance of the wirewound resistor can easily cause local oscillation

In amplifier designs, the resistor choice is very important At highfrequencies, the impedance will increase by the effect of parasiticinductance in the resistor Therefore, the placement of the gain setting

resistors should be as close as possible to the amplifier circuit to minimizethe board inductance.

In pull-up/pull-down resistor circuits, the fast switching from the transistors

or IC circuits create ringing To minimize this effect, all biasing resistorsmust be placed as close as possible to the active device and its local powerand ground to minimize the inductance from the PCB trace

In regulator or reference circuits, the DC bias resistor must be placed asclose as possible to the active device to minimize decoupling effect (i.e.improve transient response time)

2.1.3 Line Terminations

When a circuit is operating at high speeds, the impedance matchingbetween the source and destination is very important Because ofmismatching will cause signal reflection and ringing The excess RF energywill radiate or couple to other parts of the circuit, causing EMI problems.Termination of signals helps to reduce these undesirable effects

Termination not only reduce signal reflection and ringing by matching theimpedance between source and destination, but can also to slow down therising and falling edges of the signals There are several terminationmethods, each has its advantages and disadvantages Table 1 lists asummary of the termination methods

Table 1 Summary of Line Termination Methods

Series Termination

Figure 3 Series Termination Circuit

Figure 3 shows the series termination method The source terminationresistor, Rs, is added to achieve impedance matching between the source,

Zs, and the distributed trace, Z0 It can also absorb reflection from the load

Rs must be placed as close as possible to the source driver The value of Rs

is the real part in the equation: Rs = (Zo −Zs) A signal integrity simulationtool such as Mentor Graphics' HyperLynx can get a better analysis beforethe board is built

Parallel termination

Figure 4 Parallel Termination Circuit

Figure 4 shows the parallel termination method The parallel terminationresistor, RP, is added, such that Rp // ZL is matched with Z0 But this method

is not suitable for hand-held products, because of the low value of Rp(typically 50Ω), and will consume high power and requires the sourcedriver to drive a high current This method also adds a small delay by Z0L ×

Cd, where Z0L = Rp // ZL and Cd is the input shunt capacitance of the load

RC termination

Figure 5 RC Termination

Figure 5 shows the RC termination method It is similar to paralleltermination, but with addition of C1 The resistor R is same as the paralleltermination to provide impedance matching with Z0, and C1 provides thedrive current to drive the R and filter out the noise from the trace to ground.Therefore, the RC termination needs less source driver current than theparallel termination Values R and C1 depends on Z0, Tpd (round trippropagation delay), and Cd Time constant, RC = 3 × Tpd, where R // ZL =

Z0, C = C1 // Cd

Thevenin termination

Figure 6 Thevenin Termination Circuit

Figure 6 shows the Thevenin termination method It is formed by the R1pull-up and R2 pull-down resistors, such that the logic high and low canmeet the requirement of the destination load The value of R1 and R2 can bedetermined by R1 // R2 = Z0 R1+R2+ZL are such that the maximum currentcannot exceed the source driver capability If load input is open and R1 =100Ω, R2 = 100Ω, then load input voltage is:

Vref = [R2 / (R1 + R2)] x Vcc = [100 / (100 + 100)] x 3.3 = 1.65V

where Vcc is the supply voltage

2.2 Capacitor

2.2.1 Capacitor Basics

The classic definition of a capacitor is two conductive plates separated by adielectric material As charge collects on the plates, an electric field buildsacross the dielectric The amount of charge needed to create a certainpotential between the plates is referred to as capacitance and is measured inFarads Refer to Figure 7

Figure 7 The Capacitor’s structure

The capacitance can be measured by the dimensions of the plates andmaterial of the dielectric, Refer to Equation 1

Capacitance increases as the area of the plates increases since more chargecan be stored as the potential is created The distance between the platesdictates the attraction between charges stored on them As the distanceincreases, the interaction is decreased, and therefore so is the capacitance.This discussion also relates the relationship shown in Equation 2.

The charge stored on the plate is equal to the voltage potential appliedbetween plates times the capacitance, see Equation 2

The last of the basic equations involves current By definition, current is themovement of charge:

Therefore, there can only be movement of charge when the voltage

(potential between the plates) is changing In other words, if the voltagepresent is a DC voltage, there is no current flow

In summary, the size of a capacitor has a direct effect on its ability to storecharge The second determining factor of capacitance is the quality of thedielectric

2.2.2 Dielectrics

The dielectric is the material between the two conductor plates forming acapacitor It has high impedance and does not allow significant DC current

to flow from one plate to the other

Different materials used as a dielectric have varying amounts oftemperature stability, breakdown voltages and loss coefficients Thematerials shown in Table 2 are accompanied by their dielectric constant( Ꜫ ), which is the coefficient that directly relates to the capacitance of astructure through Equation 1

NA Fiber 6.00

Aluminum Electrolytic Capacitor Aluminum Oxide 7-10

Tantalum Electrolytic Capacitor Tantalum Oxide ~24

Table 2 Capacitor Dielectric Constant for Different Materials

2.2.3 Equivalent Circuit Model

The electrical characteristics of a capacitor can be defined by a seriesequivalent circuit composed of an idealized capacitance (C) and additionalelectrical components, which model all losses, an insulation resistance ofthe dielectric (RINSU), an equivalent series resistor (RESR) and an equivalentseries inductor (LESL) See Figure 8

The electrical characteristics of a capacitor are defined by:

C, the capacitance of the capacitor

RINSU, the insulation resistance of the dielectric

RESR, the equivalent series resistance, which summarizes all resistive losses

of the capacitor, usually abbreviated as “ESR”

LESL, the equivalent series inductance, which is the effective self-inductance

of the capacitor, usually abbreviated as “ESL”

Rinsu is very small, usually it is ignored from a mathematical analysis Theseother two values (RESR and RESL) represent the major amount of DC andfrequency dependent losses of the capacitive structure Capacitor type andstructure will dictate the values of these parasitic components

2.2.4 Use Capacitors to Filter out System Noise

For best EMI performance, it is important to select capacitors to have a lowESR (equivalent series resistance) value as this provides a higherattenuation to signals, especially frequencies close to the self-resonantfrequency of the capacitor in use Bypassing the system noise at the IC’spower pin is a critical aspect of the PCB design process Parallelingdifferent capacitor values helps ensure that the power supply pins see low

ac impedance across a wide band of frequencies

Decoupling capacitors

During SMPS switching, the high frequency switching noise created isdistributed along the power supply lines Decoupling capacitors can beinstalled to provide a localized source of DC power for digital devices (e.g.microprocessor, hub, etc.), and decoupling the noise to ground, thusreducing the switching noise propagating across the board Ideally, thedecoupling capacitors should be placed as close as possible to the powersupply pin to help filter out high frequency noise The low frequency noisedecoupling capacitor value should be typically between 1 µF to 100 µF Thehigh frequency noise decoupling capacitor should have typically valuebetween 0.01 µF to 0.1 µF For better EMC performance, decoupling

capacitors should be placed as close as possible to each IC, because trackimpedance will reduce the effectiveness of the decoupling function.

Ceramic capacitors are usually selected for decoupling; choosing a valuedepends on the rise and fall times of the fastest signal For example, with a100MHz clock frequency, use a 10nF capacitor; while running a 300MHzfrequency, a 3.3nF capacitor can be tried out The ESR value of thecapacitor also affects its decoupling capabilities For decoupling, it ispreferable to choose capacitors with an ESR value below 1Ω

Capacitor self-resonance

In Figure 9 the capacitor remains capacitive (blue dot line area) up to its

self-resonant frequency (blue circle) After that, the capacitor turns

inductive (red dotted line area), due to its lead length and trace inductance

Figure 9 TDK C1608JB0J106M Impedance vs Frequency

Characteristics Curve

(Courtesy from TDK:

https://product.tdk.com/info/en/contact/faq/faq_detail_D/1432773353108.html )

Table 3 lists the self-resonant frequency for two types of ceramic capacitors,one with standard 0.25 inch leads with interconnect inductance of 3.75nHand the other surface mount with interconnect inductance of 1nH We seethat the self-resonant frequency of the surface mount type is double that ofthe through hole type

Table 3 Capacitor Self-Resonant Frequencies

Another factor that affects the effectiveness of the decoupling capacitor isthe dielectric material of the capacitor Two common materials are used inthe manufacture of decoupling capacitors: barium titanate ceramic (Z5U)and strontium titanate (NPO) Z5U has a larger dielectric constant; with aself-resonant frequency from 1MHz to 20MHz NPO has a lower dielectricconstant, but a higher self-resonant frequency Therefore, Z5U is moresuitable for low frequency decoupling, while NPO is good for decoupling atover 50MHz

A common practice is to use more than one decoupling capacitors placed inparallel This configuration provides a wider spectral distribution to reducethe switching noise induced by the power supply networks Multipledecoupling capacitors connected in parallel can provide 6dB improvement

to suppress RF currents generated by active device switching

The multiple decoupling capacitors not only provide wider spectraldistribution, but also provide greater trace width such that to reduce lead

inductance Therefore, it will significantly improve the effectiveness ofdecoupling The value of the two capacitors should differ by two orders ofmagnitude to provide effective decoupling (e.g 0.1μF parallel with0.001μF) A low ESR value is more important than the self-resonantfrequency because a low ESR value can provide a lower impedance path toground such that to provide adequate decoupling by the capacitor when thecapacitor becomes inductive when over the self-resonant frequency.

2.2.5 Common Types of Capacitors and Trade-offs

Selecting a right capacitor or a group of capacitors is not easy due to theirmany types and behaviors Nonetheless, the capacitor or group of capacitors

is the components that can solve many EMC problems

The difference in frequency response of different dielectric materials mean

a type of capacitor is more suited to one application than another.Aluminum and tantalum electrolytic types dominate at the low frequencyend, mainly in reservoir and low frequency filtering applications In themid-frequency range (from kHz to tens of MHz) the ceramic capacitordominates, for decoupling and higher frequency filters Special low-lossceramic and mica capacitors are available for very high frequencyapplications and microwave circuits

The materials and structure of a capacitor will dictate its attributes, likemaximum voltage, temperature stability, parasitic, cost and size Followingparagraphs describe the characteristic on most popular type capacitors

Ceramic capacitors are the most common capacitor type They are

inexpensive, offer a wide range of values and high voltage, and providesolid performance They are usually quite low in parasitic inductance, andparasitic resistance is not usually an issue Ceramic capacitors are "non-polarized" Check out ceramic capacitors’ temperature coefficients beforeapplying in the application

COG/NPO (COG is EIA letter code, NPO is ceramic name, meansnegative-positive zero) type ceramic capacitors are the most stable with

temperature, but usually are available only in the pico-farad to nano-faradrange Special temperature coefficient ceramic capacitors are available forspecial applications, for example, compensation circuits.

Tantalum capacitors are type of electrolytic capacitor It consists of a

pellet of tantalum metal as anode, covered by an insulating oxide layer thatforms the dielectric, surrounded by conductive material as a cathode.Tantalum capacitors are the main use of the element tantalum

Tantalum capacitors are used mostly in low-voltage systems Itdistinguishes itself from other capacitors in having high capacitance pervolume and weight Tantalum capacitors have lower equivalent seriesresistance (ESR), lower leakage, and higher operating temperature thanother electrolytic capacitors, although other types of capacitors are evenbetter in these regards Tantalum capacitors are considerably moreexpensive than any other commonly used type of capacitor, so they are usedonly in applications where the small size or better performance areimportant

Tantalum capacitors are polarized; it distinctly marked positive andnegative terminals The low leakage and high capacity of tantalumcapacitors favor their use in sample and hold circuits to achieve long holdduration and some long duration timing circuits where precise timing is notcritical They are also often used for power supply rail decoupling inparallel with film or ceramic capacitors which provide low ESR and lowreactance at high frequency

Aluminum electrolytic capacitors are found in many applications such as

power supplies and computer system motherboards They are a commonchoice for low-to-medium frequency systems, and are made withcapacitance values from 0.1 μF to 2.7 F Typically they are rated from 4V to

630 V Aluminum electrolytic capacitors are polarized

The disadvantages of the electrolytic come from the electrolyte, theelectrolyte will dry up in time, causing the capacitors to gradually decrease

in capacitance Pushing the capacitor beyond its ratings (voltage, polarity,

or ripple current) will increase the pressure in the capacitor until it eithervents (loses electrolyte) or explodes The other problem is that if theelectrolytic capacitor is not used for a long time, the dielectric becomesthinner, decreasing the voltage it can withstand Electrolytic suffer fromaccelerated aging at elevated temperatures A rule of thumb is that their life

is cut in half for each 10 degree Celsius rise above ambient (25C) For allthese reasons, Aluminum electrolytic capacitors have a limited life and theuser may expect to replace them at some time in the future Electrolyticcapacitors also have a substantial amount of leakage

OSCON is an aluminum type capacitor with an organic semi-conductive

type electrolyte

Some of the superior features of the OSCON type capacitor:

Very low ESR and low ESL

ESR is stable over a wide temperature range

Wide temperature range -55 degrees C to 105 degrees C

Wide frequency range

High ripple current capability

Long life

Because of the properties, OSCON have the best quality available for thehighest price tag If you have the budget, these capacitors will providequality bypass for any circuit A much smaller value of capacitance can beused as compared to a normal aluminum electrolytic capacitor Inductorscan be reduced or eliminated Also small bypass type capacitors can often

be eliminated in an application circuit and the noise will be reduced by theOSCON The OSCON can be cost effective when designed into the circuitfrom the beginning in order to best utilize the unique properties OSCONcapacitors are polarized

Mica is a group of natural minerals Silver mica capacitors are capacitors which use mica as the dielectric Mica is in general very stable electrically,

mechanically, and chemically Mica capacitors are non-polar It has a

dielectric constant in the range of 5–7 Mica’s thermal, electrical, andchemical properties make for excellent capacitors Its capacitance changewith temperature ranges from ± 500 ppm/°C to 50 ppm/°C, depending onthe construction technique Mica capacitors exhibit very little voltagedependence, with dC/dV less than 0.1% Mica capacitors have high Q value

or conversely small power factors that are quite independent of frequency.This, combined with low inductance designs, result in capacitors that areideal for high frequency and RF filter applications Specification sheets formica capacitors usually show parameters plotted into the gaga hertz range

Film Capacitors, as a category, include any capacitor type made from

plastic materials: polyester, polypropylene, polystyrene, and so on Thereare two types of film capacitors Film-foil capacitor and Metalized filmcapacitor

Film-foil capacitor is constructed by layers of plastic film dielectric andwound alternately with metal foil electrodes Metalized film capacitors areconstructed of film dielectrics on which the metal electrodes have beenpreviously vapor-deposited

They are metalized film capacitor values from around 0.5nF to severalmicrofarads, and voltages from around 10VDC to several thousand VDC.For special applications sizes up to several thousand microfarads areavailable, but they are very large and very expensive Most film capacitorshave good to very good temperature stability, most are low in dielectricabsorption

Film capacitors are non-polar and have good AC response Self-inductanceranges from low to high, depending on the geometry of the construction Intheir simplest form, they are constructed from two pieces of foil separated

by a layer of dielectric material This is called film and foils construction,and is larger and more rugged in most cases Some film caps have ametallization layer, usually aluminum, deposited on both sides of the film.This is called metalized film, and is usually smaller and more expensive.Both types have similar performance

Following Table (Table 4) summarizes different type of capacitors andcharacteristics described above.

Table 4 Capacitor Types and Characteristics